International Journal of Business and Society, Vol. 23 No.

2, 2022, 1086-1105

TOURISM AND ECONOMIC GROWTH NEXUS IN

MALDIVES:
ASYMMETRIC ANALYSIS AND ROLE OF ICT

TK Jayaraman
Department of Economics, University of Tunku Abdul Rahman, Malaysia

Keshmeer Makun
School of Accounting, Finance and Economics, University of the South Pacific, Fiji

ABSTRACT

The relentless spread of the Covid-19 pandemic, which began in the first quarter of 2020, aided by the
periodical emergence of new variants, is continuing to inflict unprecedented damages on the global
economy. Interruptions in international travel have led to the collapse of the tourism-related service
industry, which is the backbone of the economies of many small islands and developing states. This paper
focuses on the Maldives, an island nation in the Indian Ocean, which is one of the top ten most tourism-
dependent economies. Adopting the Solow production function, the paper examines the tourism and
economic growth relationship using nonlinear autoregressive distributed lag (NARDL) methodology. The
analysis shows that tourism earnings and economic growth have a significant asymmetric relationship. The
result further reveals that the magnitude of the negative impact of the decline of a given size in tourism
earnings on growth is of much larger magnitude than that of the positive impact of the same size rise in
tourism earnings. The negative impact of tourism is found to be more pronounced in the long run. ICT
spread, measured by the rise in mobile subscriptions, is positively associated with growth and has emerged
as a significant contingent factor. We suggest policy measures to step up the recovery progress and growth.

Keywords: Tourism, economic growth, ICT, nonlinear ARDL, Covid-19, Maldives.

____________________________________
Received: 22 June 2021
Accepted: 29 April 2022
https://doi.org/10.33736/ijbs.4859.2022

1. INTRODUCTION

International tourism, which has been contributing to the economic growth of small island
nations in the Caribbean, the South Pacific, and the Indian Ocean regions, has been hit hard by
the current Covid-19 pandemic, since mid-2020. Pre-pandemic international tourism arrivals
were at record levels in 2019 with 1.4 billion, having risen from 25 million in the pre-jet year of
1950, which are attributed to the use of information communication and technology (ICT) 1 ,
hailed as the fourth industrial revolution for speeding up and making booking of travel and

Corresponding author: Department of Economics, University of Tunku Abdul Rahman, 31900 Kampar Campus, Perak,
Malaysia. Email: [email protected]
1
The ICT has revolutionized tourism industry rendering booking for travel and accommodation and arrangement of tours much
easier, more accessible, and less expensive than ever before.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1087

accommodation by “a click of the mouse”. Similarly, world tourism receipts from travel
expenditure and stay in hotels and resorts rose from US$ 901 billion in 2009 to US$1,462 billion
in 2018 at an annual average growth of 4.6 % (United Nations World Tourism Organization
(UNWTO), 2019). Thus, tourism has raised job opportunities and improved livelihoods, notably
in the informal sector. The small and medium enterprises are now meeting various interests of
travelers, such as domestic tours to places of historical sites, handicrafts, and ethnic meals and
products. These tourism activities are now more in the hands of part-time women entrepreneurs,
which has been noted as a welcome phenomenon (UNWTO, 2019).

However, the ongoing uncontrolled spread of the Covid-19 pandemic aided by the emergence of
new variants of the virus from time to time since the first quarter of 2021 and shortages of
vaccines, have totally engulfed the globe, severely disrupting economic activities and trade in
goods and services and tourism. The year-end review by Behusudi (2020) noted that in the first
half of 2020, tourist arrivals decreased by more than 65%, compared to 8% during the global
financial crisis and 17% amid the SARS epidemic of 2003. A subsequent recent review by World
Tourism Organization shows that international tourist arrival has declined by 67% in 2020 due to
the onslaught of Covid-19 pandemic (UNWTO, 2020).

The International Monetary Fund (IMF)’s October 2020 World Economic Outlook projected the
global economy to contract by 4.4% in 2020, with the tourism-dependent economies faring much
worse. The Caribbean countries were expected to experience a decline in growth by 12% while
other Pacific Island countries would experience a much deeper fall in their GDP (International
Monetary Fund (IMF), 2020). The United Nations Conference on Trade and Development
(UNCTAD) has also cautioned that when other domestic sectors might recover, the Covid-19
would have a long-lasting effect on international tourism, because of fears of the likely
continuance of the pandemic in the developed countries and the likelihood of restrictions for an
extended period on international travel (UNCTAD, 2020). With a poor vaccination rate with
much less than 100% coverage of the population in the developing countries, it looks that the
complete global coverage would take time with a minimum of two doses followed by booster
shots. Aware of these difficulties, it was estimated by UNWTO (2020) that earnings from
tourism alone would fall to US$ 910 billion in 2020 from US$ 1.2 trillion recorded in 2019 and
the recovery to levels of the pre-Covid-19 years is unlikely by 2023.

In the light of these developments, this paper takes up a study of the Maldives, an island nation,
located in the Indian Ocean, which is one of the eight-member intergovernmental organizations,
known as the South Asian Association of Regional Cooperation (SARRC). Comprising a string
of atolls, coral reefs, and low-lying coral islands, Maldives heads the list of ten topmost tourism-
dependent countries with 58% of annual gross domestic product (GDP) coming from tourism.
The forecast by UNCTAD (2020) was that the Maldives would experience a 17% decline in its
GDP in 2020. The number of international tourist arrivals declined in 2020 by 67% from 1.7
million in 2019 to 0.6 million in 2020 2. The Maldives is no stranger to shocks, whether external
or domestic. External shocks stemmed from recessions in advanced countries, which are the main
source of international tourists. The domestic shocks are due to periodical, political instability.

2
As of April 2021, the available latest tourism statistics cover estimates of fall in international tourist arrivals and growth
estimates. There are no official estimates of tourism arrivals and total foreign exchange earnings since then.
1088 TK Jayaraman, Keshmeer Makun

These setbacks, though often of short duration, do affect economic growth but the country often
bounces back, once stability returns.

