Exploring the effects of tourism capital investment on income inequality and poverty in the European Union countries

Abstract

The role of tourism in reducing inequalities has been studied and investigated in the literature. However, the specific effect of tourism investments on decreasing income inequalities, particularly given its capacity to alleviate poverty, has been little studied. Moreover, while existing research focuses on developing countries, this influence may be equally relevant in developed economies. Therefore, this article aims to understand the effects of tourism capital investment on income inequality and poverty using data from 2006 to 2019 for a sample of 24 European Union nations. To this end, the panel corrected standard errors methodology was carried out to account for data characteristics. The findings of the primary analysis suggest that tourism capital investment, international tourist arrivals, the human development index, and trade globalization contribute to mitigating income inequality and poverty across the EU region. Conversely, the age dependency ratio has a positive effect on both factors. The robustness check confirms that tourism investments-related indicators (tourism capital investment and travel and tourism direct contribution to employment) reduce income inequality and poverty in EU countries.

1 Introduction

When a significant portion of money is disproportionately concentrated among a small segment of the population, income disparities increase, and poverty issues may persist, creating numerous challenges across societal groups and population segments (Michálek and Výbošťok 2019). These inequalities between the richest and poorest have been exacerbated globally due to the transformation of the world economy, driven by increasing globalization, technological advancements, and the recent effects of the pandemic (McCaughey 2024; Soava et al. 2020).

Although these disparities are often emphasized when examining emerging economies, they are equally significant in the world’s wealthiest nations (Michálek and Výbošťok 2019). A notable example is the European Union (EU), where member states are classified as developed mainly due to their extensive economic influence—accounting for approximately 15% of global Gross Domestic Product (GDP) and 14% of international trade—and also due to their strong commitment to education and health, which underscores their status as leaders in advanced human development (Eurofound 2024).

However, economic inequalities within the EU have been further exposed, mainly due to the recent recession and the cost-of-living crisis. In fact, by 2014, over 122 million individuals—nearly 24.4% of the EU-28 population—were classified as at risk of poverty and social exclusion, marking a rise of 5 million since the financial crisis (Savoia 2024). According to McCaughey (2024), income inequalities become increasingly visible across various societal divisions: (i) between rich and poor, (ii) between men and women, (iii) between rural and urban areas, and (iv) between young and old.

Income disparities are increasing in older EU member states, while in newer members, inequalities between rich and poor have decreased. However, the share of individuals below the poverty income threshold has risen across most of these nations between 2006 and 2021 (McCaughey 2024). Simultaneously, the gender pay gap has experienced minimal variations  over the last decade, with women still earning, on average, 13% less per hour than men in 2021 (McCaughey 2024). Moreover, income inequality can also stem from differences between rural and urban areas. Specifically, employment in urban areas has increased more rapidly than in rural areas. In almost all member states, people living in cities, towns, and suburbs have been shown to receive higher incomes than those in rural areas (McCaughey 2024; Eurostat 2022a). As a result, the rural–urban gap in median incomes has deepened by nearly 20% in the past 10 years (McCaughey 2024). Finally, economic inequalities across generations have also intensified over the past decade as younger individuals face higher rates of unemployment and poverty (Chen et al. 2018). This growing disparity is attributable primarily to younger people's incomes being predominantly tied to employment (or its absence). In contrast, older people mainly rely on pensions, offering them greater economic stability (McCaughey 2024).

According to the previous statements, and despite economic progress in the EU, inequality and poverty issues seem to continue, which has prompted global initiatives such as the adoption of the 2030 United Nations Sustainable Development Goals (SDGs) (Michálek and Výbošťok 2019; Raza and Shan 2017). As part of this effort, tourism development has been recognized as an effective tool (UNWTO, 2018). Indeed, being pro-poor, tourism helps decrease economic disparities by reducing poverty, i.e., providing economic opportunities to the poorest segments of the population within this industry (Subramaniam et al. 2022). For instance, the expansion of this sector can significantly contribute to job creation, particularly among vulnerable groups—such as low-income individuals, women, and young people. In 2023, 58% of those employed in the tourism industry across the EU were women, and young individuals (aged 15–24) made up 11% of the workforce in this sector (Eurostat 2024). Hence, by providing more inclusive job opportunities and being a major source of income in these nations, tourism development can help reduce poverty and, consequently, economic disparities among socioeconomic groups, genders, and generations. Furthermore, with around 44% of the EU territory being predominantly rural, tourism growth can create a pathway for further economic development and revitalization in these regions by offering new job opportunities, helping to reduce migration, and attracting new investments (Eurostat 2022b; Rajović and Bulatović 2017).

However, if not well managed, the tourism industry can exacerbate income inequality and poverty due to its seasonality, precarity, and labor-intensive nature, while over-reliance on this sector can leave communities vulnerable to economic downturns (Bürgisser and Di Carlo 2023; Wang and Tziamalis 2023). Targeted investment and financing in this sector are crucial since they support the transition to a low-carbon economy and foster socially inclusive development (OECD 2018). Specifically, investing in the tourism sector can help increase the income of the existing workforce (Dogru and Bulut 2018). Moreover, through these investments, resilient and sustainable infrastructures can be built, contributing to the development of rural areas, helping to attract more tourists, and creating long-term employment opportunities beyond peak tourist seasons (Jeje 2021; Rajović and Bulatović 2017). Investment in training and skills programs within this industry is equally important, as it can improve job quality, ensuring workers access decent wages and career growth opportunities (Stacey 2015). Through these mechanisms, tourism capital investment can ensure that the benefits of the industry are distributed more equitably, contributing to income equality and poverty alleviation.

As was previously stated, pro-poor tourism can reduce income disparities through poverty alleviation. Therefore, this study aims to understand better the impacts of tourism capital investments on income inequality and poverty in EU nations. It led us to the following research questions: (i) Can tourism capital investment affect income inequality across the European Union? and (ii) Can tourism capital investment contribute to poverty alleviation across the European Union? To address these questions, the panel corrected standard errors (PCSE) estimator was used, as it is suitable for dealing with the characteristics of the data.

