particular occurrence in the future. Given the frequent
reliance of these methods on past experience, which in
turn requires bothexplicit and tacit assumptions
regarding the stability of relationships, the ability of
forecasting to generate long-term results and account
for unforseen events remains limited. Even short-term
forecasting can only factor in known relationships that
appear as identifiable trends, and building on these give
a picture of what may occur if change occurs along
predictable lines. The assumptions are basically those of
equilibrium and stability, in contrast to the dynamic
complexity of turbulent systems perspectives (
Laws,
Faulkner, & Moscardo, 1998
).
A number of researchers (
Witt & Song, 2001
;
Turner
& Witt, 2001
) have acknowledged the limitations of
current forecasting techniques, particularly the difficul-
ties posed by the inability to predict irregularities such as
sudden changes in consumer taste and demand. To
overcome these shortfalls researchers such as Witt, Sohn
and Turner have sought to improve the capability of
established techniques. For example,
Turner and Witt
(2001)
found that structured time series models incor-
porating explanatory variables produced the most
accurate forecasts. The identification of relevant non-
economic variables as determinants for future growth,
and the modelling of their significance pose a high level
of difficulty for forecasters. Uysal (1983 cited in
Crouch,
1994
) for example, noted that there were a number of
limitations confronting demand forecasting including:
ignoring supply factors, the omission of non-economic
factors which may have long-term consequences and the
potential for the appropriateness of variables to change.
To these, a range of other non-specified crises and
disasters including domestic and international political
factors, wars and insurrections, movements in the
international economy, and natural disasters suchas
earthquakes, cyclones or hurricanes should be added.
Although forecasters try to account for these situations
by using dummy variables to allow for the impact of
‘one-off’ events such as the two ‘oil crises’ in the 1970s
(
Witt & Song, 2001
), the problem of irregularities
continues to defy prediction.
A more sophisticated approach utilizing time varying
parameters (TVP) regression to model structural change
is one solution to the problem of predictive failure
encountered by causal tourism demand forecasting
models (
Witt & Song, 2001
). While the TVP approach
is able to simulate a range of shocks that may influence
the relationship between explanatory and dependent
variables it assumes that explanatory variables are
exogenous. Where there is some doubt about the
creditability of this assumption the vector autogressive
(VAR) modelling approachmay be more appropriate
(
Witt & Song, 2001
). In the VAR model all variables are
treated as endogenous.
While newer forecasting methods including TVP and
VAR allow researchers to model the impact of disrup-
tions they are still dependent on the parameters that are
selected for testing. It is at this point in the forecasting
process that a major problem can be identified. Little
consideration has been given to identifying the type of
unexpected disruptions that should be incorporated into
the current orthodoxy of forecasting.
Recognising the inability of current forecasting theory
to cope withth
e unexpected,
Faulkner and Russell
(2000)
put forward an alternative view suggesting that,
owing to ‘the certainty of the unexpected’, authorities
need to implement policies for coping withunexpected
disruptions to tourism flows. The long-standing New-
tonian paradigm of the relative stability of both internal
and external environments of organisations is an
inefficient theoretical basis for coping with change and
crises. Yet the assumption of change along Newtonian
lines underlies much of the current theory of forecasting.
History has many times revealed that the tide of human
events leans more to the chaotic than the ordered. If this
proposition is accepted, the norm of history is change
rather than equilibrium. Chaos theory talks of ‘trigger-
ing events’ suchas a crisis or disaster. These events need
B. Prideaux et al. / Tourism Management 24 (2003) 475–487
476
not be regarded as only destructive because they may
lead to new configurations or structures that are more
effective than those that are replaced.
A well-developed literature exemplified by journals
that include
Risk Analysis, Risk Management, Disaster
Planning and Prevention
and
Emergency Planning Digest
have recognised that there are a large range of events
that cannot be predicted with any certainity and which
lie beyond the range of predictions that standard
forecasting techniques can be expected to yield. Employ-
ment of scenarios as the basis for predicting the impact
of a range of disruptions is a widely accepted method of
planning for crises and disasters including multiple
environmental, economic and natural disasters. The
tourism literature has not begun to investigate the rich
range of techniques developed in the risk management
literature, yet this literature has the potential to yield
models, frameworks and theories that will assist tourism
forecasters and planners to cope witha range of
disasters and crises.