However, the recent jolt caused by the Covid 19 pandemic has substantial economic implications
due to the heavy reliance of the country on tourism with lingering fears of new variants. Our
objective is to empirically examine the tourism-economic growth relationship in the Maldives. In
doing so, we account for any possible asymmetry to disentangle the nature of the relationship
between the two key variables. In the economic literature, empirical studies are dominated by
linear models, which assumed the absence of any asymmetry. To the best of knowledge of the
authors, there is no quantitative study on tourism's impact on growth in the Maldives. The
objective of paper is to fill the gap with an econometric study focusing on the asymmetric
relationship between tourism and growth. The results of the study would provide more insights,
especially into the Maldivian economy, which critically depends on tourism. We employ the
nonlinear ARDL procedure (Shin et al., 2014) to examine this relationship. The asymmetric
analysis a more comprehensive picture of the effect of tourism. In addition, different from
previous studies, we look at the tourism-economic growth relationship in conjunction with ICT,
as a contingent factor in the travel and tourism industry (Salem & Twining-Ward, 2018).

The paper is organized along the following lines. The next section looks at the tourism sector, the
impact of Covid 19, and ICT trends in the Maldivian economy. The third section provides
economic literature on tourism and economic growth. The fourth section outlines the theoretical
framework, data, and methodology for empirical analysis. The fifth section presents the results of
the empirical study and the last section provides a conclusion with policy implications.

2. TOURISM IN MALDIVES

The island nation, situated in the Indian Ocean’s Arabian Sea and southwest of Sri Lanka and
India has about 1,192 islands, out of which only around 188 are inhabited with a population of
492,600. Most of them are made up of atolls with no land fit for agriculture, the rest being coral
reefs and low-lying coral islands. The capital city, Malé is situated on the most populous island
with the same name, where nearly 40% of the country’s population lives. Maldives, with about
90% of its territory covered by sea owes its growth over the last five decades to its service sector
(Salem & Twining-Ward, 2018), dominated by tourism.

About 126 island resorts are located on atolls. With more territorial sea than land, the country’s
attractive marine and coral reefs have been playing an important part in determining the course of
economic progress. The tourism industry is the main engine of the Maldivian economy,
generating fiscal revenue from taxes on the tourist services sector and expenditure by tourists
adding to the country’s foreign reserves (World Bank, 2019). A steady inflow of tourists in the
country averaged a growth rate of 7.2% per year from 2015 to 2019 and consequently, Maldives
was ranked globally fourth in international tourist arrivals among the top ten tourism-dependent
countries (UNWTO, 2019). In Asia and the Pacific, with tourism’s share at 59% of GDP and
60% of total employment (Table 1), Maldives continues to be the most tourism-dependent
country (World Bank, 2020c).
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1089

Available data series on tourism from 1995 reveal that international tourist arrivals steadily rose
from 31,500 in 1995 to 61,700 in 2004. A political crisis caused instability in 2005 hurting the
tourism industry, with tourist numbers decreasing to 39,500. However, there was a recovery over
the next three years. In 2008, the tourist arrivals reached a new high at 656,000. With another
bout of political instability in 2009, tourist numbers dipped again. Once the uncertainties
disappeared in 2010, the tourism industry began to flourish. In 2018, the figure for international
tourist arrivals stood at 1.5 million (UNWTO, 2019).

Table 1: International Tourism Expenditure: Contribution to GDP (%)

Country Region Percent of GDP Rank
Maldives Indian Ocean 57.9 1
Palau Pacific 42.2 2
Vanuatu Pacific 37.1 3
Fiji Pacific 25.0 4
Samoa Pacific 22.4 5
Cambodia East Asia 19.7 6
Thailand East Asia 12.9 7
Tonga Pacific 10.9 8
Marshall Islands Pacific 9.5 9
Federated States of Micronesia Pacific 7.9 10
Source: Asian Development Bank (ADB) (2021).

The success of tourism is attributed to steadily rising foreign investment in the tourism industry,
contributing to the development of hotel and resort infrastructure with the most sophisticated and
modern facilities. The latter has made the Maldives one of the most attractive island countries
attracting the rich, senior citizens from Europe and North America. Improvements in
infrastructure, quality, and affordable public services with impressive health and education
indicators (life expectancy at 77 years and a 100% literacy rate) have been pointed out as the
significant factors behind the development success of Maldives (World Bank, 2019), with
tourism as a major foreign exchange earner (Table 2).

Table 2: Maldives: Foreign Exchange Earnings in US$ billion (1995-2017)

Overseas Development Foreign Direct
Remittances Tourism Earnings
Assistance Investment
0.058 7.230 0.002 0.211
1995
(15.324) (1.812) (0.592) (52.884)
0.019 22.260 0.002 0.321
2000
(3.222) (3.573) (0.351) (51.415)
0.077 73.230 0.002 0.826
2005
(6.761) (4.555) (0.194) (71.001)
0.112 216.469 0.003 1.713
2010
(4.904) (8.363) (0.122) (66.186)
0.054 423.500 0.003 1.966
2011
(2.189) (15.266) (0.108) (70.863)
0.057 228.000 0.003 2.032
2012
(2.170) (7.898) (0.109) (70.405)
0.022 360.816 0.003 2.422
2013
(0.740) (10.950) (0.100) (73.505)
1090 TK Jayaraman, Keshmeer Makun

0.023 333.375 0.003 2.811

2014
(0.674) (9.016) (0.093) (76.027)
0.027 297.976 0.004 2.691
2015
(0.710) (7.251) (0.088) (65.484)
0.027 456.639 0.004 2.640
2016
(0.664) (10.345) (0.086) (59.809)
0.042 492.742 0.004 2.886
2017
(0.938) (10.127) (0.082) (59.315)
Source: World Bank (2019) and authors' calculations.
Notes: Figures in parenthesis are in percentage of GDP.

