Sage Journals: Discover world-class research

Abstract

Despite increased interest both from scholars and policymakers, most research on digital transformation in the tourism sector focuses on demand-side factors. However, supply-side factors, specifically how the adaptation of information and communication technology benefits the industry, are somewhat scant. Applying Institutions and Economic Theory, this study examines the role of digitalisation in tourism development by employing the advanced bias-corrected method of moments dynamic panel model along with the Arellano-Bond, Arellano-Bover/Blundell-Bond method of moments for around 123 countries, including economic categorisations from 1995 to 2020. The results reveal that tourism is found to be positively and significantly correlated with digitalisation. Notably, mobile phone subscriptions consistently show a significant effect on tourism demand for all countries, including developing countries. The findings imply that touristic destinations in developing economies started to follow the path of digital tourism by enhancing human connectivity between local communities and tourists through digital technologies.

Keywords

digital tourism digitalisation tourism demand international tourism panel data models information and communication technology

Introduction

Over the past few decades, hospitality and tourism technologies have driven digital transformation across all sectors, including the tourism industry (Bekele and Raj, 2025; Dogru et al., 2025; Hernandez and Lee, 2025; Kim et al., 2022; Lee and Lee, 2025). Information and communication technology has significantly contributed to the global GDP, ranking as a key contributor to global economic growth and development of countries (Lin et al., 2025; International Telecommunication Union (ITU), 2024). Furthermore, the ITU reported that in 2024, 68% of the global population (equivalent to 5.5 billion individuals) used the Internet, while broadband and mobile phone subscriptions reached 82%. The major advantage of digital technologies is that they serve as a vital tool in maintaining connectivity between individuals, governments, and businesses, particularly during times of crisis relating to natural disasters or pandemics. However, vulnerable populations in both developing and developed countries who lack access to or proficiency in digital technologies face the risk of being excluded from the post-pandemic recovery, potentially resulting in significant negative consequences. As such, the digital infrastructure, as measured by metrics such as Internet users, broadband subscriptions, and mobile phone technologies (Habibi et al., 2020), plays a crucial role in creating new service and process models, market channels, and organisational complexities including technological advancements (Myovella, 2020).

The travel and tourism industry is widely considered as an essential component of economic growth for any nation. For example, in the aftermath of the 2009 financial crisis, tourism sector emerged as a key driver of economic recovery for many countries (Sarıışık et al., 2011) that draws the attention of researchers and policymakers. While the major corpus of that research focuses on demand-side factors, while supply-side factors, specifically communication technology, are ignored. In addition, in the most recent domains regarding digital tourism, empirical and experimental pieces of evidence are missing, while conceptual studies dominate. As a result, it prevents decision-makers to formalise the role of digitalisation for tourism development at the strategy management level.

Furthermore, much of the existing research on technology in the tourism sector focuses on the effects of the individual decision-making (demand side) process (Lee et al., 2021), overlooking the institutional role (supply side) in which both people and businesses exist (Soares et al., 2021). In this study, the role of digitalisation is examined by applying the Institutions and Economic Theory by North (1992). North (1992) theorised that the existence of information asymmetric and imperfect information limits mental capacity of human being to process information, and therefore, calling for institutions to use technology to reduce uncertainties. While several studies highlighted institutional logics of service ecosystems (Polese et al., 2018) and the understanding of firm-level technology adoption (Oliveira and Martins, 2010), no study has examined the roles of digitalisation as a vehicle for the tourism industry from an institutional perspective. Moreover, as tourism sector is regarded as an information and labour incentive industry, the importance of perfect information for the tourism industry to address moral hazards, travel risks, and asymmetric information is beyond doubt.

Digital tools play a significant role in connecting both the demand (travellers and travel agencies) and supply sides (government and businesses), as shown in Figure 1. Developing digital infrastructure, by providing health and other services and accelerating internet and broadband facilities all depend on the institutions and economic capacity of a country, as initiated by North (1992). Hence, this study seeks to address three key research questions. First, how can we conceptualise the roles of digitalisation as a vehicle for tourism? Second, what are the key impacts of digitalisation on global tourism and developing evidence on several economic categorisations? Third, why do policymakers need to prioritise digital facilities for tourism development?

Figure 1.

Conceptualising roles of digitalisation as vehicle for tourism industry.

Using an advanced bias-corrected method of moments dynamic panel model along with standard dynamic panel data models for 123 countries from 1995 to 2020, astonishingly, our findings uncover the fact that digitalisation plays a significant role in determining the demand for inbound tourism in developing economies more than in developed nations. Accordingly, the study implies that the development of information technology is imperative for tourism growth in less-developed regions. In addition, integrating digital technologies in emerging economies can create an impression to potential tourists that these countries are moving towards a more modern thinking society, economic progression and technological advancement into the new century as presented in Figure 1.

The following section discusses the literature review, which then leads to the development of the hypothesis. Then, it presents the methodology used in the research, followed by empirical results and conclusion.

Literature review and hypothesis development

Conceptual framework

North (1992) developed Institutional and Economic Theory, postulating that institutions face constraints in resource allocation when efficient information feedback among human-to-human interactions is lacking. He further stressed the importance of efficient information feedback, combined with technology, to reduce the barriers of human interactions. However, the existence of information asymmetry is inevitable, and perfect information is virtually impossible to find without information technology. Löfgren et al. (2002) state that the consequences of information asymmetry and moral hazard adversely affect the relationships between demand and supply factors, like producers and consumers, which leads to uncertainty regarding quality and adverse selection in markets. Similarly, Puciato et al. (2013) argue that tourism is affected by imperfect information and moral hazard because of its complexity and information sensitivity. However, tourists often need various forms of information to make rational choices. Therefore, conforming to perfect information is a prerequisite for potential tourists and the tourist-oriented development of destinations, which eventually boosts inbound tourism for a country (Klimova et al., 2020).

Given the importance of efficient information flows, on the one hand, digital technologies can meet the demand for perfect information in an information-intensive tourism industry on a real-time basis. On the other hand, the easy access of technologies, perceived usefulness, and informativeness increase the demand for digital technologies among tourists (Ukpabi and Karjaluoto, 2017). Furthermore, tourism is recognised as an industry with a strong external spillover effect on other sectors, the regional economy, and social harmony, for which tourism security is essential. Moreover, digitalisation can be a potential driver to ensure accurate and up-to-date information for both demand-side and supply-side stakeholders (Xiang, 2018). Furthermore, Seetanah and Fauzel (2025) recently find positive interaction effects between digitalisation and economic growth on tourism development, suggesting that digitalisation provides social benefits to tourism and economic growth. Hence, the conceptual model of this study is summarised in Figure 2, where three sets of variables, i.e., digitalisation, inbound tourism, and the tourism determinant control variable, are interconnected and influencing each other to complete the whole process.

Figure 2.

Conceptual framework.

Digitalisation and international tourism

The travel industry has demonstrated consistent growth over the years, and it is projected to continue expanding, with an estimate that it will surpass 1.8 billion by 2030 (WTTC, 2017). While the travel sector grows with international arrivals, it is also witnessing an era of radical evolution in technology-dependent environments for several reasons. Firstly, the impact of technological advancements on the hospitality industry is evident in the way it influences issues related to accommodation, food and beverage establishments, and cultural heritage sites (Figueredo et al., 2017). This is primarily with the use of reviews, ratings, and tourist-generated suggestions, which can have a significant progressive impact on the decision-making process of international travellers. Figueredo et al. (2017) find that 85% of international travellers use portable devices such as iPods, tablets, or smartphones during their trips, while 97% of mobile phone subscribers share pictures of their tourism activities on social media platforms like Instagram, Facebook, and Twitter. These digital technology trends have enabled travellers to personalise their destination-related activities, resulting in a more fulfilling and satisfying travel experience that extends beyond their trip (Harris, 2017; Lee et al., 2018).