The organization of this article is as follows: the next section reviews the prior literature. Section 3 presents the data and research methodology. Section 4 offers a detailed analysis of the results, and SubSect. 4.1 checks the robustness of the outcomes. The study’s main findings are closely examined and discussed in Sect. 5. Lastly, Sect. 6 presents the main conclusions and policy implications derived from the outcomes.

2 Literature review

The relationship between tourism and economic growth is well-recognized in the literature (e.g., Dogru and Bulut 2018). Consequently, given its significant role in economic development, tourism expansion can potentially influence income inequality and poverty on a global scale (Kumail et al. 2022). Blake et al. (2008) and Incera and Fernandez (2015) highlighted the extent of this influence, identifying three key mechanisms through which tourism development affects income inequality and poverty.

First, tourism can increase demand for products and services such as accommodation, transportation, cultural services, and food and beverages, leading to higher prices (Fang et al. 2021). This phenomenon, known as the "price effect," tends to exacerbate inflation, primarily affecting wealthier households, who consume more goods and services (Fang et al. 2021; Uzar and Eyuboglu 2019). The impact on lower-income families is primarily influenced by their consumption of tourism-related goods and services, which are typically low (Njoya and Seetaram 2018). Nevertheless, a pronounced price rise for primary goods can disproportionately affect poorer individuals by reducing their purchasing power (Uzar and Eyuboglu 2019). Consequently, these dynamics must be carefully considered when evaluating the overall effects of tourism on income inequality and poverty.

Concerning the “income effect,” it suggests that when the tourism sector develops, it can significantly contribute to reducing income inequality and poverty by generating income through the creation of employment for low-skilled individuals, increasing foreign exchange earnings, and strengthening economic connections (Uzar and Eyuboglu 2019). Notably, tourism generates employment, particularly for unqualified workers within the informal sector, and boosts the income of poor individuals by expanding opportunities for self-employed workers and supporting locally-owned enterprises (Yergeau 2020; Uzar and Eyuboglu 2019). However, the extent of these benefits depends on the size of the informal sector and the prevalence of self-employment among workers. While tourism development can promote self-employment in sectors such as accommodation, food and beverage, and transportation, it carries the risk of exacerbating income inequality and poverty given that the earnings of self-employed individuals and tourism-related companies often surpass those of low-skilled workers (Kumail et al. 2022; Fang et al. 2021).

Additionally, in cases where foreign firms or multinational corporations dominate the tourism industry, external entities capture most of the income. This concentration of wealth leads to a disproportionate distribution of tourism benefits, which limits the ability of local businesses to enter the market and reduces employment opportunities for residents (Uzar and Eyuboglu 2019; Alam and Paramati 2016).

Finally, the third effect is the "tax revenue effect," which relates the expansion of the tourism sector to income inequality and poverty through the taxation of goods and services within this industry (Kumail et al. 2022). As a result, if governments strategically increase expenditures on tourism-related activities, such as infrastructure investment, it can boost tax revenue (Blake et al. 2008; Incera and Fernandez 2015). Consequently, this growth in tax revenue generated from tourism can help reduce income inequality and poverty by redistributing the collected taxes to support disadvantaged groups.

However, there is no consensus on the outcomes; some studies suggest that tourism can mitigate income inequality and poverty through the “price effect,” “income effect,” and “tax revenue effect,” while others have reached opposite conclusions. This disparity in the results can be attributed to the fact that tourism development's influence on income distribution and poverty largely depends on the magnitude of each mechanism, with the level of national income playing a key role in determining the overall outcome (Kumail et al. 2022).

2.1 Empirical research on tourism development and income inequality

In line with the previous statement, several authors have investigated the impacts of tourism on income inequality across different countries/regions, employing various estimation techniques and finding mixed results (e.g., Subramaniam et al. 2022; Uzar and Eyuboglu 2019). For instance, Subramaniam et al. (2022) studied the impacts of tourism development on income inequality in 9 countries between 2001 and 2016. The empirical findings reveal that tourism revenues are major drivers of decreasing income inequality. On the contrary, Uzar and Eyuboglu (2019) examine the same relationship for Turkey, and their results indicate that tourism revenues increased income inequality between 1974 and 2015.

The literature focused on the EU, and the region of interest in this research appears limited. However, some authors have included this group of countries in their investigations as highly developed economies (e.g., Ghosh and Mitra 2021; Nguyen et al. 2020). As an example, Nguyen et al. (2020) explored the effects of the tourism sector on income inequality across 97 countries between 2002 and 2014. These authors provided estimates for the entire sample and further categorized their analysis into three subsamples: 30 low and lower-middle-income economies, 25 upper-middle-income economies, and 42 high-income economies. The findings suggest that tourism development (measured by six tourism indicators) increases income inequality for low- and lower-middle-income and upper-middle-income economies but reduces income inequality across high-income nations. Following the same guideline, Ghosh and Mitra (2021) investigated the impacts of the tourism industry on income inequality for a panel of 41 emerging, developed, and highly developed countries from 1995 to 2016. The outcomes indicate that tourism receipts reduce income inequality in developing countries, increase income inequality in developed countries, and have a non-statistically significant effect in highly developed countries.

2.2 Empirical research on tourism development and poverty

Concerning the impacts of tourism on poverty, the results on this topic are far from conclusive (see Zhao 2020; Mahadevan and Suardi 2017). Zhao (2020) explored the influence of tourism on poverty reduction in 29 Chinese provinces between 1999 and 2014. Using tourism receipts and arrivals to represent tourism development, these authors argued that both indicators contribute to poverty alleviation. Moreover, Mahadevan and Suardi (2017) studied the impact of tourism growth on poverty for a panel of 13 tourism-intensive economies between 1995 and 2012. This study provides empirical evidence that tourism receipts fail to alleviate poverty.