2.1. Support Factors

As Maldives is quite remote from its source markets, the role of ICT became critically important
as early as the first decade of the New Millennium. The country was one of the early countries,
which recognized the need for better ICT infrastructure, for connecting the atolls and promoting
tourism, and sought international assistance for strengthening the ICT sector together with the
development of the energy sector, which received considerable attention with large domestic
investments as well as FDI inflows.

The Third Tourism Master Plan (TTMP) covering a five-year period (2007–11) recommended
speeding up the adoption of electronic methods of doing business in tourism and related ancillary
sectors supporting the tourism industry. Loan assistance from the ADB enabled Maldives in
carrying out an ICT development project amounting to US$9.50 million 3, which was completed
in late 2009 for the establishment of a network connecting the government, and its agencies in
Malé and atolls, and for the installation of internet kiosks in Malé. The ICT Development Index
(International Telecommunication Union, 2017) based on ten indicators comparing it across the
countries over the last ten years places Maldives way ahead of the other SAARC countries (Table
3).

Acknowledging the role of ICT in tourism development as immense, as it helped to overcome the
hurdles posed by distance from source markets in regard to flight booking and purchase of air
tickets, and booking for accommodation, and tours, thereby reducing the traveling cost for air
travelers to a substantial extent, World Bank (2021a) describes ICT as the game-changer. It is
also visualized that ICT would play a significant role in other sectors in the future (World Bank,
2021a).

Table 3: ICT Indicators for SAARC Member Countries for 2017

SAARC member Mobile cellular subs per 100 Percent of individual using ICT
countries inhabitant internet Index
Afghanistan 66 10.6 1.95
Bangladesh 77.88 18.25 2.53
Bhutan 88.78 41.77 3.69
India 86.95 29.55 3.03
Maldives 222.99 59.09 5.25

3The project completed a network via fiber optic cable linking government agencies in Malé and outer islands. Following which,
various applications of electronic delivery of public services have been developed including a framework for E-government
services and storage hosted within the data center at the National Center for Information Technology.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1091

Nepal 111.7 19.69 2.88

Pakistan 71.39 15.51 2.42
Sri Lanka 118.49 32.05 3.91
Source: International Telecommunication Union (2017).

2.2. Covid-19 Impact on the Economy

While the global economy contracted by 3.5% in 2020, the Maldivian economy shrunk by 33%
in the first nine months of 2020. The reason behind this is the dominance of the tourism sector,
whose contribution to GDP and domestic employment is 58.6% and 59.6% respectively. Only
555,494 tourists visited Maldives in 2020 –the lowest number since 2002, at 67% lower than the
2019 record high of 1.7 million tourists (World Bank, 2021b)

As domestic cases of Covid-19 infection began to rise, Maldives closed its borders. From end-
March to mid-July 2020, the economy stood still (World Bank, 2020a). By late July 2020 things
appeared to have eased, the country allowed tourists who were screened and cleared at both
points of origin and entry. One atoll-one resort policy also helped the health authorities as the
resorts served as quarantine centers for those who were suspected of infection, as the resort
workers were vaccinated earlier in advance to receive the arrivals. By these measures, tourism
was revived. World Bank’s latest assessment is that Maldives contracted by 28% in 2020.
However, the decline is much more than 13.1%, which was the recorded rate of decline in 2005
following the Indian Ocean tsunami in 2004 (World Bank, 2020b)

As tourist arrivals rose to 100,000 in January and February 2021, though still 42% below pre-
pandemic levels, it was estimated by World Bank (2021a) that Maldives would grow by 17% in
2021 and 11.5% in 2022. However, the optimism was a short-lived one. New variants of the
pandemic emerged in Europe and India in late March 2021, destroying hopes of any speedy
recovery in source markets. Maldives had to impose travel restrictions once again, by putting an
end to staying in resorts, except those stand-alone resorts on atolls with no population except the
resort workers. As world shortages in vaccines developed, initial gifts of the Indian manufactured
vaccine were found inadequate. Maldives had to obtain vaccines from World Health
Organization under a special procedure. In the limited period of restrictions, Maldives stepped up
the vaccination program4. As of the end of April 2021, the country (with a population of 540,000)
has reached the vaccination rate of 84 per 100 people.

2.3. Macroeconomic Indicators

With the drop in tourism, not only did foreign exchange receipts fall from $752 million in
December 2019 to $ 570 million in 2020, but government revenues also plummeted. Although
non-priority expenditures were reduced, total expenditures in 2020 were only 4.5% lower than in
2019, because of the rise in public spending on health, social and economic relief measures. As a
result, the fiscal deficit reached 20% of the estimated GDP. The large public and external
financing gaps were mostly met by support from bilateral donors, with India playing a major role
and international financial institutions. Claims on the central government increased due to a

4
Total numbers infected were 47, 431 and recovered were 29,086 with deaths at 1050. Total number of doses were 451, 601
(Trading Economics, 2021).
1092 TK Jayaraman, Keshmeer Makun

larger quantum of issuance of T-bills and T-bonds to the State Bank of India as part of India’s
financial assistance of $250 million.

Maldives monetized the deficit, obtaining $ 262 million through short-term advances from the
Maldives Monetary Authority (MMA) in July 2020. Total public and publicly guaranteed debt
rose from $4.4 billion or 78.4% of GDP in 2019 to USD 5.6 billion or 139.3% of estimated GDP
in 2020. With the sharp decline in travel receipts, shortages of US dollars led the parallel market
rate to widen. The MMA activated a $ 400 million swap agreement with the Reserve Bank of
India to reduce pressures on the exchange rate (World Bank, 2021a). Table 4 with key
macroeconomic indicators including foreign exchange reserves illustrates the grave situation
faced by Maldives.

Table 4: Maldives Key Macroeconomic Indicators

Indicator 2019 2020 (estimate) 2021 (forecast)
Growth rate (%) 7 -28 17.1
Inflation (%) 0.2 -1.4 2.5
Fiscal Balance (% of GDP) -6.6 -20.1 -1.85
Current Account Balance (% of GDP) 26.8 -26.3 -27.1
Debt (% of GDP) 78.4 139.3 135.2
Foreign Res($ mill) 762.4 569.6 NA
Source: World Bank (2021a) and ADB (2021).