Secondly, the emergence of the digital revolution has significantly altered the transformation of tourism industry (Bekele and Raj, 2025). Tourists can access information about their intended destinations, including details about climate and weather patterns, available accommodations, scenic attractions, geopolitical and economic conditions, as well as the ability to make travel bookings, engage in online shopping and payments, and capture memories (Adeola and Evans, 2019; Law et al., 2018). Recent studies have shown that technologies like smartphones and internet connectivity can have growth enhancing impacts on the tourism industry as international tourists can now easily access travel information from abroad, thus, increasing their confidence in worldwide destinations, including language and cultural barriers, which were previously impossible (Tay, 2020).

Thirdly, digital tourism is characterised by consumer demand, technological innovations, and industry functions. From the consumer demand perspective, it enables tourists to access up-to-date information on reservations at a fraction of the cost and provides relevant information regarding destinations, resorts, and hotels, as well as a suite of activities (Athari and Çağlar Onbaşıoğlu, 2019; Buhalis and Law, 2008; Li and Buhalis, 2005; Mills and Law, 2004; O’Connor et al., 2001). In this regard, digital tools can promote the word-of-mouth effect through reviews and post-travel reports on websites and travel forums (Gelb and Sundaram, 2002). Subsequently, it can reduce time lags due to lengthy information searches and the uncertainty through quick access to the internet that may arise from expensive travel and bad experiences (Spencer, 2019). Furthermore, numerous service providers utilise a range of interactive tools, such as push SMS, promotional emails, and live chat, to engage with potential tourists regarding pricing and special promotions. Consequently, this leads to cost savings through reduced prices, commissions, or airfares, which in turn allows tourists to allocate more funds towards their intended destinations (Clemons et al., 2002; Luo et al., 2004).

Finally, global tourism is experiencing high inter-industry competition, so destinations are striving to increase their competitiveness to attract more tourists (Dogru and Suess, 2021). Literature reveals a series of determinants where technology, infrastructure, digital capability, and technological readiness, including wired and wireless internet connectivity, are very common and conform to the access of information. Thus, the following hypothesis can be constructed:

Digitalisation is positively associated with tourism development.

Data, model, and methodology

Data

The dependent variable of this study is tourism demand, which typically involves the assessment of visitor arrivals, the corresponding international tourist receipts within the host countries and is frequently employed within academic discourse (Saha and Yap, 2014). International tourist arrivals (TA), which refers to international inbound tourists (overnight visitors), are the number of tourists who travel to a country other than that in which they have their usual residence but outside their usual environment for a period not exceeding 12 months. The data on inbound tourists refers to the number of arrivals, not the number of people travelling. Thus, a person who makes several trips to a country during a given period is counted as a new arrival each time. On the other hand, international tourism receipts (TR) are used for the robustness check and are calculated by counting expenditures by international inbound visitors, including payments to national carriers for international transport. The data are in current U.S. dollar. These receipts include any other prepayment made for goods or services received in the destination country and receipts from same-day visitors, except when these are important enough to justify separate classification. International tourist arrivals data strictly identify tourists as foreign tourists in a country, whereas tourism receipts include expenditure by both domestic and international tourists.

Digitalisation of a country is the prime focus and independent variable of this study. The present research aims to determine variables based on the significance and accessibility of data. Hence, the process of digitalisation has been operationalised through the utilisation of mostly used indicators such as mobile phone subscriptions (MS) and fixed broadband subscriptions (FBS). MS represents the total mobile cellular telephone subscriptions that offer voice communications. The succeeding variable pertains to the metric of FBS, which refers to the invariable subscriptions for the provision of high-speed access to the publicly accessible internet through a Transmission Control Protocol/Internet Protocol (TCP/IP) connection, with downstream speeds matching or surpassing 256 kilobits per second (kbit/s).

In addition, the current study incorporates various control variables, particularly related to country risks, to effectively model tourism demand. According to Athari et al. (2023), country risks arising from domestic economic and political uncertainties can adversely affect tourism development, which is detrimental to international tourism. Moreover, Carmignani (2005) finds that strong institutions are essential for maintaining political stability by reducing the Gini coefficient and increasing average income, growth, and income of the poor people. In this study, we utilise real GDP to comprehensively assess the level of economic advancement achieved by the target nations because an elevated level of economic development can yield superior tourism products and infrastructure, including digital infrastructure. Furthermore, the consumer price index (CPI) is incorporated in this research as a proxy of living costs at the intended tourist destination (Song and Li, 2008). The higher CPI value can indicate higher affordability for a particular economy (refer to Wang, 2009). Regarding country-specific characteristics, political stability can influence tourists’ risk perceptions of travelling to a destination. Nations with high political stability levels are considered safe destinations for tourists, and they witness more tourism activities and international visitors than those politically unstable countries (Athari et al., 2021).

Table 1 presents the list of all variables included in the current research. The data used in this study are obtained from the World Bank’s World Development Indicator (WDI), which has been accessible since 1995 and contains the latest records up to 2020. The selection of this time frame served as the basis for the analytical period. The reliability of the WDI as a metric for assessing digitalisation has been established in numerous studies, such as Habibi and Zabardast, 2020; Kozłowski et al., 2021; Kumar et al., 2019. The WDI dataset has been constructed using cross-country surveys and expert polls.

Table 1.

List of variables used.

Variable type	Variable name	Label	Measurement and description
Dependent variable	International tourist receipt	TR	Amount of USD received from international travellers
Dependent variable	International tourist arrivals	TA	Number of international travellers arrived
Independent variable	Mobile cellular subscriptions	MS	Total mobile phone subscriptions
Independent variable	Fixed broadband subscriptions	FBS	Total fixed broadband subscriptions
Control variable	Real gross domestic product	RGDP	GDP is converted to international dollars using purchasing power parity rates
	Consumer price index	CPI	Price index converted to US dollars
	Political stability index	PS	Governance indexing ranging from 1 (lowest political stability) to 10 (highest political stability)

Source: World Bank’s World Development Indicator.

Furthermore, our investigation is based on the data availability from the WDI. Due to extensive missing data, 123 countries are selected with the most data values. These countries are categorised into two subgroups, namely developed and developing nations. Supplemental Table 1 provides a list of the chosen countries for this research.

Figure 3 presents the average of annual percentage change in international tourism and digitalisation data between 1995 and 2020 by countries. However, in the year 2020, the outbreak of COVID-19 pandemic caused the closure of international travel borders globally, affecting both international tourist arrivals and tourism receipts. In fact, the data shows that the pandemic event alone affected developing countries worse than developed countries in terms of tourism revenue. In other words, the annual growth rates of global tourism flow were relatively stable before the pandemic. In addition, Figure 3 reveals that the mobile cellular and fixed broadband subscriptions were at their peak in the year 2000 due to the rise of the technology revolution, where technology devices became more personal and portable and the massive growth of social media (World Economic Forum, 2020). However, the average annual increase in digital subscriptions presents declining trends for developed and developing nations over the years, indicating the momentum of technology adoption has reached its stability and steady state level (DATAREPORTAL, n d).

Figure 3.

The average of annual percentage change in international tourism and digitalisation data between 1995 and 2020.

Panel unit root test

This research delves into an examination of the effects of digitalisation on tourism demand. To achieve consistent variance and facilitate unambiguous interpretation of coefficients, the present study employs the natural logarithm transformation for the considered variables, which is commonly used in panel data analysis. Next, we adopt Karavias and Tzavalis (2014) panel unit root test to determine the stationarity of the data in the presence of structural breaks, trends and cross-sectional dependency. According to Chen et al. (2022), structural breaks are unanticipated shocks that have permanent effects on model parameters, and hence, disregarding structural breaks in unit root tests can mislead the conclusion of accepting an incorrect hypothesis.