One direction that should be considered is a synthesis
between risk specification, identification and manage-
ment, and forecasting. In sucha synthesis, forecasting,
using current techniques, could be based on revised
variables determined by forward looking scenarios or
risk analysis as an alternative to the current reliance on
variables based on historical relationships. The risk
literature has demonstrated the validity of scenario
building as the basis for risk management.
Haimes,
Kaplan and Lambert (2002, p. 383)
for example state
that ‘‘
y
. It is clear that the first and most important
step in a quantitative risk analysis (QRA) is identifying
the set of risk scenarios. If the number of risk scenarios
is large, then the second step must be to filter and rank
the scenarios according to their importance’’. Ranking
of risk, where the level of probability of occurrence and
the degree of impact can be established, provides data
that can then be used as a basis for forecasting. Of
course the possible range of scenarios is large and there
is some need to rank risk scenarios by the probability of
their occurrence on a scale that must start with highly
probable through to improbable. It is also apparent that
ranking must include the flexibility to adjust the scale of
probability. Prior to the September 11 terrorist attack
on the USA, an incident of this nature could be
described as a highly improbable risk but after the
attack the level of risk moved to highly probable.
Where then, does this leave the study of forecasting
future growth trends in tourism? Where change is slow
and ordered, and therefore relatively predicable, fore-
casting may yield a high degree of accuracy as claimed
by forecasters. On the other hand, where events follow
the normal course of history and exhibit a tendency to
sudden, large-scale instability and unpredictability,
forecasting looses its potency and an alternative form
of prediction is required.
Faulkner (2001)
described a
variety of situations that could be classified as either
crises or disasters, but which were in the main single
events attributable that can be classified as either
management failures (crises) or unpredictable cata-
strophic change (disaster). Beyond individual crisis or
disaster events lie complex situations where many
factors coalesce to impact on the harmony of the
tourism industry. Rather than pursue a quest for
definitional precision there is a need to examine complex
situations where numerous crises, disasters and political
system failures act in unison to create one or more
shocks on the tourism industry. Conditions in Indonesia
between 1996 and 2002 created sucha situation where a
number of events had serious consequences for that
nation’s tourism industry.
By being able to determine the magnitude of the
problem and identify its cause as either natural or
human or a mix of the two, actions to minimise the
impact of crisis and disaster can be implemented.
According to
Faulkner (2001)
, a synthesis of the
characteristics of disaster or crisis situations based on
researchby
Fink (1986, p. 20)
,
Keown-McMullan (1997,
p. 9)
and
Weiner and Kahn (1972, p. 21)
identified the
following key factors:
*
A triggering event, which is so significant that it
challenges the existing structure, routine operations
or survival of the organisation. Trigger events may
include political crises, religious or ethnic tensions,
economic decline and climate change;
*
Characterised by ‘fluid, unstable, dynamic’ situations
(
Fink, 1986, p. 20
).
*
High threat, short decision time and an element of
surprise and urgency;
*
A perception of an inability to cope among those
directly affected; and
*
A turning point, when decisive change, which may
have both positive and negative connotations, is
imminent. As
Keown-McMullan (1997, p. 9)
empha-
sise, ‘even if the crisis is successfully managed, the
organisation will have undergone significant change.
Understanding the path of the events following a
shock may contribute to a methodology of identifying
risk situations and thereby assist in providing some
warning of when such events may occur, and how
they may evolve.
Faulkner’s (2001)
Tourism Disaster
Management Framework provided one example of an
operational model of this type. Once it can be admitted
that current techniques have limitations, particularly in
the areas of risk and uncertainty, the way is open to look
for new methods that move beyond the Newtonian
assumption of stability. Ignoring the possibility of
disruptions in the terms stated in this discussion may
lead to a prolonging of the event and exacerbation of its
effects as remedial action is considered in the heat of
B. Prideaux et al. / Tourism Management 24 (2003) 475–487
477
unexpected unfolding events, rather than in the calm of
prior contingency planning based on new forecasting
methods.
A typical large-scale disruption precipitates complex
movements away from the previous relationships which
usually trend towards stability and equilibrium. During
a situation of this nature multiple events and their
follow-on affects may prolong the period of disequili-
brium unless there is some mechanism that can assist to
re-establish a new equilibrium situation. Chaos theory
(
Faulkner, 2001
) provides an insightful paradigm for the
investigation of changing complex situations where
multiple influences impact on non-equilibrium systems.