3. LITERATURE REVIEW5

The tourism activities and economic growth literature date back to 1997, largely to the
pioneering study by Sheldon (1997). Since then, substantial research literature on tourism
activities and growth nexus has emerged. The subject was examined mostly using two settings.
One is a country-specific setting, and the other is a panel and cross-country setting. Some of the
examples of country-specific studies on tourism and growth are Durbarry (2004) for Mauritius;
Nowak et al. (2007) for Spain; Kumar (2014) for Kenya; Ishikawa and Fukushige (2007) for
Japan; Katircioglu (2009) for the Turkish; Dritsakis (2004) for Greece. Examples of panel studies
include Wu et al. (2018); Holzner (2010); Kumar and Kumar (2013); Lee and Chang (2008);
Narayan et al. (2010); Roque and Raposo (2016) and Seetanah (2011).

Empirical investigations have confirmed the positive impact of tourism activities on economic
growth. However, the outcomes of the causality direction test varied among these studies. This
could be due to different proxies of variables, datasets, and analysis techniques. Further, other
factors such as differences in cultural traditions, political situation, and economic policies could
be the reason for such results (Ozturk, 2010). Table 5 presents a summary of well-known studies
on tourism and economic growth.

5 This section heavily draws upon Jayaraman and Makun (2020, 2022).
 Table 5: Tourism - Growth Literature Review: A Summary

Author Period Country Frequency Variables Methodology Causality Effect

Balaguer and Cantavella- Tourist earnings, exchange

1975-1998 Spain Quarterly VECM T→GDP +
Jorda (2002) rate

Tourism earnings, capital

Durbarry (2004) 1952-1999 Mauritius Annual VECM T↔GDP +
stock, human capital, labor

Cortez-Jimenez and Paulina Tourist earnings, capital

1954-2000 Italy Annual VECM T↔GDP +
(2006) stock, human capital

Quarterly and
Kim et al. (2006) 1971-2003 Taiwan Tourist earnings VECM T↔GDP +
Annual

Lee and Chang (2008) 1990-2002 OECD Annual Tourist earnings Panel T→GDP +

Narayan et al. (2010) 1988-2004 PICs Annual Tourist earnings Panel FMOLS T→GDP +

Panel of Islands
Seetanah (2011) 1990-2007 Annual Tourist earnings GMM T↔GDP +
(19)
Tourist earnings, political
Tang and Tan (2015) 1975-2011 Malaysia Annual VECM T→GDP +
stability

Tourist earnings, capital

Stauvermann et al. (2018) 1980-2014 Sri-Lanka Annual ARDL T→GDP +
stock, exchange rate, labor
Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT

Notes: GDP - Gross Domestic Product. ARDL - Autoregressive Distributed Lag approach. na refers to not applicable. T→GDP - causality relationship from
tourism to GDP. T↔GDP - the bidirectional relationship amid tourism and GDP. PICs - Pacific Island Countries. + is a positive effect of tourism on GDP.
1093
1094 TK Jayaraman, Keshmeer Makun

4. THEORETICAL FRAMEWORK

We adopt the neoclassical growth model of Solow (1956) augmented by tourism and ICT and
apply the dynamic ARDL bounds test technique designed by Pesaran et al. (2001). The bounds
test within the ARDL framework has been widely employed in recent years to investigate long-
run relations amongst time series factors with reliable econometric tests. This methodology is
also robust and efficient with minor sample observations (Ishida, 2015). Further, an extended
tourism-growth model that accounts for the asymmetric effect of tourism by a nonlinear ARDL
(NARDL) is also introduced.

The Solow (1956) framework, which is along the lines of the Cobb-Douglas function, takes into
consideration of Hicks-neutral technological development, the real GDP per-capita ( yt )
equation is expressed as:

yt  Ak 
t t , 0   1 (1)

Where yt is real GDP per capita; At is technology stock; kt is capita stock per capita; α is the
capital share. Accordingly, technological progress is represented by:

At  Aoegt (2)

where At is the cumulative technology in period t, Ao is the initial stock of technology, t is time
and g is the exogenous progress of technology. The pace of technical progress determines the
output growth, which is exogenous in the model. Thus, it is possible to include other growth-
enhancing variables. Besides capital stock, we introduce the variables representing tourism and
mobile cellular subscription - a proxy for ICT into the aggregate technology function.
Consequently, it is possible to adopt the following:

At  f (TEt , MOBt ) (3)

Here, TE is tourism earnings as a share of GDP and MOB is mobile subscription per 100
inhabitants. The aforesaid variables are shift factors, with capital stock as the fundamental factor
determining per-capita output. Hence, the initial Cobb-Douglas function is amended as:

yt  A0e1TEt 2MOBt kt3 (4)

For econometric estimation purposes, we employ the model in the natural log (l) form. The
logarithmic model has two advantages. First, the process of utilizing the variables in log form
reduces various errors including those arising from high standard error. Secondly, the magnitudes
of the coefficients indicate the elasticity of the variables. The stochastic model in its logarithmic
form for estimation purposes is expressed as:
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1095

lyt  0  1lkt  2lTEt  3lMOBt  t (5)

Based on this specification, we test the following hypotheses: (i) the fundamental variable,
namely, real capital per capita stock (lk) is positively associated with real output per capita (ly);
and (ii) the other explanatory variables, tourism earnings (lTE) and mobile subscription (lMOB)
are directly related with ly.

4.1. Data

The data series for the Maldives, covering a period of 1995 to 2019 employed in the study, are
drawn from various sources. Data series of real GDP per capita in constant US$ (2010), tourism
earnings expressed as a percent of GDP, and mobile subscriptions represented by a number per
100 inhabitants are sourced from World Development Indicators (WDI). The data series on
capital stock per capita in constant US$ (2010), which are available only up to 2019, is obtained
from Penn World Tables. We use quarterly observations (96) obtained by resorting to the cubic
spline procedure.6

The summary statistics and correlation analysis in levels are shown in Panel A and B of Table 6,
respectively. Panel A shows the maximum values, mean, and variance of the variables.
Regarding correlation, it is seen the capital stock per capita (k) is directly correlated at 0.97 and
remarkably has a greater correlation with GDP per capita (y). Tourism earnings (TE) and ICT
indicator (MOB) are also seen positively associated with per capita GDP (y) at 0.52 and 0.95
respectively.