In this study, we run the panel unit root test based on two conditions. The first condition includes at least one unknown structural break and cross-sectional dependency, while the second condition extends the first condition by adding a time trend. Table 2 presents the unit root test for all variables. The results show that, without a time trend, all variables reject the null hypothesis of unit root processes. Hence, the data are stationary when structural break and cross-sectional effects are considered. However, when including a time trend in the unit root test, the level data for TA, TR, FBS, RGDP, CPI and PS cannot reject the null hypothesis of unit root processes. In other words, these variables are non-stationary and integrated at the order one, I (1), which are caused by time effects. To overcome this issue, we transform the data by first differencing the data to omit the time effects. Mathematically, it can be written as:

{\tilde{X}}_{i t} = \ln (X_{i t}) - \ln (X_{i t - 1})

, where X is a numerical variable, i is cross-country and t is time script. We re-run the panel unit root test for the first-differenced data and all variables become stationary or I (0). Therefore, the research uses first-differenced data to generate reliable and consistent estimations. Supplemental Table 2 summarises statistical descriptions of the data used in this research.

Table 2.

Panel unit root tests.

Variable	Without time trend	With time trend
Level data
L (TA)	−21.761***	3.669
L (TR)	−7.406***	−2.116
L (MS)	−57.523***	−20.792***
L (FBS)	−42.725***	−5.029
L (RGDP)	−37.307***	7.298
L (CPI)	−36.835***	7.599
L (PS)	−3.224***	−1.773
First differenced data
DL (TA)	−70.078***	−39.493***
DL (TR)	−81.329***	−45.676***
DL (MS)	−60.202***	−31.112***
DL (FBS)	−42.287***	−23.703***
DL (RGDP)	−47.947***	−25.167***
DL (CPI)	−35.906***	−23.309***
DL (PS)	−53.392***	−27.221***

Note: The test uses Karavias and Tzavalis (2014) panel unit root test statistic with at least one structural break and cross-sectional dependency. The null hypothesis is that all panel time series are unit root processes. ***Denotes significance at the 1% level. L(.) indicates the natural logarithm operator. DL(.) denotes log difference operator. TA: international tourist arrivals; TR: tourism receipts; MS: mobile cellular subscriptions; FBS: fixed broadband subscriptions; RGDP: real gross domestic product; CPI: consumer price index; PS: political stability index.

The model and estimation procedures

In this study, for simplicity, we propose a linear panel data model where the log-difference of tourism demand is a function of its lag-dependent variable and log-difference digitalisation and economic indicators. Furthermore, as the COVID-19 event impacted global travel markets significantly, including the one-off event in our model, is crucial. Therefore, mathematically, our estimation model is written as follows:

Y_{i t} = β_{0} + β_{1} Y_{i t - 1} + β_{2} D_{2020, i} + β_{3} {Y e a r}_{i} + X_{i t}^{'} β + W_{i t}^{'} γ + e_{i t}

(1)

X_{i t} = [\begin{array}{l} {D L (M S)}_{i t} \\ {D L (F B S)}_{i t} \end{array}]

W_{i t} = [\begin{array}{l} \begin{array}{l} {D L (R G D P)}_{i t} \\ {D L (C P I)}_{i t} \end{array} \\ {D L (P S)}_{i t} \end{array}]

e_{i t} = μ_{i} + u_{i t}

where

Y_{i t}

= log-difference of tourism demand for country i at time t. Two proxy variables for

Y_{i t}

are used in this research, namely the log differences of international tourist arrivals (DL (TA)) and tourism receipts (DL (TR)).

D_{2020, i}

= dummy variable to capture the COVID-19 pandemic that occurred in 2020 for each country i,

{Y e a r}_{i}

= year effects for each country i,

X_{i t}

= a set of log difference of digitalisation indicators for country i at time t,

W_{i t}

= a set of log difference of control variables for country i at time t,

β a n d γ

= coefficient matrix, DL(.) = log difference operator, MS = mobile cellular subscriptions; FBS = fixed broadband subscriptions; RGDP = real gross domestic product; CPI = consumer price index; PS = political stability index;

μ_{i}

= individual country i, and

u_{i t}

= i. i.d. error term where

u_{i t} \sim N (0, σ^{2})

As part of the robustness check, we constructed three empirical models to determine whether the focus variables, namely MS and FBS, have consistent outcomes regarding significance and coefficient signs across different econometric models.

Model 1

Y_{i t} = f (Y_{i t - 1}, {D L (M S)}_{i t}, {D L (F B S)}_{i t}, {D L (R G D P)}_{i t}, {D L (C P I)}_{i t}, D_{2020, i}, {Y e a r}_{i})

Model 2

Y_{i t} = f (Y_{i t - 1}, {D L (M S)}_{i t}, {D L (F B S)}_{i t}, {D L (R G D P)}_{i t}, {D L (C P I)}_{i t}, D_{2020, i}, {Y e a r}_{i}, {D L (P S)}_{i t})

Model 3

Y_{i t} = f (Y_{i t - 1}, {D L (M S)}_{i t}, {D L (F B S)}_{i t}, {D L (R G D P)}_{i t}, {D L (C P I)}_{i t}, D_{2020, i}, {Y e a r}_{i}, {D L (R G D P)}_{i t} * D i g i t a l i s a t i o n i n d i c a t o r)

Model 1 focuses on the effects of digitalisation on tourism demand using economic variables as the control variables. Model 2 extends Model 1 by including political stability as the control variable for country risk, whereas Model 3 expands Model 1 by analysing the interaction between real GDP and digitalisation, as a higher level of income of a country makes the country able to invest more in digital infrastructure, which in turn enhances inbound tourism industry.

Preliminary tests on normality, slope heterogeneity and cross-sectional dependency

We conducted several preliminary tests using the standard linear panel data models to examine the existence of non-normality, slope heterogeneity and cross-sectional dependency. For our testing, the standard linear panel data models are based on panel generalised least squared estimations¹. Table 3 shows that the standard linear models reject the null hypothesis of normality and slope homogeneity when both tourist arrival and tourism receipt data are used as the dependent variables. This shows that using standard linear panel models when the residuals are non-normal and slope coefficients are heterogeneous can yield invalid inferences and biases in estimations (Alejo et al., 2015; Bersvendsen and Ditzen, 2021). However, Table 3 reveals that the models cannot reject the null hypothesis of cross-sectional independence, indicating that cross-sectional units are independent and may not cause the heterogeneity in slope coefficients. In conclusion, the models may likely suffer from misspecification under the assumption of standard linear panel data models, and therefore, dynamic panel data estimations are proposed in this study.

Table 3.

Preliminary tests on normality, slope heterogeneity and cross-sectional dependency.

	DL (TA)	DL (TR)
Normality test on $e_{i t}$ :	18.91	9.65
Chi-squared statistics [p-value]	[0.000]	[0.008]
Normality test on $μ_{i}$ :	10.48	6.84
Chi-squared statistics [p-value]	[0.005]	[0.033]
Testing for slope heterogeneity:	2.113	13.160
Delta test statistics [p-value]	[0.035]	[0.000]
Testing for cross-sectional dependence:	1.69	1.05
CD test statistics [p-value]	[0.091]	[0.295]

Note. The normality test examines the normality in standard panel-data one-way error-component models, based on the skewness and kurtosis in the individual-specific and residual errors (Galvao et al., 2013). Testing for slope heterogeneity in panels is based on the standard delta test that weighs the difference between the cross-sectional unit-specific and pooled estimates (Bersvendsen and Ditzen, 2021). Testing for cross-sectional dependence is based on the test statistics developed by Pesaran (2015) to examine the existence of weak cross-sectional dependence across units.