Regarding the sample considered in this investigation, to the best of our knowledge, only Boghean and State (2019) explored the relationship between tourism development and poverty reduction in EU countries. First, they examined the correlation between tourism revenues and poverty levels, demonstrating a direct connection between tourism receipts and poverty levels. Second, they conducted a cluster analysis, and the findings revealed distinct clusters of countries with varying outcomes. For instance, where tourism plays a substantial role in the economy (such as in Croatia or Greece), the sector has a stronger impact on reducing poverty levels since it significantly contributes to GDP and supports local employment. Conversely, in countries with lower tourism revenues (such as Bulgaria or Romania), the impact of the tourism sector on poverty alleviation is less pronounced.

2.3 Capital investment in travel and tourism, income inequality and poverty

As noted in existing literature, tourism revenues are often considered the main indicator representing the tourism sector and its development; however, tourism capital investment can be an equally important indicator to proxy these factors (Kumail et al. 2022).

Hence, considering the main focus of this paper, previous literature exploring the effects of tourism capital investment on income inequality and poverty was considered. There is a limited amount of literature dedicated to the direct impacts of tourism investments on income inequality; however, the research of Kumail et al. (2022), Fang et al. (2021), and Chi (2020) can be cited. According to Kumail et al. (2022), who evaluated the influence of tourism development on income inequality in South Africa from 1996 to 2020, tourism capital investment has a diminishing effect on income inequality. Additionally, Fang et al. (2021) analyzed the impacts of tourism development on income inequality in developing and developed economies between 1995 and 2014, including travel and tourism capital investment as one of the dimensions of tourism development. The outcomes reveal that tourism capital investment decreases income inequality in the full sample and developing countries, while it does not exhibit statistical significance in developed countries. Similarly, Chi (2020) also examined this relationship in emerging and developed economies between 1995 and 2015. The results align with the findings of Fang et al. (2021).

The prior literature on tourism capital investment and poverty also appears scarce. For instance, Banerjee et al. (2015) researched the potential relationship between tourism investments and poverty in Haiti. They found that the US$36 million tourism investment decreased unemployment and headcount poverty. In addition, Deskins and Seevers (2011) analyzed the impact of public spending on tourism promotion in fostering tourism development in the United States between 1985 and 2003. They found that increased public spending on tourism promotion led to higher tourism and employment growth in U.S. states with low initial tourism expenditure. This study can endorse the idea that public investment in tourism can help alleviate poverty by stimulating economic activity and job creation, especially in underserved areas. In particular, there is a lack of studies for the EU that directly investigate the relationship between tourism investments and poverty levels.

Based on this brief literature review, it becomes clear that numerous empirical studies have explored the influence of tourism development on income inequality and poverty. However, some research gaps remain. First, the findings are inconclusive, highlighting the need for further investigation. Second, instead of focusing on the effects of tourism revenues, this article explores the impacts of tourism capital investment, international tourist arrivals, and travel and tourism's direct contribution to employment on income inequality and poverty. Third, this analysis aims to determine whether tourism capital investment contributes to a more equitable income distribution and if this can be connected with its ability to overcome poverty. Finally, as previously mentioned, much of the research has concentrated on emerging regions and nations. Simultaneously, the relationship in developed economies, including those in the European Union, has been largely underexplored. Therefore, additional research within this region is necessary to provide a more comprehensive understanding of the relationship between tourism, income inequality, and poverty. Despite benefiting from advanced economies and ranking higher on national development indicators, these countries can still experience underlying issues such as high unemployment and poverty in some areas.

3 Data and methodology

This study will analyze the effects of tourism capital investment on income inequality and poverty, using annual data from 2006 to 2019 for a panel of 24 EU countries. More precisely, the countries included were Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Slovak Republic, Slovenia, Spain, and Sweden. These nations were selected because the tourism sector significantly boosts their economies and share a common cultural background (Silva et al. 2024; Paramati et al. 2018). The chosen time horizon focuses on the pre-COVID-19 years.

The econometric techniques were computed using the statistical software STATA 17.0. Table 1 displays the variables, the definitions, and the sources.

Table 1 Variables description

The dependent variable of the first panel will be represented by the Gini coefficient of equivalized disposable income (GINI) to measure income inequality (e.g., Fang et al. 2021), i.e., how income is distributed across a population compared to a scenario of perfect equality (Hasell 2023). This indicator was retrieved from "Eurostat [tessi190]," and its values range between 0 and 100, with 0 indicating perfect income equality and 100 indicating perfect income inequality.

Notably, the Gini index of equivalized disposable income was chosen because it reflects income after taxes and transfers, making it a better indicator of the amount individuals have available for saving and spending.

The dependent variable of the second panel will be measured by the share of people below 40% of median income (POV), which was sourced from "Our World in Data." This variable will represent the poverty rate, i.e., the percentage of individuals living in households with an income below 40% of the median (e.g., Granados 2013).

Concerning the interest variables, the capital investment in travel and tourism per capita (CIPC) will represent the sector's total gross fixed capital formation (Demir et al. 2020). It encompasses the following categories: tourism-specific fixed assets, non-tourism-specific assets invested by tourism industries, and tourism-related infrastructure (OECD et al. 2017). The indicator CIPC was achieved through the ratio between "CI"—collected from the "World Travel & Tourism Council"—and the total population (POP)—collected from the "World Development Indicators—World Bank."

In addition, the variable international tourist arrivals (TA) will also be incorporated as an interesting variable in the analysis since it proved to be positively affected by the investments within the tourism sector (Jeje 2021). This indicator can provide valuable insights into the sector's performance and the broader impact of such investments. TA was retrieved from the "World Development Indicators—World Bank," it measures overnight visitors traveling to a country outside their usual residence for up to 12 months for purposes unrelated to earning income in the destination country.

The control variables will be the human development index (HDI), the age dependency ratio (AGE) as a percentage of the working-age population, and trade globalization (TGLOB). HDI was sourced from the "Human Development Reports." At the same time, AGE and TGLOB were retrieved from the "World Development Indicators—World Bank" and the "KOF Index of Globalization," respectively. All control variables were chosen based on prior research.