Table 6: Summary Statistics of the Variables

Panel A: Descriptive measures
y k TE MOB
Mean 6633.38 22926.88 60.32 81.69
Median 6736.81 21036.26 61.23 82.8
Maximum 8971.13 41946.7 81.09 206.29
Minimum 4600.67 8727.15 37.57 0.01
Std. Dev. 1159.8 10695.57 12.74 70.7
Skewness 0.14 0.25 -0.42 0.14
Kurtosis 1.92 1.59 2.18 1.46
Jarque-Bera 4.61 8.31 5.11 9.13
Probability 0.1 0.09 0.08 0.11
Panel B: Correlation matrix
y k TE MOB
y 1
k 0.972 1
TE 0.525 0.59 1
MOB 0.959 0.994 0.64 1
Source: Authors’ calculations. Note: y is per capita RGDP. k is per capita real capital stock. TE is tourism earnings, as a
percent of GDP. MOB is a mobile cellular subscription per 100 persons, a proxy for ICT. TE*MOB is the interaction of
ICT and tourism.

6
The cubic spline interpolation procedure, which was noted to be a robust technique of transforming annual data to quarterly
series (Ajao et al., 2012), has been extensively used in the quantitative analysis.
1096 TK Jayaraman, Keshmeer Makun

4.2. Methodology and Estimation Procedure

We employ the bounds test procedure of Pesaran et al. (2001), within Autoregressive Distributed
Lag (ARDL) context, which has some technical advantages. It allows for conducting
cointegration tests even when variables are of mixed order of integration, that is, I(0) or I(1)
(Pesaran et al., 1999). It is also considered more suitable than other multivariate approaches if the
sample observation is small (Narayan, 2005). The ARDL estimates are also found to be
consistent (Pesaran & Shin, 1999). The following unrestricted model based on the ARDL
framework is used to examine the existence of cointegration:
n n
lyt   0  1lyt 1   2lkt 1  3lTEt 1   4lMOBt 1   1lyt i   2 lkt i 
i 1 i 0 (6)
n n

  lTE
i 0
3 t i   4 lMOBt i   t
i 0

Here,  is the difference factor and denotes the short-run impacts. Specifically, two steps are
required in this method. First, Equation (6) is regressed using the OLS method. Second, the
confirmation of cointegration is determined by imposing constraints on the coefficients of the
lagged level variables. The null hypothesis is that there is no cointegration
( H 0 : i1  i 2  i 3  i 4  i 5  0 ) and the alternative hypothesis is that there is
cointegration between the variables ( H1 : i1  i 2  i3  i 4  i5  0 ). To test the null
hypothesis, the estimated F-statistics from coefficient restrictions are compared with critical
values. According to Narayan (2005), the critical values of Pesaran et al. (2001) are for larger
sample studies and cannot be applied to small samples. Consequently, he tabulated critical values
based on a small sample study. Therefore, we employ Narayan’s (2005) critical values. The null
hypothesis is dismissed when F-statistic is greater than the critical value at a given level of
significance and if it is less than, the null hypothesis is not rejected.

Next, the long-run elasticity estimates of y with respect to each of the included variables are
estimated using the ARDL model. The presence of cointegration requires an error correction
model (ECM) to account for the short-run effect of ICT and tourism and to test the strength of the
long-run estimates in Equation (6). The ECM is stated in Equation (7) as:

n n n n
lyt  0   1lyt i   2lkt i   3lTEt i   4lMOBt i   ECM t 1   t (7)
i 1 i 0 i 0 i 0

In Equation (7),  is a coefficient of the ECMt 1 and is expected to have a negative sign,
reflecting the convergence process in the model.

4.2.1. Asymmetric model

The asymmetric estimation is the nonlinear extension of Pesaran et al.'s (2001) ARDL linear
model, which is a single long-run cointegration and short-run dynamic process. In the nonlinear
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1097

ARDL, we resort to Equation (6) following Shin et al. (2014). Under this scenario, positive and
negative shocks of tourism are examined and their impacts on GDP are not expected to be the
same. The asymmetric version of Equation (6) is presented as below:
n n
lyt   0  1lyt 1   2lkt 1  3lMOBt 1   4lTEt   4lTEt   1lyt i   2 lkt i  (8)
i 1 i 0
n n n

  lMOB
i 0
3 t i   4  lTEti   4  lTEti   t
i 0 i 0

Where lTEt and lTEt are the tourism’s partial sum decomposition in a positive and negative
n n
effect, respectively, and computed as lTEt   lTEt   max(lTEt , 0) and
t 1 t 1
n n
lTEt   lTEt   min(lTEt , 0) . Where lTEt  lTE0  lTEt  lTEt . The
t 1 t 1

 4  4
elasticity coefficient of lTEt and lTEt is computed as:     and 

 .
1 1
The error correction representation of Equation (8) yields the following:

n n n n
lyt  0   1lyt i   2 lkt i   3lMOBt i   4 lTEti
i 1 i 0 i 0 i 0
(9)
n n
 4 lTE   5lTE * MOBt i   ECM t 1   t
 
t i
i 0 i 0

The error correction term (  ECMt 1 ) estimates the equilibrium asymmetric relationship in the
specified model and the associated parameter (  ) captures the adjustment speed after shock.
The short-run positive and negative changes in tourism are captured by 4 and 4
respectively. To test for the asymmetry, the Wald test is applied. The null hypothesis
( Hnull :     ) for long-run symmetry is tested against the alternative hypothesis
( Halt :     ). Similarly, the short-run symmetry of remittances is tested by evaluating the
n n
null hypothesis ( 
i 0
4

  4  ).
i 0
1098 TK Jayaraman, Keshmeer Makun

5. RESULTS AND DISCUSSION

Before undertaking the econometric exercise, we conducted unit toot tests to determine the order
of integration. Results show all variables are stationary in differences, except MOB which was
already stationary in level. Thus the maximum order of integration is one, which confirms the
appropriateness of the NARDL cointegration procedure.