Dynamic panel data models

Endogeneity is another issue for estimating equation (1), where the explanatory variables on the right-hand side of the equation must be strictly exogenous regressors, and the correlation between the variables and $e_{i t}$ is zero for all i and t (Baltagi, 2008: p. 121). Since $Y_{i t}$ is the function of $e_{i t}$ , equation (1) has violated the strictly exogenous condition because $Y_{i t - 1}$ is correlated with the error term. In other words, the issue of endogeneity cannot be ignored. To overcome the endogeneity problem, Arellano and Bond (1991) and Blundell and Bond (1998) proposed a generalised method of moments (GMM) estimator in dynamic panel data models that incorporates a combination of lagged level and first difference data as instrument variables. Blundell and Bond (1998) claim that the GMM estimator is robust because the instruments are good predictors for endogenous variables. However, Breitung et al. (2022) and Bun and Windmeijer (2010) argue that the method suffers from weak instrument problems when the variance of individual-specific effects ( $σ_{μ}^{2}$ ) is greater than the variance of the idiosyncratic errors ( $σ_{u}^{2})$ . Therefore, to improve the use of GMM in dynamic panel data analysis, Breitung et al. (2022), hereafter as BKH, introduce a biased-corrected method of moments estimation. The model first computes the within-group least squares and subsequently adjusts the sample moment conditions to estimate the coefficients. To explain mathematically, setting $θ = {(β_{0}, β_{1}, β_{2}, β^{'}, W^{'})}^{'}$ , we can re-write equation (1) as:

e_{i t} (θ) = Y_{i t} - (β_{0} + β_{1} Y_{i t - 1} + β_{2} D_{2020, i} + β_{3} {Y e a r}_{i} + X_{i t}^{'} β + W_{i t}^{'} γ), and

{\bar{e}}_{i t} (θ) = T^{- 1} \sum_{t = 1}^{T} e_{i t} (θ)

The estimated variance of the within-group least square can be expressed as:

{\hat{σ}}_{T i}^{2} (θ) = {(T - 1)}^{- 1} \sum_{t = 1}^{T} [e_{i t} (θ) - {\bar{e}}_{i t} (θ)] e_{i t} (θ), in such that E [{\hat{σ}}_{T i}^{2} (θ)] = σ_{i}^{2}

The moment conditions suggested by BKH can be specified as follows:

m_{T i} (θ) = T^{- 1} \sum_{t = 1}^{T} [(\begin{array}{c} Y_{i t - 1} - {\bar{Y}}_{- 1, i} \\ x_{i t} - \bar{x_{i}} \end{array}) - (\begin{array}{c} b_{T} {\hat{σ}}_{T i}^{2} (θ) \\ 0 \end{array})] e_{i t} (θ) = 0

(2)

where

b_{T} = - T^{- 2} \sum_{t = 0}^{T - 2} \sum_{s}^{t} β_{1}^{s}

, in such that

E [m_{T i} (θ)] = 0

From equation (2), the estimators can then be solved as follows:

\hat{θ} = \arg \begin{array}{c} \min \\ θ \end{array} {(\sum_{i = 1}^{N} m_{T i} (θ))}^{'} (\sum_{i = 1}^{N} m_{T i} (θ))

(3)

Equation (3) solves the fixed effects biased-corrected estimators. To obtain the random effects 2-step biased-corrected estimators, equation (4) below is the extension from equation (3) by including a weighting matrix, as follows:

{\hat{θ}}^{*} = \arg \begin{array}{c} \min \\ θ \end{array} {(\sum_{i = 1}^{N} m_{T i} (θ))}^{'} W (\sum_{i = 1}^{N} m_{T i} (θ))

(4)

where

W = {(N^{- 1} \sum_{i = 1}^{N} m_{T i} (\hat{θ}) {m_{T i} (\hat{θ})}^{'})}^{- 1}

According to Kripfganz and Breitung (2022), the biased-corrected dynamic panel data model presents several advantages. The model can compute fixed- and random-effects estimators, and the method of moments specified in the model follows an asymptotic distribution. Moreover, the model works well with small sample properties. Hence, a biased-corrected dynamic panel GMM model is employed as the main econometric technique to estimate the relationship between digitalisation and tourism. We further tested our models using Arellano-Bond and Bond-Blundell methods to check for robustness.

Estimation results

The empirical analysis of the relationship between digitalisation and tourism demand is conducted using cross-country panel data. Based on the unit root test results, the annual percentage changes (log difference) of the dependent and independent variables are estimated. We first start with the scatter plots of percentage changes in digitalisation and its impact on tourism demand in a country. The horizontal and vertical axes in Figure 4 measure the percentage changes in digitalisation and tourism variables, respectively. In Figure 4(a), scatter plots for all countries (upper panel), shows that the growth in MS leads to increase in tourist arrivals as most of the dots are in the positive quadrant, with some exceptions. Similar movements are seen when FBS is measured on the right panel. However, the movements are a bit limited compared to the MS. The lower panel in Figure 4(a) displays a similar relationship between digitalisation and tourism revenue. Figures 4(b) and 4(c) show the same relationship for developed and developing countries, respectively. The findings indicate that nations characterised by high amounts of digitalisation exhibit greater appeal to travellers, exemplified by the notable influx of international tourists into countries such as the United States and Australia. Conversely, realms of limited digitalisation are inclined to discourage tourism, as evidenced by the scant number of visitors to destinations such as Burundi and Sudan.

Figure 4.

(a)Scatter plots of log difference tourism and digitalisation data for all countries. (b) Scatter plots of log difference tourism and digitalisation data for developed countries. (c) Scatter plots of log difference tourism and digitalisation data for developing countries.

Next, the relationship between digitalisation and inbound tourism is more rigorously estimated by incorporating control variables. The results for all countries are reported in Table 4. The findings reveal that the DL (MS) coefficient is positive and significant in all models (Models 1–3). In other words, a 1% increase in DL (MS) leads to 0.09% growth in tourist arrivals, ceteris paribus, supporting our hypothesis that mobile subscription can enhance international tourist flows in a country. The result is consistent with Irvine and Anderson (2008), who find that information and communication technology is seen as enhancing personal service by promoting quality of service. However, fixed broadband services have shown growth enhancing effects in attracting international tourists, but they are not significant. In addition to the varying coefficients among the indicators of digitalisation, it is evident that mobile phone subscriptions exhibit the most prominent coefficient with higher magnitudes. Essentially, the provision of readily accessible mobile phone subscription services plays a pivotal role in attracting tourists, as it enables them to maintain seamless connectivity while offline or online. In this context, the eminence of technology readiness is more crucial than the inherent allure of the destination itself. For example, the United States is one of the superior nations in terms of internet accessibility and as an exceptional tourist destination. Conversely, Sao Tome represents one of the least developed nations with limited access to the internet and is the least visited tourist destination.

Table 4.

Bias-corrected estimation of linear dynamic panel data models for all countries.