First, the human development index (HDI) is a composite index that incorporates education, life expectancy, and per capita income indicators, and its overall score ranges from 0 to 1. It should be stressed that most researchers examine education's impacts on income inequality and poverty (e.g., Pal 2024; Omar and Inaba 2020). However, given that HDI provides a more comprehensive set of observations in data, it was used in this article. This index has been proven to affect income inequality (Santiago et al. 2022; Theyson and Heller 2015). For instance, Theyson and Heller (2015) studied the impact of HDI on income inequality. They concluded that during the early stages of a country's development, improvements in HDI led to a decrease in income inequality. However, as a country develops, this is followed by a brief period of increased inequality and a further decline in income inequality. Furthermore, the human development index can effectively measure human resources' well-being and is highly correlated with poverty (Cashin et al. 2001). Studies have demonstrated that improvements in HDI are associated with reduced poverty levels (e.g., Lestari et al. 2022; Widiastuti et al. 2022).

When it comes to the age dependency ratio (AGE), the ratio of young and old dependents to the working-age population, it seems to be an important influencing factor of income inequality, as it is strongly related to the distribution of household income and social expenditure (Fang et al. 2021; Mao 2016). The main conclusion is that the age dependency ratio positively affects income inequality (e.g., Fang et al. 2021; Wang and Naveed 2021; Mao 2016). For example, Fang et al. (2021) conclude that the age dependency ratio has a rising effect on income inequality in developing and developed countries. Moreover, the age dependency ratio can also affect poverty rates (see Belu et al. 2024; Cruz and Ahmed 2018). Following the outcomes of Belu et al. (2024), the old-age dependency ratio appears to have a rising effect on the at-risk-of-poverty rate among European Union countries.

Lastly, trade globalization (TGLOB) is a sub-dimension of economic globalization and includes the following components: trade in goods, trade in services, and trade partner diversity (for further details, see Gygli et al. 2019). Its relationship with income inequality has been the subject of many investigations (e.g., Han et al. 2023; Asteriou et al. 2014; Çelik and Basdas 2010). For instance, we can cite the study of Asteriou et al. (2014), who explored the impact of trade globalization (proxied by trade openness) on income inequality in the European Union region. These authors found that trade openness contributes to reducing income inequality in all EU countries, which is consistent with the findings of Han et al. (2023). TGLOB seems related to income inequality and poverty (e.g., Signoret et al. 2020; Ganić, 2019). Evidence of this is provided by Signoret et al. (2020), who investigated the effect of trade openness on poverty rates in EU regions. These authors found that trade and poverty are strongly and negatively correlated, with areas more open to trade exhibiting lower poverty rates.

As mentioned earlier, this investigation aims to test empirically the impacts of tourism capital investment on both income inequality and poverty. Therefore, following the best econometric practices, a set of preliminary tests must be performed before the estimation to check the adequacy of the data, namely: (i) the Pesaran Cross-sectional Dependence (CD) test; (ii) the Maddala and Wu (MW) test for variables without cross-sectional dependence; (iii) the Cross-sectionally augmented Im-Pesaran-Shin (CIPS) test for variables with cross-sectional dependence; (iv) the correlation matrix; and (v) the Variance Inflation Factor (VIF) test.

The descriptive statistics are displayed in Table 2. All variables were converted into natural logarithms (represented by "L"). From the descriptive statistics outcomes, it is possible to observe a lack of observations for international tourist arrivals (LTA). Specifically, these gaps for LTA occurred in Estonia in 2006 and 2007 and in Lithuania in 2006.

Table 2 Descriptive statistics

First, the Pesaran CD test was carried out, and the presence of cross-sectional dependence for all variables, except for LGINI, was confirmed. The following step was the performance of 1st and 2nd generation unit root tests. The Maddala and Wu 1st generation unit root test were computed for LGINI, which does not exhibit cross-sectional dependence. The 2nd generation unit root CIPS test was applied to the remaining variables. The findings revealed that all variables are stationary. Lastly, the correlation matrix and the VIF test were also computed to assess collinearity and multicollinearity, respectively. The outcomes indicate that correlation and multicollinearity are far from a problem for estimating both models. The results of the preliminary tests are not included in the manuscript but can be provided upon request.

Considering that the characteristics of the data will not skew the outcomes, the functional forms of Model I and Model II can be represented as follows, respectively (Neves et al. 2018):

$${LGINI}_{it}={\beta }_{0}+{{\beta }_{i1}LCIPC}_{it}+{{\beta }_{i2}LTA}_{it}+{{\beta }_{i3}LHDI}_{it}+{{\beta }_{i4}LAGE}_{it}+{{\beta }_{i5}LTGLOB}_{it}+{\varepsilon }_{it}$$

(1)

$${LPOV}_{it}={\Upsilon }_{0}+{{\Upsilon }_{i1}LCIPC}_{it}+{{\Upsilon }_{i2}LTA}_{it}{{+\Upsilon }_{i3}LHDI}_{it}+{{\Upsilon }_{i4}LAGE}_{it}+{{\Upsilon }_{i5}LTGLOB}_{it}+{\mu }_{it}$$

(2)

where ${LGINI}_{it}$ and ${LPOV}_{it}$ represent the dependent variable of Model I and Model II, respectively; ${\beta }_{0}$ and ${\Upsilon }_{0}$ represent the intercept; ${\beta }_{ik}$ and ${\Upsilon }_{ik}$ denote the estimated coefficients, which represent the relationship between explanatory variables and the dependent variable, with $k=1,\dots ,5$; and ${\varepsilon }_{it}$ and ${\mu }_{it}$ designate the error terms.

In addition to inspecting data, a battery of specification tests was computed, namely: (i) the F-test; (ii) the Breusch and Pagan Lagrangian multiplier (LM) test for random effects; (iii) the Hausman test; (iv) the Modified Wald test; (v) Pesaran's test of cross-sectional independence; and (vi) the Wooldridge test.