In these conditions of uncertainty, fruitful approaches to
strategy formulation need to incorporate contingencies
for the unexpected. Chaos theory demonstrates that
there are elements of system behaviour that are
intrinsically unstable and not amenable to formal
forecasting. If this is the case, a new approach to
forecasting is required. Possible ways forward may
include political audits and risk analysis to develop a
sense of the possible patterns of events allowing these to
be factored into projections of future tourism activity
using a series of scenarios. The latter may involve the
use of a scenario building approachthat may incorpo-
rate elements of
van der Heijden’s (1997)
strategic
conversion model, elements of the learning organisation
approachbased on a structured participatory dialogue
(
Senge, 1990
) or elements of risk management described
by
Haimes et al. (2002)
. Which ever direction is taken,
there are a number of factors that must be identified and
factored into considerations of the possible course of
events in the future.
3. Factors that may influence tourism flows
Throughout recorded human history it has proved
impossible to predict the future, although many
have tried.
Faulkner (2001)
, for example, cites the
importance of oracles in classical Greece and noted
their failures. What is known of the future is that there
are a number of circumstances that may exert influence
on the course of events in following years. Events that
disrupt the tourism industry can be divided into three
groups:
3.1. Trends
Trends describe a range of possible future trends that
can be identified in the present and which, unless
remedial action is taken, will cause some magnitude of
disruption in the future. The degree of impact of these
trends will depend on how Governments and industry
respond to these trends to mitigate the worst of the
possible range of outcomes. One example is the future
impact of low fertility rates in developed countries which
may see increased tensions in the future between retirees
and workers (
Willmott & Graham, 2001
). For example,
the Japanese population is projected to decline by 17.9
million between 2000 and 2050 while the number of 60
plus persons will climb to 42% of the population
(
Sayan, 2002
). Population changes of this magnitude are
also occurring in Korea, Italy and Spain and in the
future will have a significant impact on tourism as the
national tax base falls but consumption of health
services escalates.
3.2. Crises
Crises can be described as the possible but unexpected
result of management failures that are concerned with
the future course of events set in motion by human
action or inaction precipitating the event
.
Events of this
type include the Foot and Mouth outbreak on UK
farms in 2001, the Chernobyl disaster and the Exxon
Valdez oil tanker wreck. Examples of crises that may
occur at some point in the future include:
*
The impact of AIDS particularly in Sub-Saharan
Africa and potentially in the Indian subcontinent and
the Russian Federation (
Quinn-Judge, 2001
);
*
An increase in militant religious fundamentalism;
*
Nuclear war in Asia;
*
Financial meltdowns including global recession; and
*
Terrorism employed to achieve political or religious
objectives.
3.3. Disasters
Disasters can be described as unpredictable cata-
strophic change that can normally only be responded to
after the event, either by deploying contingency plans
already in place or through reactive response.
Dursch-
mied (2000)
cites a number of examples from history
where unexpected weather turned the tide of battle
including the typhoon that saved Japan from a Mongol
invasion in 1281. More recent examples include the
Kobie earthquake, the 1997 El Nino climate effect and
the 2002 floods in Europe and China. Events of this
nature occur regularly but at undeterminable frequency,
intensity and location. Examples of future disasters that
the Tourism industry could begin to prepare for include:
*
Natural disasters of all types including floods,
droughts and earthquakes;
*
Long-term natural climate change separate from the
current concern over human induced global warming;
and
*
A pandemic perhaps caused by a new strain of flu or
other unknown disease.
B. Prideaux et al. / Tourism Management 24 (2003) 475–487
478
3.4. Change in the structure of government, social
organisation or economic structure
To these previous classes of disruptions may be added
other factors (
Prideaux, 1999
) that while neither a
disaster nor a crisis, may precipitate significant change
in the organisation of international tourism:
*
Development of new trading blocks where nations
join together in regional political and economic
unions suchas the European Union (EU);
*
The future direction of capitalism;
*
Demographic change in terms of ageing populations
in developed economies as well as growing popula-
tions in many underdeveloped nations;
*
A continuing searchfor political identity by ethnic
and religious groups causing further fragmentation in
a number of nations;
*
Sacristy, particularly of farming lands, water, marine
resources and non-renewable energy; and
*
Environmentalism, particularly if global warming
continues.