Table 7: Unit root test result

Variables ADF Tstat
In Level In the first Diff.
ly -0.642 (0.854) -5.766 (0.000)*
lk -1.261 (0.644) -5.312 (0.000)*
lTE -2.318 (0.168) -3.369 (0.014)**
lMOB -5.898 (0.000) -2.951 (0.043)**
Notes: Critical values for the ADF test are based on Mackinnon (1996). The length lag is based on Akaike Information
Criterion (AIC). The null hypothesis is that a series has a unit root (non-stationary). *, ** and *** indicate 1%, 5%, and
10% level of significance.

Results of the bounds test for the nonlinear model are provided in Table 8. The estimated F-
statistics is 6.24 for Equation (7). It is more than the critical value at one percent, confirming that
there is a long-run nonlinear association. Similar specifications are tested with different
dependent variables however, none of the equations emerged with significant F-statistics. Thus, it
is confirmed that there is only one nonlinear cointegration equation, with ly as the dependent
variable.

Table 8: Bound F-test Results for Cointegration

Dependent series Estimated F-statistics
ly 6.24
lk 1.04
lTE(+) 3.16
lTE(-) 2.76
lMOB 1.98
Critical Value a
Significance level Lower limit Upper limit
1% 4.483 6.320
5% 3.120 4.560
Notes: a Narayan (2005) critical bounds values from Case D: restricted intercept, and no trend.

In the nonlinear analysis, which seeks to investigate the existence of the asymmetric relationship,
the impact of tourism on economic growth is disaggregated by positive and negative partial sum
decomposition. The results reported in Table 9 confirm the existence of the asymmetric effect of
tourism earnings on economic growth.

Table 9: Asymmetric Test

Null hypothesis Long run Short-run
Symmetric effect of tourism = 4.26
earnings on real per capita (0.000)*
X 2 (1) = 5.83
GDP (0.001)*
Note: * is statistical significance at 1% level.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1099

Following the asymmetric test, we estimate the long-run and short-run nonlinear model and trace
out the impact of tourism on economic growth. Table 10 reports the findings of this analysis. In
the long run, the capital stock and output per capita are positively related. On average, the share
of capital stock is 0.33, which is in line with stylized values for developing countries (Rao et al.,
2010). This finding is supplemented by our initial result in Table 6 where the capital stock has a
significant positive correlation with output. Maldives as an emerging economy has also
experienced an increase in capital stock over time.

Tourism earnings (both negative and positive decomposition) emerged with a positive sign and is
statistically significant in the long run and short run. The estimates reveal that an increase in
tourism earnings (positive partial sum decomposition) by one percent, in the long run, causes
about 0.46% growth in per capita real GDP. A decrease in tourism earnings (negative partial sum
decomposition) by one percent leads to a decline in the per capita real GDP by about 0.54%. The
findings indicate that a decrease in tourism earnings by one percent has a larger negative impact
on economic growth when compared to a positive outcome of an increase by the same size in
tourism earnings. Similarly, in the short run, a one percent increase in tourism earnings increases
per capita GDP by 0.11%, and a one percent decline in tourism earnings will reduce income by a
much larger extent, about 0.33%. The negative impact of tourism is much more pronounced both
in the long run and in the short run.

The impact effect of ICT measured by mobile subscription (MOB) is positive and statistically
significant both in the short run and long run. The estimated coefficient of 0.01 in long run and
0.02 in the short run implies that a one percent increase in ICT leads to less than one percent
increase in economic growth, respectively. The significant positive effect confirms the vital role
of ICT in the tourism and economic growth process. This finding is consistent with the World
Bank’s (2021b) observation that ICT is a game-changer for tourism and travel. Finally, the
cointegrating equation (ECMt-1) which shows the speed of adjustment to stability after shock is
quite slow. It is estimated that there is about 43% convergence in the first period implying
recovery to long-run equilibrium would take about two years.
 1100

Table 10: Asymmetric Long Run and Short-Run Estimates

Long Run Estimates Short Run Estimates Model Diagnostics

Variables Coefficient Variables Coefficient Test Statistics

lk 0.335*** (0.059) ∆lk 0.069** (0.032) DW 2.08

lTE(+) 0.457** (0.011) ∆lTE(+) 0.111* (0.002) R-square 0.69

lTE(-) 0.536* (0.004) ∆lTE(-) 0.329* (0.000) R-bar-sq 0.654

lMOB 0.012* (0.000) ∆lMOB 0.002** (0.004) SC 0.420 (0.518)

Constant 4.096* (0.000) Constant 3.535* (0.000) NORM 0.800 (0.374)

TK Jayaraman, Keshmeer Makun

Trend 0.004* (0.000) Trend 0.002* (0.005) FF 0.479 (0.489)

ECMt-1 -0.431* (0.000) HET 0.811 (0.368)

Notes: Values in brackets are probability values. The lag length based on Schwartz Information Criterion is set to one. * and ** designates 1% and 5% levels of
statistical significance. DW is Durbin Watson statistics to account for autocorrelation. SC is a test for serial correlation, NORM is a normality test, FF is to test the
functional form and HET is a test for heteroscedasticity.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1101

The model diagnostic tests show that the results of the cointegration and error-correcting model
are reliable. We take into consideration the following: (i) Lagrange multiplier test of serial
correlation (X2sc); (ii) Ramsey’s RESET test for correct functional form (X2ff) using square of
the fitted values; (iii) Jarque-Bera’s normality (X2NORM) test; and (iv) test for heteroscedasticity
(X2HET) using the regression of squared residuals on squared fitted values. The findings of the
aforementioned tests are reported in Table 8. It indicates that the specified model does not suffer
from any classical econometric issue for the reasons that the null hypotheses that (i) there is no
serial correlation, (ii) sample is normally distributed, (iii) the functional form is correct and (iv)
the presence of homoscedasticity cannot be rejected. Further, the parameter stability is confirmed
by the cumulative sum of recursive squares (CUSUM) and cumulative sum of squares of
recursive residuals (CUSUMQ) in Figure 1 and 2. Therefore, results of the NARDL bounds test,
long-run elasticity estimates and ECM are reasonably reliable.