The F-test was conducted to assess fixed effects, while the Breusch and Pagan LM test was used to check for random effects. The Hausman test was then performed to compare fixed and random effects estimators. The outcomes of these tests are presented in Tables 3 and 4 (see the next section). The F-test and Breusch and Pagan LM test revealed that fixed and random effects are preferred over OLS. Hence, the Hausman test was computed, confirming that the fixed effects are the appropriate estimators for Model I and Model II. The subsequent specification tests were determined based on the Hausman test results.

The Modified Wald test was performed to check the presence of heteroscedasticity, the Pesaran's test to check the presence of contemporaneous correlation, and the Wooldridge test to check the presence of first-order autocorrelation. The results suggested the presence of heteroscedasticity and first-order autocorrelation in both models and contemporaneous correlation only in Model II. The outcomes of the specification tests are not displayed in the manuscript to preserve space but are available upon request.

Despite no evidence of contemporaneous correlation in Model I, individual analysis of the variables suggests that cross-sectional dependence is present in most of them (Neves et al. 2018). As a result, the traditional fixed effects estimator will produce inefficient coefficients and biased standard errors (Marques and Fuinhas 2012). To deal with these disturbances, and given that in our sample, the number of years ($T=14$) is smaller than the number of countries ($N=24$), the PCSE are the most appropriate estimator.

PCSE, proposed by Beck and Katz (1995), addresses the correlation in standard errors within units, across groups, and panel heteroscedasticity. A positive point of this method is its ability to use the full structure of panel data, including information from all periods, to more accurately estimate residuals for each unit when calculating the variance of the error term. Moreover, this method is preferred to both Ordinary Least Squares (OLS) and the fixed effects estimator since, as mentioned earlier, it produces efficient coefficients and correct standard errors when non-spherical errors are present.

4 Empirical results

Remembering the goal of this article, which is to assess the impact of tourism capital investment on income inequality and check if this is connected to its ability to influence poverty, two models were estimated, and this section presents the outcomes obtained. Precisely, four structures were specified for each model: (i) OLS to serve as a benchmark; (ii) PCSE(HET) to deal with heteroscedasticity; (iii) PCSE(HET-AR1) to address both heteroscedasticity and first-order autocorrelation; and (iv) PCSE(CORR) with options for contemporaneous correlation.

Table 3 presents the results of Model I, which reflect the impacts on income inequality. Strong stability and consistency can be observed across estimators, as the coefficients exhibit similar signals and magnitudes. Furthermore, the Wald test also confirms the consistency of the models since it proved to be statistically significant.

Table 3 Results of OLS and PCSE—Model I

Examining Table 3 reveals that increased capital investment in travel and tourism and rising international tourist arrivals contribute to reducing income inequality in the European Union across all estimates. Furthermore, the outcomes reveal that the variable HDI is also highly significant in achieving income equality. More precisely, the results reveal that a 1% increase in the HDI corresponds to a reduction of approximately 1% in income inequality among European Union nations. Concerning the impact of trade globalization, it also reduces income inequality. In other words, higher levels of trade globalization have a diminishing effect on the Gini coefficient. Conversely, the age dependency ratio drives income inequality in this region, meaning that a higher proportion of dependents relative to the working-age population significantly exacerbates economic disparities.

Model II was estimated to evaluate whether the role of tourism capital investment in influencing income inequality is linked to its capacity to affect poverty (see Table 4). All estimations of Model II also exhibit strong consistency for the same reasons—the similarity of their impacts and the statistical significance of the Wald test.

Table 4 Results of OLS and PCSE—Model II

The findings indicate that a higher proportion of capital investment in travel and tourism and international tourist arrivals are significantly associated with reductions in poverty rates across EU nations, as evidenced by the consistently negative and statistically significant coefficients across all model specifications. Similarly, HDI proves to be a key driver of poverty alleviation, implying that major human development is reducing the poverty rate in EU countries. On the other hand, the results exhibit a positive and statistically significant effect from the age dependency ratio to the poverty rate, suggesting that higher levels of dependents relative to the working-age population increase poverty levels. Lastly, trade globalization has a negative impact on the poverty rate, meaning that higher levels of trade globalization decrease poverty in this region.

For further clarity, Fig. 1 visually summarizes the key findings from Tables 3 and 4.

Fig. 1

Summary of the main results. Notes: Own elaboration

4.1 Robustness check

In this subsection, the robustness and consistency of the previously obtained results are evaluated. Therefore, both models will be estimated using the same methodology but with different specifications. In the first panel, the dependent variable will be replaced by the Gini coefficient of equivalized disposable income before social transfers (INEQ), obtained from the "Eurostat [ilc_di12c]." This indicator is the sum of the 'Gini coefficient of disposable income' and 'tax and transfer effects' and measures market income inequality (Kim 2019).

Concerning the dependent variable of the second panel, it will be replaced by the poverty headcount ratio at $2.15 a day (2017 PPP) (ABS_POV) and was sourced from the "World Development Indicators – World Bank" to measure poverty (see Anser et al. 2020). This variable represents the percentage of individuals with an income below $2.15 per day. The values of the poverty headcount ratio range from 0%, representing no poverty, to 100%, representing extreme poverty across the entire population.

This indicator corresponds to the previous $1.90/day poverty line (2011 PPP). As prices have risen, one international dollar in 2017 buys fewer goods and/or services than in 2011. Therefore, to account for inflation and changes in purchasing power, the World Bank adjusted the poverty line in 2022 (Hasell 2022).

Additionally, instead of TA, the travel and tourism direct contribution to employment (EMP) was obtained from the "World Travel & Tourism Council." It will be used jointly with the main indicator—CIPC. This dimension was included since it also proved to be significantly and positively influenced by tourism investments and will serve as a measure of tourism's capacity to generate economic activity (Nguyen et al. 2024; Barišić and Cvetkoska, 2019).