Consideration of these trends, crises, disasters and
changes to the structure of the economy or system of
government is the first step to developing some capacity
to factor disruptions into tourism forecasting. The
concept of trends does afford some degree of predic-
ability allowing the tourism industry to develop
responses prior to the impact of the identified trend.
Disasters on the other hand, can generally only be
responded to after the event, either by deploying
contingency plans already in place or through reactive
response. Crises may offer some scope for prediction
based on the premise that after a particular type of crisis
has passed, analysis of its causes should enable greater
predictability of similar problems in the future. The
impacts of single or multiple disruptions occurring
simultaneously or in a sequence may create unexpected
political, social and/or economic conditions that cause
the decline of some destinations, growth of new
destinations or radical disturbances to global tourism
flows should also be considered.
To date, forecasts of future patterns of tourism
growthh
ave used a range of econometric demand
models. However, it is apparent that there is a need to
develop new techniques that identify risk and events that
may cause future disruptions. Where the likelihood of
disruption is small, techniques of this type are not
normally required. Where history has identified areas of
political disruption, use of these techniques, in conjunc-
tion withexisting forecasting tools, may be an appro-
priate course for forecasting. For example, disruptions
sometimes generate waring signs or are initiated by
triggering events. Identification of warning signs and
trigger events may extend warning periods and allow
forecasters to produce revised forecasts using standard
TVP, VAR or similar models.
4. Role of government in the unexpected
There has been limited discussion about the mechan-
isms that may be used to assist tourism cope with the
certainty of the unexpected, and a general criticism of
the literature relating to tourism forecasting is that
limited attention has been given to the impact of
unforseen political and economic crises on policy
development.
Friedman (1999)
suggested that the
vulnerability of individual countries to shocks has
increased exponentially as a consequence of the increase
in inter-locking systems associated withglobalisation,
political alliances and modern communication technol-
ogy. Examples of researchinclude
Clements and
Georgiou (1998)
, who examined the impact of political
instability on tourism flows in Cyprus, and
Prideaux and
Kim (1999)
who analysed the impact of the Asian
financial crisis on bilateral tourism between Australia
and Korea.
Henderson (1999)
compared the impact of
the crisis on Indonesia and Thailand finding that
tourism is vulnerable to outside forces suchas economic
conditions and suggested that there is a need for a
response strategy to cope with the unexpected. Further
researchof this nature can be expected from analysis of
the impacts of September 11, 2001.
Government responses to shocks are important and
will often affect the rate of recovery of the tourism
industry, however, we find little in the tourism literature
to assist governments to prepare for the unexpected, and
cope withits impact. Aside from limited use of scenario
planning, governments rely on forecasts to develop
budgets, policies and plans in the absence of other
methods of predicting the future. Policy frameworks
enacted by government provide the incentives as well as
the constraints around which destinations must work as
they seek to attract investment and encourage visitation.
Hall (1994)
, for example, noted that there was a need to
take into account the political context within which
tourism development occurred. In an analysis of barriers
to US tourism to Africa
,
Brown (2000)
noted that
political risk, defined as risks that arise from the action
of governments or political forces and which interfere
withor prevent foreign business transactions, can
disrupt tourism flows. Risks of this nature may inhibit
the flow of international tourists for a number of
reasons including the unwillingness of foreign investors
to support or extend lending facilities, the inability of
intermediaries to undertake financial transactions and
for airlines to operate into an uncertain logistics
environment.
Richter (1995, 1999)
has published a
number of studies examining the impact of political
events, including episodes of violence, on national
B. Prideaux et al. / Tourism Management 24 (2003) 475–487
479
tourism industries. In a recent article
Richter (1999)
commented on the impact on the tourism industries of
the Philippines, Pakistan and Sri Lanka caused by
political disruption. It is salient to compare the
consequences of political tensions in these countries
withth
e recent situation in Indonesia. Comparisons
with the Philippines are particularly relevant, as
Indonesia appears to have experienced the same general
course of events that occurred in the Philippines after
the removal of Marcos from the Presidency in 1986.
According to
Richter (1999)
, risk assessment and
political audits may be useful tools to have in place to
assist nations to recover quickly from political disasters.