	Dependent variable: DLTA			Dependent variable: DLTR
	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
One lag dependent variable	0.085** (0.036)	0.067* (0.039)	0.066 (0.040)	0.108 (0.076)	−0.019 (0.116)	−0.019 (0.115)
DLMS	0.085*** (0.029)	0.109*** (0.038)	0.083* (0.042)	0.082** (0.036)	0.087* (0.052)	0.034 (0.054)
DLFBS	0.006 (0.009)	0.012 (0.012)	0.011 (0.012)	−0.005 (0.010)	0.024* (0.052)	0.022 (0.014)
DLRGDP	0.884*** (0.127)	0.824*** (0.120)	0.857*** (0.121)	0.949*** (0.152)	1.231*** (0.242)	1.295*** (0.212)
DLCPI	0.034 (0.041)	0.049 (0.041)	0.056 (0.040)	−0.013 (0.064)	0.107 (0.072)	0.122* (0.072)
YR2020	−1.326*** (0.049)	−1.335*** (0.049)	−1.336*** (0.068)	−1.070*** (0.048)	−1.059*** (0.052)	−1.062*** (0.052)
DLPS		0.183*** (0.067)	0.179*** (0.068)		0.354*** (0.096)	0.347*** (0.097)
DLRGDP × DLMS			0.564 (0.417)			1.149*** (0.394)
Year effect	0.002* (0.001)	0.002* (0.001)	0.002* (0.001)	−0.003*** (0.001)	−0.003*** (0.001)	−0.003*** (0.001)
Arellano–bond test for zero autocorrelation in first-differenced errors of order 2: Z statistics [p-value]	2.530 [0.011]	2.287 [0.022]	2.388 [0.017]	1.041 [0.298]	1.014 [0.311]	1.038 [0.299]
Hansen test of the overidentifying restrictions:	12.054 [0.061]	14.280 [0.046]	19.055 [0.015]	10.740 [0.097]	16.242 [0.0230]	20.224 [0.010]
Chi-squared statistics [p-value]	12.054 [0.061]	14.280 [0.046]	19.055 [0.015]	10.740 [0.097]	16.242 [0.0230]	20.224 [0.010]
Hausman test:	16.377 [0.012]	21.182 [0.004]	26.340 [0.001]	10.631 [0.101]	19.658 [0.006]	23.653 [0.003]
Chi-squared statistics [p-value]	16.377 [0.012]	21.182 [0.004]	26.340 [0.001]	10.631 [0.101]	19.658 [0.006]	23.653 [0.003]

Note: All coefficients are the fixed effects bias-corrected method of moments estimators proposed by Breitung et al. (2022), except DLTR in Model 1 which is based on the random effects bias-corrected method of moment estimators. Models 2 and 3 are robustness tests that include political stability and the interaction effect between DLRGDP and DLMS, respectively, to determine consistency using the bias-corrected method of moments estimations in Model 1. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.

Although, both fixed broadband and mobile subscriptions are reliable digital tools for both residents and businesses, however, mobile broadband provides a lot more flexibility as fixed broadband is tailored for stationary locations. In contrast, mobile broadband is designed for users who need internet connections while on-the-go. Quaglione et al. (2020) argue that mobile broadband offers have become the main wireless market option for accessing internet in the rural areas in EU member states. With the deployment of third-fourth-and-fifth-generation mobile standard enhances data transfer speed and relaxes the existing service constraints and thereby reduces the gap with the better performing wired options and increases its appeal to those territories where wired networks are considerably unviable. Moreover, Prieger (2013) finds that the availability and adoption of high-speed fixed broadband are lower in rural than urban areas in the US. However, mobile broadband holds a great potential to benefit rural areas through economic development. Hence, it is reasonable to argue that fixed broadband has some limitations for the tourists to access information in the tourist places that are in rural and remote areas. Also, over the years mobile technology has grown at a much faster rate that postulates more flexibility and access to both tourists and the provider of tourism services which in turn helps enhancing tourism industry. Overall, dependable and simple internet or hotline accessibility plays an important role in improving the satisfaction of travellers during their vacation. For instance, the travel applications, encompassing Tripadvisor, Airbnb, and Booking.com, serve as travel planning and support tools to assist tourists in efficiently executing all scheduled activities before and during the holiday period, utilising online connectivity for operations and maintenance. As a result, information and communication connectivity are extremely important for the advancement of the leisure industry.

The coefficient for lagged growth in tourist arrivals is positive and significant at a 5%–10% level of significance, except in Model 3 of Table 4, indicating an opposite catch-up effect where growth in tourist arrivals in a previous year boosts growth of tourist numbers for the subsequent year. It is reasonable to argue that digitalisation provides better information to tourists through feedback, which supports North’s Institutional and Economic Theory. If tourists are satisfied, then the consequence of tourists’ satisfaction, word of mouth (Lang and Hyde, 2013), attracts more tourists in the coming years.

The coefficients for the real GDP growth reveal the expected positive sign with 1% significant level in all models, suggesting that high economic development in a tourist destination increases inbound tourism by providing better infrastructure. The coefficients of CPI do not show the expected sign and are insignificant. Interestingly, the coefficients of 2020 year dummy is negative and highly significant (1% level of significance) with a greater magnitude in all models. The results confirm the worldwide adverse impact of the COVID pandemic due to lockdowns and border closures in many countries.

The coefficients for political stability reveal a boosting effect. They are highly significant (Models 2 and 3), indicating that political stability is a crucial factor for inbound tourism, as tourists feel safe and less risky while enjoying travel. A 1% increase in DL (PS) leads to 0.18% growth in tourist arrivals, ceteris paribus, manifesting that political stability generates an international tourism flow in a country. The result is consistent with Athari et al. (2023), suggesting that political stability encourages tourism growth. Finally, the moderation effect between real GDP and digitalisation shows a positive sign in boosting international tourism.

The results are very similar when tourism revenue is replaced with tourist arrivals (Table 4), and the magnitude of the coefficients, particularly real GDP, political stability and the moderation coefficient, are much stronger when TR is the dependent variable. The moderation coefficient is positive and highly significant, suggesting that a higher level of GDP supports digital infrastructures, which generate positive outcomes for the tourism industry (Model 3), Table 4. Also, the models (both in TA and TR) are tested for serial correlation and overidentifying restrictions for endogeneity, and the results satisfy no serial correlation and control for endogeneity. The next subsection examines the impact of digitalisation on inbound tourism for both developed and developing countries to identify differences in the effects of digitalisation.

Developed and developing countries

Our evidence shows that the impact of digitalisation variables may exhibit variation. The digital incorporation into various information systems in developed nations is facilitated by a sound technological design, substantial investment in IT equipment, and a robust infrastructure. These factors collectively confer a favourable position for these nations in their capacity to embrace digital transformation. The estimation results for MS reveal a positive but not always statistically significant impact on tourism when DLTA is used as the dependent variable (Table 5). For instance, tourist arrivals upsurge by 0.15% in response to a 1% increase in mobile phone subscription in Model 1 of Table 5, but in other models, the coefficients are insignificant. The MS coefficients are not significant and, in some cases, negative association when DLTR is the dependent variable (Table 5). It is reasonable to argue that the process of obtaining a mobile phone subscription can be seen as a cumbersome and costly endeavour from a short-term user standpoint. Moreover, a fixed broadband connection has a detrimental influence on TA and TR for developed nations, arguably because the expansion of FBS seems to have reached a steady state level. Therefore, the benefit of further improvement of FBS does not provide much development in tourism demand in the developed countries.

Table 5.

Bias-corrected estimation of linear dynamic panel data models for developed countries.