Finally, instead of TGLOB, both models will use the financial development index (FD) as a control variable. This measure, sourced from the "International Monetary Fund," comprises nine sub-indices that capture the development of financial institutions and markets, focusing on their depth, accessibility, and efficiency (de Haan et al. 2022). FD has been shown to affect income inequality (see Qehaja-Keka et al. 2023; Baiardi and Morana 2018). For example, Qehaja-Keka et al. (2023) examined the influence of financial development on income inequality in the 27 Member States of the EU. They concluded that economic inequality in these nations tends to increase as financial systems become more advanced. At the same time, a growing body of literature addresses the connection between financial development and poverty, and the results are far from consensual (e.g., de Haan et al. 2022; Kaidi et al. 2019). Following the results of de Haan et al. (2022), who examined this nexus for a panel of 84 countries, financial development does not directly influence poverty reduction, either in the full sample or when the analysis is divided into developed and developing economies.

Therefore, the equations representing Model III and Model IV can be expressed as follows, respectively:

$${LINEQ}_{it}={\beta }_{0}+{{\beta }_{i1}LCIPC}_{it}+{{\beta }_{i2}LEMP}_{it}+{{\beta }_{i3}LHDI}_{it}+{{\beta }_{i4}LAGE}_{it}+{{\beta }_{i5}LFD}_{it}+{\varepsilon }_{it}$$

(3)

$${LABS\_POV}_{it}={\Upsilon }_{0}+{{\Upsilon }_{i1}LCIPC}_{it}+{{\Upsilon }_{i2}LEMP}_{it}{{+\Upsilon }_{i3}LHDI}_{it}+{{\Upsilon }_{i4}LAGE}_{it}+{{\Upsilon }_{i5}LFD}_{it}+{\mu }_{it}$$

(4)

where ${LINEQ}_{it}$ and ${LABS\_POV}_{it}$ denote the dependent variable of Model III and Model IV, respectively; ${\beta }_{0}$ and ${\Upsilon }_{0}$ denote the intercept; ${\beta }_{ik}$ and ${\Upsilon }_{ik}$ represent the coefficients, which translate the linkage between independent indicators and the dependent variable, with $k=1,\dots ,5$; and ${\varepsilon }_{it}$ and ${\mu }_{it}$ designate the error terms.

Preliminary and specification tests were conducted to assess the characteristics of our data and to confirm the suitability of the PCSE methodology, respectively. The existence of cross-sectional dependence in all variables was confirmed except for ABS_POV. Moreover, all variables were found to be stationary, with low levels of collinearity and multicollinearity among them. The results of these tests are available upon request.

Regarding the outcomes of the specification tests, the F-test and the Breusch and Pagan LM test indicate that fixed and random effects are preferred over OLS, respectively. The Hausman test corroborates that the fixed effects estimator is the most appropriate for Model III and Model IV. These tests are presented in Table 5.

Table 5 Results of robustness check—Model III and Model IV

Moreover, the Modified Wald test and the Wooldridge test indicate the presence of heteroscedasticity and first-order autocorrelation in both models—lastly, the findings of Pesaran’s test point to contemporaneous correlation only in Model IV. Once again, the outcomes of this group of specification tests are available upon request.

After confirming the suitability of the PCSE methodology, Model III and Model IV were estimated, and the results are displayed in Table 5.

The findings demonstrate the strong robustness of both models across all estimations, highlighted by the similarity of their effects and the statistical significance of the Wald test.

The findings generally remain consistent with those displayed in Tables 3 and 4. More precisely, the outcomes of Table 5 confirm that indicators related to tourism investments (in this case, represented by tourism capital investment and the direct contribution of tourism to employment) reduce income inequality and poverty across European Union nations. Additionally, as in previous specifications, the human development index has a negative and statistically significant effect on income inequality and poverty, while the age dependency ratio positively impacts both factors. The results also suggest that the financial development index has an augmenting effect on income inequality, while it shows an absence of statistical significance for poverty in these countries.

5 Discussion and comparison between models

This section discusses the findings to understand how tourism capital investment promotes income equality and alleviates poverty in the EU. The answer to the first research question of this paper is that tourism capital investment can be an essential instrument for reducing income inequality in European Union nations. This result contrasts with the findings of Chi (2020) and Fang et al. (2021), who found no statistical significance in highly developed countries. Two factors may explain this discrepancy: (i) the period analyzed by these authors is different from the one used in this investigation, extending only to 2015 and 2014, respectively; and (ii) their investigations do not focus exclusively on the EU, which is the most visited region in the world and received around 37% of the global tourist arrivals in 2019 (European Court of Auditors 2021).

Indeed, as shown in the figure below, capital investment in travel and tourism has been increasing—despite a slight decrease in 2016—during the most recent years included in this study. This increase may explain why it has become statistically significant in our analysis (Fig. 2).

Fig. 2

Tourism capital investment in Travel and Tourism in the European Union between 2006 and 2019. Notes: Own elaboration using World Travel & Tourism Council (WTTC) data. Data is in billion euros (real prices)

Moreover, addressing the second research question, tourism capital investment also seems to affect poverty negatively in this region. These outcomes suggest that tourism capital investment contributes to a more equitable income distribution in EU nations, potentially through its role in alleviating poverty.

In the European Union, tourism expansion is considered an essential strategy for boosting both competitiveness and regional development, with this progress being heavily reliant on substantial investments within the sector (Setoodegan et al. 2022; Paramati et al. 2018). While these investments contribute to the development of destinations, they also have a significant socio-economic impact, contributing to the well-being of EU citizens by supporting employment and enhancing consumer experiences (European Court of Auditors 2021). In this context, the positive effects of tourism investments on income distribution and poverty reduction may be closely aligned with broader efforts undertaken by the European Union to promote equitable growth. More precisely, the programs launched by the European Union provided public funding for investments in the tourism sector.

For instance, EU initiatives that strengthen tourism investments and position them as key tools for poverty reduction and promote income equality include the European Social Fund (ESF) and the Employment and Social Innovation (EaSI). The ESF serves as a main instrument for enhancing employment quality within the tourism industry by training workers—i.e., assisting companies dealing with restructuring or a scarcity of skilled labor and offering skill development programs for individuals from vulnerable backgrounds (European Commission 2016). Hence, this program can reduce poverty and promote income equality across different societal groups.