Where there is an element of corruption evident,
planning and management may be less adaptable
to external shocks because of rigidities that reduce
response options.
One common element that emerges from the preced-
ing discussion is the role of governments in inadvertently
creating as well as managing unforseen events. In the
case of Indonesia and the Philippines during the political
upheavals that heralded the demise of the Suharto and
Marcos administrations, the government appeared to be
the cause of the crisis. The corrupt and undemocratic
nature of the Marcos and Suharto regimes eventually
generated sufficient opposition that mass unrest even-
tually unseated the incumbent governments. In other
instances the government’s response to crisis may be a
critical element in the manner in which national tourism
industries cope withdeclining tourism flows. If govern-
ments withdraw support for tourism promotion or
tourism development the impact may be exacerbated.
Conversely, if governments assist the industry, as in the
case of Thailand during the Asian Financial Crisis
(
Henderson, 1999
), the impacts may be minimised.
Analysis of the reactions of governments through policy
mechanisms to such crises and the results of those
actions appears to be important areas that warrant
further research.
To illustrate the issues outlined above, the paper
presents a case study describing the main causes of
disruption to the inbound tourism sector of Indonesia
during 1997–2002. The difficulty of forecasting in the
face of unexpected and complex disruptions to tourism
flows is illustrated by analysing the projections of
bilateral tourism flows between Indonesia and Australia
developed by the
Tourism Forecasting Council (TFC)
(1997b, 1998, 2000, 2001)
.
5. Case study—Indonesia
During the 10 year period 1987–1997, Indonesia
achieved a 475 per cent increase in inbound tourism
witharrivals climbing from 1,060,000 to 5,036,000. A
significant feature of the growth during this period was a
shift from traditional European, US and Australian
markets to intra-regional travel from Asia, a result of
the high and sustained growth of Asian economies.
WTO (1994)
had estimated that the growth rate
throughout the 1990s would average between 13 and
15 per cent per annum, a target that appeared achievable
until a series of events disrupted the Indonesian
economy from 1997 onwards. Between 1993 and 1997
total arrivals grew by 152.4% while receipts rose from
$US3.99 billion to $US5.44 billion (
WTO, 1999
). By
1997 foreign exchange earnings from tourism accounted
for 10.2% of Indonesia’s exports (
WTO 1999
).
5.1. Factors effecting Indonesian tourism during
1997–2002
In the period 1997–2002 Indonesia experienced 10
major shocks that received widespread international
publicity and resulted in sharply reduced activity in the
tourism sector. Many of the factors listed stem from
pressures that have existed in Indonesian society for
many decades and which surfaced as a consequence of
adverse political and economic factors.
I.
Smoke haze
. Negative media reporting of the
annual smoke haze resulting from illegal burning of
forests in Sumatra and Borneo. The haze was particu-
larly bad in 1997
II.
The Asian financial crisis
. The crisis lead to a large
and rapid fall in the value of the Indonesian Rupiah
resulting in a substantial increase in unemployment,
business failure and increase in price of many imports
including a number of staple food items.
III.
Political unrest
. Associated with the fall of the
Suharto regime commenced in late 1997 political unrest
reached a peak in May 1998. Much of the unrest
emanated from students factions in Jakarta’s universi-
ties who were pressing for democratic reforms.
IV.
Ethnic unrest
. Commencing in 1997 and appar-
ently sparked by the rapid deterioration of the domestic
economy and rising unemployment, many Chinese
communities and businesses in Java were targeted by
rioters. Ethnic unrest also flared in Kalimantan in 1999
between the native Dayaks and the immigrant popula-
tion of Madurese leaving 50,000 internally displaced
Madurese.
V.
Religious unrest.
Commencing withthe attacks on
Chinese Christian communities in Java in 1997, sectar-
ian unrest spread into a number of provinces including
Ambon. The trouble continued into 2000 in the Maluku
Islands where there were frequent clashes between
Christians and Muslims. In some instances religious
unrest was related to ethnic tensions.
VI.
Rebellion and political unrest.
Separatist move-
ments have been active in Aceh, East Timor and Irian
Jayra for several decades and subject to vigorous
suppression by the Indonesian military. Suppression of
B. Prideaux et al. / Tourism Management 24 (2003) 475–487
480
the Fretilin pro independence movement in East
Timor has regularly featured in the Western media
Do'stlaringiz bilan baham: |