	Dependent variable: DLTA			Dependent variable: DLTR
	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
One-lagged dependent variable	0.195* (0.103)	0.214* (0.111)	0.144* (0.077)	0.180*** (0.061)	0.155 (0.098)	0.179*** (0.065)
DLMS	0.148** (0.069)	0.158 (0.112)	0.091 (0.064)	−0.033 (0.079)	−0.067 (0.115)	−0.033 (0.017)
DLFBS	−0.015 (0.010)	−0.039 (0.034)	−0.027** (0.011)	−0.031*** (0.012)	0.074* (0.042)	−0.032* (0.017)
DLRGDP	0.736*** (0.107)	0.727*** (0.123)	0.696*** (0.055)	1.202*** (0.247)	1.183*** (0.217)	1.198*** (0.264)
DLCPI	−0.165 (0.283)	0.253 (0.299)	0.077 (0.225)	0.850* (0.458)	1.369*** (0.481)	0.854* (0.466)
YR2020	−1.356*** (0.102)	1.349*** (0.100)	−1.254*** (0.085)	−0.955*** (0.077)	−0.962*** (0.075)	−0.955*** (0.077)
DLPS		0.287 (0.214)			0.096 (0.116)
DLRGDP × DLFBS			0.399*** (0.124)			0.037 (0.393)
Year effect	0.001 (0.001)	0.001 (0.001)	0.001 (0.001)	−0.004*** (0.001)	−0.002 (0.001)	−0.0049*** (0.001)
Arellano–bond test for zero autocorrelation in first-differenced errors of order 2: Z statistics [p-value]	1.656 [0.098]	1.623 [0.105]	1.621 [0.105]	−1.116 [0.264]	−1.025 [0.305]	−1.112 [0.266]
Hansen test of the overidentifying restrictions:	10.141 [0.119]	11.926 [0.064]	10.708 [0.152]	8.013 [0.156]	7.561 [0.272]	8.110 [0.230]
Chi-squared statistics [p-value]	10.141 [0.119]	11.926 [0.064]	10.708 [0.152]	8.013 [0.156]	7.561 [0.272]	8.110 [0.230]
Hausman test:	9.916 [0.128]	13.389 [0.037]	10.949 [0.141]	13.059 [0.023]	7.665 [0.264]	11.696 [0.069]
Chi-squared statistics [p-value]	9.916 [0.128]	13.389 [0.037]	10.949 [0.141]	13.059 [0.023]	7.665 [0.264]	11.696 [0.069]

Note. All coefficients are the fixed effects bias-corrected method of moments estimators proposed by Breitung et al. (2022), except DLTA in Model 3 which is based on the random effects bias-corrected method of moment estimators. Models 2 and 3 are robustness tests that include political stability and the interaction effect between DLRGDP and DLFBS, respectively, to determine consistency using the bias-corrected method of moments estimations in Model 1. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.

Whilst the digitalisation factors may not have an equal impact on tourism development in developed nations, they are uniformly positive and highly noteworthy when viewed from the lens of a developing country’s perspective. For example, an increase of 1% in mobile phone subscriptions and fixed broadband subscriptions—all other conditions remaining constant—results in a respective growth of 0.078% and 0.011% in DLTA (Model 1 of Table 6). Additionally, the volume of tourism revenue (DLTR) displays a corresponding improvement of 0.089% and 0.014% as presented in Model 1 of Table 6. Furthermore, the observed pattern of control variables remains consistent across all previous estimations, with some exceptions. The observed findings highlight that utilising digitalisation in tourism development is crucial in developing countries. In this regard, all facets of digitalisation are deemed positive and, to some extent, significant. Mobile phone subscriptions are relatively more crucial than broadband subscriptions. Moreover, the requirement of extensive infrastructural facilities for fixed broadband utilisation could pose a significant financial challenge for institutions, particularly in underdeveloped economies.

Table 6.

Bias-corrected estimation of linear dynamic panel data models for developing countries.

	Dependent variable: DLTA			Dependent variable: DLTR
	Model 1	Model 2	Model 3	Model 1	Model 2	Model 3
One-lagged dependent variable	0.059 (0.037)	0.036 (0.040)	0.037 (0.042)	−0.030 (0.125)	−0.046 (0.125)	−0.046 (0.126)
DLMS	0.078** (0.032)	0.105** (0.041)	0.047 (0.075)	0.089 (0.054)	0.099* (0.059)	0.072 (0.062)
DLFBS	0.011 (0.012)	0.015 (0.013)	0.015 (0.012)	0.014 (0.013)	0.022 (0.015)	0.022 (0.015)
DLRGDP	0.984*** (0.197)	0.889*** (0.189)	0.758*** (0.229)	1.411*** (0.374)	1.294*** (0.347)	1.236*** (0.379)
DLCPI	0.036 (0.043)	0.041 (0.043)	0.046 (0.043)	0.061 (0.073)	0.093 (0.068)	0.095 (0.067)
YR2020	−1.306*** (0.056)	−1.320*** (0.055)	−1.337*** (0.058)	−1.077*** (0.067)	−1.095*** (0.067)	−1.102*** (0.071)
DLPS		0.172** (0.071)	0.171** (0.073)		0.406*** (0.106)	0.406*** (0.106)
DLRGDP × DLMS			1.131 (1.209)			0.533 (0.715)
Year effect	0.002 (0.001)	0.002 (0.001)	0.002 (0.001)	−0.002 (0.002)	−0.003 (0.002)	−0.002 (0.002)
Arellano–bond test for zero autocorrelation in first-differenced errors of order 2: Z statistics [p-value]	1.737 [0.083]	1.559 [0.119]	1.596 [0.111]	1.144 [0.253]	1.049 [0.294]	1.086 [0.278]
Hansen test of the overidentifying restrictions:	10.522 [0.104]	10.496 [0.162]	14.782 [0.064]	10.818 [0.094]	12.844 [0.076]	14.829 [0.063]
Chi-squared statistics [p-value]	10.522 [0.104]	10.496 [0.162]	14.782 [0.064]	10.818 [0.094]	12.844 [0.076]	14.829 [0.063]
Hausman test:	16.925 [0.010]	21.299 [0.003]	24.585 [0.002]	12.841 [0.046]	16.629 [0.020]	18.176 [0.019]
Chi-squared statistics [p-value]	16.925 [0.010]	21.299 [0.003]	24.585 [0.002]	12.841 [0.046]	16.629 [0.020]	18.176 [0.019]

Note: All estimations are based on the fixed effects bias-corrected method of moments estimators proposed by Breitung et al. (2022). Models 2 and 3 are robustness tests that include political stability and the interaction effect between DLRGDP and DLMS, respectively, to determine consistency using the bias-corrected method of moments estimations in Model 1. ***, ** and * denote significance at the 1%, 5% and 10% levels, respectively.

Robustness results

The results using Arellano-Bond and Arellano-Bover/Blundell-Bond linear dynamic panel data regression are reported in Supplemental Tables 3–8. The results are very similar to bias-corrected dynamic panels, with some exceptions. Mobile phone subscriptions show a positive effect but not always significant. In contrast, fixed broadband results are mostly insignificant. In other words, mobile subscriptions are more crucial for tourism than fixed services. Interestingly, lagged dependent variables show a catching-up effect, but not always significant. However, an increase in real GDP and political stability boosts inbound tourism, but COVID dummy poses a significant hostile effect on the tourism industry overall. Furthermore, mobile subscriptions are vital in attracting tourist flows in developing countries. Other than the methodology for the robustness check, we have used alternative measures for both dependent and independent variables, and the results remain the same with some exceptions. Overall, the findings support our hypothesis that digitalisation is a vehicle for tourism growth. Most importantly, mobile phone subscriptions are vital for enhancing the growth effect. Economic development and political stability can further enrich the effect.

Conclusions and further implications

The tourism industry, as an economic booster, has been fundamentally changing due to the adaptation of digital technologies. This paper empirically investigates the effects of digitalisation on the travel and leisure industries worldwide. The models are tested using a dynamic panel setting, along with the categorisations of developed and developing countries.

Theoretical implications

This study provides some theoretical implications. First, the study applies Institutional and Economic Theory to understand the critical role of institutions in ensuring perfect information flow using technologies. If a country aims to grow inbound tourism, the policymakers should focus on improving its level of digitalisation to maintain a competitive advantage. Furthermore, information technology helps the tourism industry revisit and update organisational policies and strategies in response to changes in the external environment. Second, this study contributes to the development of literature in the fields of tourism and information technology. It proves the critical role that digitalisation can play in improving international tourist flows and revenues for a country. This study has gone a step further and expanded the literature by unravelling the positive linear relationship between digitalisation and tourism by adding broadband connections into account for information, communication and technology adaptation. Therefore, including digitalisation variables into the conventional tourism demand model can improve forecasts of international tourist flows and revenues to a country.