Figure 1: CUSUM
30

20

10

-10

-20

-30
98 00 02 04 06 08 10 12 14 16 18

CUSUM 5% Significance

Figure 2: CUSUMQ
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
06 07 08 09 10 11 12 13 14 15 16 17 18 19

CUSUM of Squares 5% Significance

1102 TK Jayaraman, Keshmeer Makun

6. CONCLUSION

In this paper, we investigated Maldives’ tourism and economic growth nexus in an augmented
Solow production functional model. Employing a non-linear ARDL model, we find the existence
of an asymmetric relationship between tourism earnings and economic growth. The results reveal
that while the positive partial sum decomposition of tourism earnings increases economic growth,
negative partial sum decomposition of the same size in tourism earnings reduces economic
growth to a larger extent, both in the long run and short. Further, it is found that ICT has a
significant, positive effect on economic growth.

From a policy perspective, the adverse asymmetric impact of tourism needs to be minimized with
the appropriate response. Given the severe extent of the decline in international tourism from the
overseas source markets, expansion of domestic and intra-regional tourism can be encouraged.
The domestic tourism of Maldives, as a percentage of spending on total travel, is presently small.
On the other hand, the number of intra-regional trips by destination to the Maldives as a percent
of total trips within the SAARC region is much higher, most of which are from India and Sri
Lanka. Promotion of intra-regional tourism within SAARC offers greater scope in the short-run,
as the resumption of international travel from outside the region to Maldives would take a long
time as travelers’ confidence level has to be restored. The intra-regional travel mobilization
efforts could include air travel bubble agreements between countries within the SARRC region,
and they can be bilaterally and multilaterally be negotiated.

In the recovery phase, Maldives may consider special efforts such as waiving visa requirements
and offering attractive discounts on travel fares and accommodation charges to attract Indian and
other regional Sri Lankan tourists. Middle-class families from India have been reported to be
spending more on international travel. Also, Indian movie producers, who prefer exotic, scenic
beauty and sandy beaches for shooting their movies, can now be targeted by Maldives.
Confidence of tourists can be built further by improved sanitary and hygienic conditions and
strict enforcement of Covid-19 protocols. The recent fast track package of US$7.3 million to
Maldives by the World Bank is specifically directed toward strengthening its public health
preparedness.

Just as there are downsides of Covid-19 impact in terms of loss of lives and livelihoods, there are
upsides, which would give confidence to policymakers towards strengthening efforts for
economic diversification and reducing dependency on tourism alone. As the World Bank (2021b)
has characterized ICT's role in tourism in Maldives as “A Digital Dawn”, the pandemic has
brought to light the hidden potential during the lockdowns in several spheres, including payment
systems, telehealth, online hygiene education, and many other digital applications. Given the
high degree of broadband and mobile internet penetration in the country compared to other
countries in South Asia, there is great scope for improvements in many areas of economic
development, especially in small and medium enterprises, manufacturing consumer goods and
their supplies for domestic consumption, delivering health and education services across the
small, dispersed islands for more balanced development. More importantly, other sectors such as
fisheries and marine products development which have export potential can now be taken up
more seriously in collaboration with foreign commercial interests, who are specialized in these
spheres.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1103

ACKNOWLEDGEMENT

The authors thank the anonymous reviewers and the editor of the International Journal of
Business and Society for helpful comments and suggestions.

REFERENCES

Ajao, I. O., Ibraheem, A. G., & Ayoola, F. J. (2012). Cubic spline interpolation: A robust method
of disaggregating annual data to quarterly series. Journal of Physical Sciences and
Environmental Safety, 2(1), 1-8.
Asian Development Bank. (2021). Key indicators for Asia and the Pacific [Data file].
https://www.adb.org/data/key-indicators/main
Balaguer, J., & Cantavella-Jorda, M. (2002). Tourism as a long-run economic growth factor: The
Spanish case. Applied Economics, 34(7), 877-884.
Behsudi, A, 2020. Wish you were here. Finance and Development, 57(4), 36-39.
Cortez-Jimenez, I., & Paulina, M. (2006). A further step into the ELGH and TLGH for Spain and
Italy (Fondazione Eni Enrico Mattei Working Paper No. 118.2006). https://www.fe
em.it/m/publications_pages/NDL2006-118.pdf
Dritsakis, N. (2004). Tourism as a long-run economic growth factor: An empirical investigation
for Greece using causality analysis. Tourism Economics, 10(3), 305-316.
Durbarry, R. (2004). Tourism and economic growth: The case of Mauritius. Tourism
Economics, 10(4), 389-401.
Holzner, M. (2010). Tourism and economic development: The beach disease? Tourism
Management, 32(4), 922-933.
International Monetary Fund. (2020). World economic outlook, October 2020: A long and
difficult ascent. https://www.imf.org/en/Publications/WEO/Issues/2020/09/30/world-eco
nomic-outlook-october-2020.
International Telecommunication Union. (2017). Measuring the information society: The ICT
development index [Data file]. https://www.itu.int/net4/ITU-D/idi/2017/index.html
Ishida, H. (2015). The effect of ICT development on economic growth and energy consumption
in Japan. Telematics and Informatics, 32(1), 79-88.
Ishikawa, N., & Fukushige, M. (2007). Who expects the municipalities to take the initiative in
tourism development? Residents’ attitudes of Amami Oshima Island in Japan. Tourism
Management, 28(2), 461-475.
Jayaraman, T. K., & Makun, K. (2020). Tourism-growth nexus in Pacific Island countries: A
panel study on ICT as a contingent factor. Tourism Economics, 26(3), 371-388.
Jayaraman, T. K., & Makun, K. (2022). Asymmetric analysis of tourism and economic growth in
South Asian countries: Lessons for policymakers towards mitigating the adverse effects
of Covid-19. Journal of Applied Economic Sciences, 2(72), 49 – 55.
Katircioglu, S. T. (2009). Revisiting the tourism-led-growth hypothesis for Turkey using the
bounds test and Johansen approach for cointegration. Tourism Management, 30(1), 17-
20.
Kim, H. J., & Chen, M. H. (2006). Tourism expansion and economic development: The case of
Taiwan. Tourism Management, 27(5), 925-933.
Kumar, R. R. (2014). Exploring the nexus between tourism, remittances, and growth in
Kenya. Quality and Quantity, 48(3), 1573-1588.
1104 TK Jayaraman, Keshmeer Makun