The EaSI program also promotes higher-quality jobs in tourism-related activities to overcome social exclusion and poverty across Member States (European Commission 2016). It includes initiatives such as the Programme for Employment and Social Solidarity (PROGRESS), the European Job Mobility (EURES), and the EaSI Guarantee Financial Instrument. PROGRESS supports research and experimentation in social policies by testing innovative approaches on a small scale and expanding successful ones. EURES assists workers in finding employment in other EU countries and helps companies recruit internationally through mobility schemes. It also covers part of the costs for Small and Medium Enterprises (SMEs)—representing most of the EU tourism business—to train newly hired workers and support their integration into new roles. Lastly, the EaSI Guarantee Financial Instrument supports establishing and developing small businesses or social enterprises, covering investment, working capital, leasing, and start-up costs such as licenses and other expenses.

In addition, the Cohesion Fund (CF) was developed to reduce economic and social inequalities between and within Member States and supports investment in environmental and trans-European transport networks (European Court of Auditors 2021). This program targets Member States whose Gross National Income per capita is less than 90% of the EU average (European Court of Auditors 2021). Transportation infrastructure is crucial for driving tourism by providing easier access to existing destinations, increasing tourist arrivals, and significantly boosting local economic activities (Kanwal et al. 2020). Additionally, constructing or maintaining such infrastructure can lead to the development of new tourist destinations (Virkar and Mallya 2018). This aspect can be particularly important for less developed and rural areas of the EU, as it helps improve local living standards by generating employment, preserving essential services, increasing incomes, slowing depopulation, and encouraging generational renewal—ultimately reducing poverty (Šajn and Finer 2023; Nazneen et al. 2021).

Therefore, CF investments in transportation infrastructure can be classified as tourism-related infrastructure investments.¹ While these investments may not always be specifically designed for tourism, they benefit this industry and are included in the sector's total gross fixed capital formation (see Sect. 3). As a result, this program supports the development of the sector, particularly in less developed areas, which in turn generate both "income" and "tax revenue" effects—helping to alleviate poverty and thereby reducing economic disparities between urban and rural regions (one of the main challenges within the EU).

The effects of international tourist arrivals contribute to reducing income inequality across the EU region, aligning with the findings of Nguyen et al. (2020), who found that tourist arrivals decrease economic disparities in highly developed countries. Furthermore, once again, the role of tourism in reducing economic disparities can also be attributed to its capacity to benefit the deprived since international tourist arrivals also have a negative impact on poverty. These negative effects are probably linked to the potential of tourism investments to expand capacity at a destination. More precisely, continuous investment in constructing new tourism-related infrastructures or maintaining existing facilities is extremely important to meet the growing demand from higher international tourist arrivals (Paramati et al. 2018). For instance, destinations can better manage their carrying capacity by constructing new accommodations, expanding tourist facilities, and increasing airport capacity, thus preventing over-tourism and mitigating upward pressure on prices (Paramati et al. 2018; WTTC 2015). Therefore, when well-managed, these inflows stimulate local economic activity by generating revenue and promoting economic growth, ultimately benefiting residents and low-income individuals (Subramaniam et al. 2022). Consequently, this generates both "income" and "tax revenue" effects, offering pathways out of poverty, mainly for disadvantaged groups (such as low-income individuals or women), and thereby decreasing economic inequalities.

Tourism and travel’s direct contribution to employment contributes to decreasing income inequality across the European Union, and it is also possibly linked to its negative influence on poverty. These outcomes are supported by Incera and Fernández (2015), who noted that tourism could be a key solution for income inequality and poverty issues due to its ability to create numerous job opportunities, particularly for workers with lower skill levels. This effect of the direct contribution of tourism to employment on both income equality and poverty reduction can be attributed to investments within the sector, which have proven to be the main catalysts for job creation (Nguyen et al. 2024). Specifically, increasing investments in the tourism sector accelerate industry growth by generating higher expected returns and fueling greater tourism demand, which is stimulated by rising income levels (Nguyen et al. 2024). As demand increases, more jobs are created (Li et al. 2018), addressing labor needs while reducing disparities. In summary, higher investment in the tourism sector drives employment generation and overall economic activity, increasing household incomes ("income effect") and boosting government revenues ("tax revenue effects").

Moreover, tourism investments through the previously mentioned initiatives (ESF and EaSI) are probably amplifying the benefits of tourism's direct contribution to employment in terms of economic equality and poverty alleviation. These programs not only support the creation of jobs but also improve the quality of employment by ensuring better working conditions and fair wages for workers in the industry.

Regarding the influence of the HDI, the outcomes suggest that it is a driver in reducing income inequality and poverty, corroborating previous research findings (Widiastuti et al. 2022; Theyson and Heller 2015). Economic expansion is associated with higher incomes, increased public budget revenue, and expanded public spending on health and education, all of which foster human development (Simionescu et al. 2024). Thus, the increase in economic output leads to higher levels of human development, which ensures that individuals have the necessary resources and opportunities to prosper, thereby enhancing the quality of human capital and consequently addressing both income inequality and poverty (Simionescu et al. 2024; Widiastuti et al. 2022). Over the past two decades, the European Union’s economy has grown by approximately 27% (European Commission 2024a). Moreover, public expenditure on education has proven to play a crucial role in alleviating poverty among adults in Europe (Hidalgo-Hidalgo and Iturbe-Ormaetxe 2018). These facts support the previous argument, suggesting that economic growth has probably led to improvements in HDI, which, in turn, has enhanced human capital quality and can explain its positive impact on reducing poverty and, consequently, income inequality across the EU.