Practical implications

Digitalisation has shown a positive and significant relationship with tourism. Hence, digitalisation should be fundamental to successful tourism development, especially in developing countries. Tourism stakeholders have the responsibility to collaborate and develop policies and applications that will extend the benefits of information, communication and technology to tourism, travel, and hospitality enterprises. The current findings can offer some practical implications. Firstly, this study provides macro-level insights and guidelines for policymakers to select and address the dominating digitalisation factors that can effectively influence the tourism industry. For instance, the relationship between the variables of digitalisation and tourism arrivals, as well as tourism receipts, was found to be positive. Therefore, they should consider investments in information technology to develop sustainable tourism. Secondly, the current results are also useful for the government and other concerned regulatory bodies in formulating legal guidelines ensuring ethical information sharing in the tourism industry. Not only does the digitalisation of tourism facilitate sustainable tourism development, but it also helps the government pursue the Sustainable Development Goals. Thirdly, accelerating digitalisation in the tourism industry will benefit tourists. For example, potential tourists can get updated information on costs, quality of services, accessibility to facilities in host countries, and make rational decisions. Fourthly, government incentives to support information technology industries will improve inter-connectivity among other sectors, such as tourism and hospitality, resulting in a win-win situation for all the stakeholders.

Future research

The empirical study of the nexus between digitalisation and tourism development is rare in the literature. Our current paper has developed a comprehensive understanding of the relationships between digital technology and tourism, and the importance of technology in breaking the barrier of information asymmetry for institutions and society. However, several scopes of this topic can be extended for future research. It would be interesting to apply other contesting theories, such as open system and stakeholder theories, to explore how individual tourism firms can benefit from digital tourism through capital investment in artificial intelligence and how rising artificial intelligence can impact tourism. Furthermore, given the emergence of fake news in the digital world, examining how digitalisation can cause negative externalities will enrich the existing literature.

Supplemental Material

Supplemental Material - Digitalisation as a vehicle of expansion of tourism industry

Supplemental Material for Digitalisation as a vehicle of expansion of tourism industry by Nehad Laila Sanju, Ghialy Yap, Shrabani Saha, and Mahfuzur Rahman in Tourism Economics

Footnotes

ORCID iD

Ghialy Yap

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

Note

Author biographies

Dr. Nehad Laila Sanju serves as an Assistant Professor in the Department of Management Information Systems at the University of Dhaka, Bangladesh. Her research encompasses a range of multidisciplinary topics, such as data-driven information technologies, digital innovation, tourism information systems, and the sustainability of corporate governance, with a significant number of publications in these areas. In this study, she has played a key role in conceptualising ideas, organising data, and analysing findings.

Dr Ghialy Yap is a lecturer at the School of Business and Law, Edith Cowan University, Western Australia. She is passionate about quantitative financial economics and the application of economics in risk analysis and business strategies. Her recent research analyses the effects of political, geopolitical, and climate risks on tourism. Her publications appear in tourism journals such as the Journal of Travel Research, Tourism Economics, Current Issues in Tourism, and the International Journal of Tourism Research. Her current contributions include developing methodology, empirical analysis, and editing.

Dr Shrabani Saha is a Professor of Development Economics at the University of Lincoln, United Kingdom. Her research interests include corruption, economic growth, development economics, international trade issues, and political instability and tourism. The main areas of research focus on the role of institutions in economic development across nations. She has been also engaged in research involving political instability, freedom, and their effects on tourism demands and economic growth. Her contributions in the paper include conceptualization, analyzing the results, helping in writing original draft, and editing.

Dr Mahfuzur Rahman is an Associate Professor at the University of Lincoln, UK. His current academic and research interests are in sustainability and innovation, with a focus on emerging economies. Dr Rahman is also an Academic Adviser to the Commonwealth Scholarship Commission, UK.

References

Adeola

Evans

(2019) Digital tourism: mobile phones, internet and tourism in Africa. Tourism Recreation Research 44(2): 190–202.

Alejo

Galvao

Montes-Rojas

, et al. (2015) Tests for normality in linear panel-data models. STATA Journal: Promoting Communications on Statistics and Stata 15(3): 822–832.

Arellano

Bond

(1991) Some tests of specification for panel data: Monte carlo evidence and an application to employment equations. The Review of Economic Studies 58(2): 277–297.

Athari

Çağlar Onbaşıoğlu

(2019) The effects of global competitiveness and risk on international tourism inflows: the case of western european countries. New Trends in Management Studies.

Athari

Alola

Ghasemi

, et al. (2021) The (un) sticky role of exchange and inflation rate in tourism development: insight from the low and high political risk destinations. Current Issues in Tourism 24(12): 1670–1685.

Athari

Alola

(2023) A global perspective of the role of domestic economic, financial and political risks in inbound tourism. International Journal of Emerging Markets 18(10): 4191–4213.

Baltagi

(2008) Econometric Analysis of Panel Data. Wiley.

Bekele

Raj

(2025) Digitalization and digital transformation in the tourism industry: a bibliometric review and research agenda. Tourism Review 80(4): 894–913.

Bersvendsen

Ditzen

(2021) Testing for slope heterogeneity in stata. STATA Journal: Promoting Communications on Statistics and Stata 21(1): 51–80.

10.

Blundell

Bond

(1998) Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics 87(1): 115–143.

11.

Breitung

Kripfganz

Kazuhiko

(2022) Bias-corrected method of moments estimators for dynamic panel data models. Econometrics and Statistics 24: 116–132.

12.

Buhalis

Law

(2008) Progress in information technology and tourism management: 20 years on and 10 years after the internet – the state of eTourism research. Tourism Management 29: 609–623.

13.

Bun

Windmeijer

(2010) The weak instrument problem of the system GMM estimator in dynamic panel data models. The Econometrics Journal 13(1): 95–126.

14.

Carmignani

(2005) Efficiency of Institutions, Political Stability and Income Dynamics, Public Economics. University Library of Munich. Working Paper 0503007.

15.

Chen

Karavias

Tzavalis

(2022) Panel unit-root tests with structural breaks. STATA Journal: Promoting Communications on Statistics and Stata 22(3): 664–678.

16.

Clemons

Hann

Hitt

(2002) Price dispersion and differentiation in online travel: an empirical investigation. Management Science 48(4): 534–549.

17.

DATAREPORTAL (n.d.) Digital around the world. Retrieved from. https://datareportal.com/global-digital-overview

18.

Dogru

Suess

Sirakaya-Turk

(2021) Why do some countries prosper more in tourism than others? Global competitiveness of tourism development. Journal of Hospitality & Tourism Research 45(1): 215–256.

19.

Dogru

Line

Mody

, et al. (2025) Generative artificial intelligence in the hospitality and tourism industry: developing a framework for future research. Journal of Hospitality & Tourism Research 49(2): 235–253.

20.

Figueredo

Cacho

Thome

, et al. (2017) Using social media photos to identify tourism preferences in smart tourism destination. In: 2017 IEEE International Conference on Big Data (Big Data), 11–14 December 2017, Boston, MA, USA.

21.

Galvao

Montes-Rojas

Sosa-Escudero

, et al. (2013) Tests for skewness and kurtosis in the one-way error component model. Journal of Multivariate Analysis 122: 35–52.

22.

Gelb

Sundaram

(2002) Adapting to “word of mouse.”. Business Horizons 45(4): 15–20.

23.

Habibi

Zabardast

(2020) Digitalisation, education and economic growth: a comparative analysis of Middle East and OECD countries. Technology in Society 63(September): 101370.

24.

Harris

(2017) Digital trends that are transforming the travel industry in 2017. Retrieved from. https://www.huffingtonpost.com/entry/digital-trends-that-are-transforming-the-travelindustry_us_58c6cc3ae4b0c3276fb78790

25.