Kumar, R. R., & Kumar, R. (2013). Exploring the developments in urbanization, aid dependency,
sectoral shifts and services sector expansion in Fiji: A modern growth
perspective. Global Business and Economics Review, 15(4), 371-395.
Lee, C. C., & Chang, C. P. (2008). Tourism development and economic growth: A closer look at
panels. Tourism Management, 29(1), 180-192.
Narayan, P. K. (2005). The saving and investment nexus for China: Evidence from cointegration
tests. Applied Economics, 37(17), 1979-1990.
Narayan, P. K., Narayan, S., Prasad, A., & Prasad, B. C. (2010). Tourism and economic growth:
A panel data analysis for Pacific Island countries. Tourism Economics, 16(1), 169-183.
Nowak, J. J., Sahli, M., & Cortés-Jiménez, I. (2007). Tourism, capital good imports and
economic growth: Theory and evidence for Spain. Tourism Economics, 13(4), 515-536.
Ozturk, I. (2010). A literature survey on energy–growth nexus. Energy Policy, 38(1), 340-349.
Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approaches to the analysis of
level relationships. Journal of Applied Econometrics, 16(3), 289-326.
Pesaran, M. H., Shin, Y., & Smith, R. P. (1999). Pooled mean group estimation of dynamic
heterogeneous panels. Journal of the American statistical Association, 94(446), 621-634.
Rao, B. B., Tamazian, A., & Singh, R. (2010). What is the long run growth rate of the East
Asian Tigers? Applied Economics Letters, 17(12), 1205-1208.
Roque, V., & Raposo, R. (2016). Social media as a communication and marketing tool in tourism:
An analysis of online activities from international key player DMO. Anatolia, 27(1), 58-
70.
Salem, T., & Twining-Ward, L. (2018). Voice of Travelers-Leveraging User-Generated Content
for Tourism Development. https://openknowledge.worldbank.org/bitstream/handle/1098
6/30451/130052-WP-PUBLIC-Sept-25-2pm-DC-TheVoiceofTravelers.pdf?sequence=1
&isAllowed=y
Seetanah, B. (2011). Assessing the dynamic economic impact of tourism for island
economies. Annals of tourism research, 38(1), 291-308.
Sheldon, P. J. (1997). Tourism information technology. Cab International.
Shin, Y., Yu, B., & Greenwood-Nimmo, M. (2014). Modelling asymmetric cointegration and
dynamic multipliers in a nonlinear ARDL framework. In R. Sickles, & W. Horrace
(Eds.), Festschrift in honor of Peter Schmidt (pp. 281-314). Springer.
Solow, R. M. (1956). A contribution to the theory of economic growth. The Quarterly Journal of
Economics, 70(1), 65-94.
Stauvermann, P. J., Kumar, R. R., Shahzad, S. J. H., & Kumar, N. N. (2018). Effect of tourism on
economic growth of Sri Lanka: Accounting for capital per worker, exchange rate and
structural breaks. Economic Change and Restructuring, 51(1), 49-68.
Tang, C. F., & Tan, E. C. (2015). Tourism-led growth hypothesis in Malaysia: Evidence based
upon regime shift cointegration and time-varying Granger causality techniques. Asia
Pacific Journal of Tourism Research, 20(sup1), 1430-1450.
Trading Economics. (2021). Economics indicators [Data file]. https://tradingeconomics.com/mal
dives/indicators.
United Nations Conference on trade and Development. (2020). Handbook of statistics.
https://unctad.org/system/files/official-document/tdstat45_en.pdf.
United Nations World Tourism Organisation (2020). International tourist numbers could fall 60-
80% in 2020. https://www.unwto.org/news/covid-19-international-tourist-numbers-coul
d-fall-60-80-in-2020.
 Tourism and Economic Growth Nexus in Maldives: Asymmetric Analysis and Role of ICT 1105

United Nations World Tourism Organisation. (2019). Global and regional tourism performance.
https://www.unwto.org/tourism-data/global-and-regional-tourism-performance.
World Bank (2020c). Covid 19 and tourism in South Asia: Opportunities for sustainable regional
outcomes. World Bank.
World Bank (2021a). Maldives development update: A digital dawn. World Bank.
World Bank (2021b). South Asia focus, spring 2021: South Asia vaccinates. World Bank.
World Bank. (2019). World development indicator database [Data file]. www.worldbank.org/dat
a/countrydata/countrydata.html
World Bank. (2020a). World Bank fast-tracks $7.3million Covid-19 support to Maldives.
https://www.worldbank.org/en/news/press-release/2020/04/02/world-bank-fast-tracks-7
3-million-covid-19-support-to-maldives.
World Bank. (2020b, April 2). World Bank approves $12.8 million to support Maldives workers
impacted by Covid-19. https://www.worldbank.org/en/news/press-release/2020/06/09/w
orld-bank-approves-128-million-to-support-workers-in-maldives.
Wu, T. P., Wu, H. C., Liu, S. B., & Hsueh, S. J. (2018). The relationship between international
tourism activities and economic growth: Evidence from China’s economy. Tourism
Planning & Development, 15(4), 365-381.