The impacts of the age dependence ratio suggest that it increases income inequality and poverty, which aligns with the findings of the literature (Belu et al. 2024; Fang et al. 2021). As was previously noted, this indicator is connected to the income distribution of families and social expenditure (Fang et al. 2021; Mao 2016). Particularly in the European Union, an increase in life expectancy has led to a decline in the working-age population (15–64 years) as the population ages (European Commission 2024b). This reduction in the working-age population exerts pressure on the capacity to sustain pensions, potentially leading to higher poverty rates among older adults (Belu et al. 2024). Moreover, in 2019, the age dependency ratio was registered at 54.9%, which means there were approximately two working-age individuals for each dependent youth or elderly citizen in the EU (Kiss 2021), underscoring the economic pressure on the labor force. Considering that a larger proportion of dependent citizens is commonly associated with increased income inequality (Dorn et al. 2022), this trend provides further insight into the positive impacts previously observed.

Concerning the effects of trade globalization on income inequality and poverty, the results indicate that greater trade globalization is associated with reducing income inequality and poverty in the EU. These findings are consistent with previous investigations on developed economies (see Han et al. 2023; Signoret et al. 2020). Generally, countries with stronger global economic linkages tend to have more equitable employment opportunities, higher labor productivity, and improved wage distribution (Chi 2023; Çelik and Başdas 2010), contributing to poverty alleviation and helping to overcome income disparities. In the context of the EU region, the second-largest exporter of products globally, exports supported about 38 million jobs in 2019 (European Parliament 2019). In addition, economic globalization benefits consumers primarily through lower import tariffs and increased competition, which decreases the prices of goods and services, thus enhancing purchasing power and raising living standards. For example 2017, these price reductions saved EU consumers approximately €600 per person (European Parliament 2019). These are possible reasons for trade globalization’s positive influence on economic equality and poverty alleviation across the EU.

Finally, the financial development index positively affects income inequality. Still, it lacks statistical significance on poverty, aligning with the outcomes of Qehaja-Keka et al. (2023) and de Haan et al. (2022). While a well-developed financial system offers numerous advantages—such as turning savings into productive investments, supporting trade, enhancing borrower monitoring to improve efficiency, and spreading risk (de Haan et al. 2022), these benefits often disproportionately favor wealthier individuals, who are more integrated into formal financial systems (Greenwood and Jovanovic 1990). This unequal distribution of benefits to more affluent individuals could explain the positive impact of financial development on the Gini coefficient. Furthermore, a lower-income fraction of the population, who are typically excluded from the formal financial system, lack access to such advantages and instead depend on informal household linkages for capital (Claessens and Perotti 2007; Rajan and Zingales 2003; Greenwood and Jovanovic 1990). This exclusion probably leads to the inability of financial development to influence the poverty headcount ratio.

6 Conclusion, policy implications, and further research

This article evaluates the effects of tourism capital investment on income inequality and poverty, centered on a panel of 24 EU nations, using data from 2006 to 2019. In order to achieve these goals, two models were estimated: (i) Model I with the Gini coefficient as the dependent variable and (ii) Model II with the share of people below 40% of median income as the dependent variable. Given that $N<T$ and that cross-sectional dependence, heteroscedasticity, first-order autocorrelation, and contemporaneous correlation (in Model II and IV) are present, the PCSE estimator is the most suitable.

Focusing on the results of our main analysis, tourism capital investment, international tourist arrivals, and the human development index appear to contribute to reducing income inequality and poverty in European Union countries. Contrariwise, the age dependency ratio is a highly significant barrier to achieving income equality and poverty alleviation across these nations. Moreover, trade globalization appears to have a negative effect on income inequality and poverty levels across the EU.

The outcomes of the robustness check confirm that the indicators related to tourism investments (tourism capital investment and tourism direct contribution to employment) negatively influence income inequality and poverty in the EU region. Additionally, the findings corroborate the negative effects of HDI on income inequality and poverty and the positive impact of the age dependency ratio on both factors. Lastly, the empirical estimations reveal that the financial development index exacerbates income inequality while lacking statistical significance on EU poverty.

Based on the main findings of this research, investments in the tourism sector can be a valuable tool for achieving Goals 1 and 10 of the 2030 United Nations SDGs. However, the line between tourism's positive and negative impacts is thin. For this reason, while such investments should continue growing, governments must carefully manage them to maximize their benefits. Thus, to maximize the efficiency of tourism investments—considering that tourism activities are currently fragmented across sectors such as transportation and accommodation, each managed by individual EU departments—policymakers should establish a coordinating entity within existing tourism-related structures, specifically tasked with ensuring that sector concerns are consistently integrated into policy development and regulatory efforts.

Furthermore, EU governments must keep increasing investments in the tourism sector, particularly for SMEs, which represent the majority of tourism businesses in this region but often have a weak capital base. Creating a funding mechanism or low-interest loan program for them can stimulate investment in small-scale, locally owned businesses, which may promote entrepreneurship, create jobs, and broadly distribute economic benefits, addressing poverty and income inequality issues.

At the same time, policymakers should continue to promote financing for social protection schemes (e.g., unemployment benefits, pensions) specifically designed for workers in the tourism sector, with a focus on areas with high poverty rates. These schemes can provide workers with a safety net during economic instability, reducing their risk of falling deeper into poverty. Consequently, as more workers achieve financial security, income inequality may decrease as the economic benefits are spread more evenly across the population.

Governments must also establish Tourism Development Zones within the EU. These are designated areas, primarily less economically developed or remote regions, where incentives (such as streamlined regulatory processes or tax breaks) encourage businesses to invest in tourism. These zones would help distribute the economic benefits of tourism more equitably across the population, ensuring that both urban and rural areas, particularly those that have historically received less investment, can share in the resulting growth and job creation.

Notwithstanding its contributions, this investigation also presents some limitations. Therefore, for future research, it could be interesting to extend the analysis to include middle- and low-income countries, thereby enhancing the applicability of the findings while offering valuable insights into the diverse global tourism investment landscape. Moreover, this work uses data from 2006 to 2019. Hence, expanding the period to cover COVID-19 and recent events such as the Russia-Ukraine war could shed light on how economic shocks influence the effect of tourism capital investments on income inequality and poverty. However, this is not possible yet due to the lack of data.