Hernandez

Lee

(2025) Hospitality and tourism technology and organizational performance: an integrated framework of HTT business value and future research agenda. International Journal of Hospitality Management 126: 104092.

26.

International Telecommunication Union (2024) Measuring digital development – Facts and Figures 2024. ITU. Available at: https://www.itu.int/hub/publication/d-ind-ict_mdd-2024-4/.

27.

Irvine

Anderson

(2008) ICT (information communication technology), peripherality and smaller hospitality businesses in Scotland. International journal of entrepreneurial behavior & research 14(4): 200–218.

28.

Karavias

Tzavalis

(2014) Testing for unit roots in short panels allowing for a structural break. Computational Statistics & Data Analysis 76: 391–407.

29.

Kim

Lee

, et al. (2022) Key initiatives for digital transformation, green new deal and recovery after COVID-19 within the construction industry in Korea. Sustainability 14(14): 8726.

30.

Klimova

Bogomazova

Anoprieva

, et al. (2020) Digital supply chain management in the tourism and hospitality industry: trends and prospects. International Journal of Supply Chain Management 9(3): 843–849.

31.

Kozłowski

Brzozowska-rup

Piotrowska-Piątek

(2021) Information and communication technologies in food and accommodation sector in EU countries: sticker or challenge for tourism development? Technological Forecasting and Social Change 171(120941). Available at: https://doi.org/10.1016/j.techfore.2021.120941

32.

Kripfganz

Breitung

(2022) Bias-corrected estimation of linear dynamic panel data models. In: 28th UK Stata Conference, London, 8 September 2022. Retrieved from. https://www.stata.com/meeting/uk22/slides/UK22_Kripfganz.pdf

33.

Kumar

Stauvermann

Kumar

, et al. (2019) Exploring the effect of ICT and tourism on economic growth: a study of Israel. Economic Change and Restructuring 52(3): 221–254.

34.

Lang

Hyde

(2013) Word of mouth: what we know and what we have yet to learn. Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior 26: 1–18.

35.

Law

Chan

ICC

Wang

(2018) A comprehensive review of mobile technology use in hospitality and tourism. Journal of Hospitality Marketing & Management 27(6): 626–648.

36.

Lee

(2025) Special issue on “Information and communication technology and economic implications in hospitality and tourism”. Tourism Economics 31(2): 159–164.

37.

Lee

Chung

, et al. (2018) Tourists’ happiness: are there smart tourism technology effects? Asia Pacific Journal of Tourism Research 23(5): 486–501.

38.

Lee

Ahn

Shin

, et al. (2021) Integrating technology to service innovation: key issues and future research directions in hospitality and tourism. Journal of Hospitality and Tourism Technology 12(1): 19–38.

39.

Buhalis

(2005) Predicting Internet Usage for Travel Bookings in China. In: Frew

A.J.

(ed). Information and Communication Technologies in Tourism 2005. Vienna: Springer, 429–439. Available at: https://doi.org/10.1007/3-211-27283-6_39

40.

Lin

Shin

(2025) The economic impacts of information and communication technologies in the tourism and hospitality industry: a systematic review of the literature. Tourism Economics 31(2): 165–182.

41.

Löfgren

K -G

Persson

Weibull

(2002) Markets with Asymmetric Information: The Contributions of George Akerlof, Michael Spence and Joseph Stiglitz. The Scandinavian Journal of Economics 104(2): 195–211. Available at: http://www.jstor.org/stable/3441066.

42.

Luo

Feng

Cai

(2004) Information search behaviour and tourist characteristics: the internet vis-a`- vis other information sources. Journal of Travel & Tourism Marketing 17(2/3): 15–25.

43.

Mills

Law

(2004) Handbook of Consumer Behaviour, Tourism and the Internet. Harworth Hospitality Press.

44.

Myovella

Karacuka

Haucap

(2020) Digitalisation and economic growth: a comparative analysis of Sub-Saharan Africa and OECD economies. Telecommunications Policy 44(2): 101856.

45.

North

(1992) Institutions and economic theory. The American Economist 36(1): 3–6.

46.

Oliveira

Martins

(2010) Information technology adoption models at firm level: review of literature. In: The European Conference on Information Systems Management. Academic Conferences International Limited, p. 312.

47.

O’Connor

Buhalis

Frew

(2001) The transformation of tourism distribution channels through information technology. In: Laws

Buhalis

(eds) Tourism Distribution Channels: Practices, Issues and Transformations. Thompson, pp. 332–350.

48.

Pesaran

(2015) Testing weak cross-sectional dependence in large panels. Econometric Reviews 34: 1089–1117.

49.

Polese

Botti

Grimaldi

, et al. (2018) Social innovation in smart tourism ecosystems: how technology and institutions shape sustainable value co-creation. Sustainability 10(1): 140.

50.

Prieger

(2013) The broadband digital divide and the economic benefits of mobile broadband for rural areas. Telecommunications Policy 37(6–7): 483–502.

51.

Puciato

Łoś

Mrozowicz

(2013) Franchising as a way of reducing moral hazard in the tourism market. ARGUMENTA OECONOMICA 1(30).

52.

Quaglione

Matteucci

Furia

Marra

Pozzi

(2020) Are mobile and fixed broadband substitutes or complements? New empirical evidence from Italy and implications for the digital divide policies. Socio-Economic Planning Sciences 71. Available at: https://doi.org/10.1016/j.seps.2020.100823

53.

Saha

Yap

(2014) The moderation effects of political instability and terrorism on tourism development: a cross-country panel analysis. Journal of Travel Research 53(4): 509–521.

54.

Sarıışık

Sarı

Halis

(2011) Tourism Sector in Order to Recovering From the Recession: Comparison Analyses for Turkey. Procedia Social and Behavioral Sciences 24: 181–187. Available at: https://doi.org/10.1016/j.sbspro.2011.09.070

55.

Seetanah

Fauzel

(2025) The moderating role of digitalisation in the tourism-growth nexus: evidence from small island economies. Journal of Policy Research in Tourism, Leisure and Events 17(2). Available at: https://doi.org/10.1080/19407963.2023.2201888

56.

Soares

ALV

Mendes-Filho

Gretzel

(2021) Technology adoption in hotels: applying institutional theory to tourism. Tourism Review 76(3): 669–680.

57.

Song

(2008) Tourism demand modelling and forecasting- A review of recent research. Tourism Management 29(2): 203–220.

58.

Spencer

(2019) ICT and Caribbean Tourism: Challenges and Opportunities for Small Island Developing States. Palgrave Macmillan, 95–114.

59.

Tay

(2020) The impact of adoption of information technologies on international tourism. Proceedings of the 2020 6th International Conference on E-Business and Applications (ICEBA 2020), Kuala Lumpur, Malaysia, 25–27 February 2020. Association for Computing Machinery, 60–64. https://doi.org/10.1145/3387263.3387265

60.

Ukpabi

Karjaluoto

(2017) Consumers’ acceptance of information and communications technology in tourism: a review. Telematics and Informatics 34(5): 618–644.

61.

Wang (2009) The impact of crisis events and macroeconomic activity on Taiwan’s international inbound tourism demand. Tourism Management 20: 75–82. Available at: https://doi.org/10.1016/j.tourman.2008.04.010

62.

World Economic Forum (2020) How has technology changed- and changed us-in the past 20 years? Retrieved from. https://www.weforum.org/stories/2020/11/heres-how-technology-has-changed-and-changed-us-over-the-past-20-years/

63.

WTTC (2017) Travel and Tourism Economic Impact 2017. World travel and tourism council. Yemen. Retrieved from. https://www.wttc.org/

64.

Xiang

(2018) From digitisation to the age of acceleration: on information technology and tourism. Tourism Management Perspectives 25: 147–150.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.37 MB