Tourism demand and economic growth in Spain: New insights based on the yield curve

Highlights

• We investigate the dynamic correlation between tourism demand growth and expected macroeconomic conditions of Spain.

• We find that the tourism-expected economic growth relationship is time varying and volatile in sign and magnitude.

• We find that positive correlations are also observed when there is a regime shift in the Spanish economy.

• Key geopolitical and economic events affect the said relationship.

• We do not report notable differences across the different origin countries.

Abstract

This study examines for the first time the dynamic relationship between tourism growth and expected macroeconomic conditions of the destination country using a DCC-GARCH model. The focus is on the Spanish economy in which monthly tourist arrivals data from 1998 to 2017 were collected for five key origin countries and around the world. To capture expected macroeconomic conditions, the Spanish term structure of interest rates is used. The results suggest that the tourism-expected economic growth relationship is time varying without any country-specific differences in the behaviour of the correlations. Importantly, positive correlations reportedly coincides with a regime shift in the Spanish economy; whereas negative correlations are evident when expected economic conditions are stable. It is also shown that the aforementioned relationship is influenced by key geopolitical and economic events (the 9/11 terrorist attacks, Global Financial Crisis and the ECB's quantitative easing programme). Finally, policy implications derived from the main findings are discussed.

Introduction

This study aims to shift the focus of the tourism economics literature to the dynamic relationship between tourism growth and expected macroeconomic conditions. The study focuses on Spain given that it is among the top ten (10) destination countries globally in terms of the tourism contribution to its economy (WTTC, 2017). Tourism contributes (both directly and indirectly) in excess of 14% of Spanish GDP, providing 2.6 m jobs directly and through related industries, which represents 14% of the total workforce (WTTC, 2017b). These figures place tourism as the second most important sector in the Spanish economy, only behind the retail industry (WTTC, 2017c).

The importance of the tourism industry on world destinations is well documented in the relevant literature (see, for instance, Cárdenas-García, Sánchez-Rivero, & Pulido-Fernández, 2015; Dogru & Sirakaya-Turk, 2017; Dogru & Bulut, 2018). It is argued that the importance of the tourism sector to the wider economy stems from the fact the former provides both direct and indirect effects to the latter, in terms of income, employment and infrastructure, among others. Such importance is more prevalent in the European context given the economic effects of the global financial crisis on tourism (Song, Dwyer, Li, & Cao, 2012) and the fact that several EU member-countries are among the top tourism destinations in the world (UNWTO, 2017).

Despite ample evidence on the impact of tourism on economic growth (see, for instance, Lee, Moon, & Mjelde, 2010; Li, Blake, & Cooper, 2011), such a relationship is by no means conclusive. It is acknowledged that the tourism-led economic growth hypothesis is indeed among the most widely accepted hypothesis in the tourism economics literature (some recent studies include Sugiyarto, Blake, & Sinclair, 2003; Parrilla, Font, & Nadal, 2007; Ivanov & Webster, 2013; Dogru & Bulut, 2018). Nevertheless, there is evidence that the conservation hypothesis also holds which maintains that economic conditions are conducive to tourism income generation (see Antonakakis, Dragouni, Eeckels, & Filis, 2017; Aslan, 2014). More recently, authors opine that the feedback hypothesis is able to explain the relationship between tourism income and economic growth, suggesting that there is a strong interdependency among the two (see, Chen & Chiou-Wei, 2009; Perles-Ribes, Ramón-Rodríguez, Rubia, & Moreno-Izquierdo, 2017; Antonakakis et al., 2017, among others). Finally, the neutrality hypothesis, which posits that tourism and economic growth are actually independent, also finds support in some studies (see, for instance, Katircioglu, 2009; Tang & Jang, 2009; Tugcu, 2014).

The aforementioned causal relationships have been largely examined through a variety of econometric techniques, including Autoregressive Distributed Lag (ARDL), Vector Error Correction Model (VECM) and Vector AutoRegressive (VAR) models and Granger Causality tests. The variables used in these studies primarily involve tourist arrivals or tourism income (as a proxy for tourism growth) and GDP growth.

Given the extant literature surrounding this research area, it is beyond the scope of this paper to provide a detailed account of existing studies. Rather, the aim is to highlight the three main innovations of this paper. First, this study draws attention to the research on the potential interrelationship between tourism growth and expected (rather than current) macroeconomic (performance) prospects of the destination country. Thus, unlike previous studies that focus their interest on GDP (i.e. current macroeconomic performance) when examining the link between tourism and economic growth, this paper investigates the interdependency between tourism and a key economic leading indicator; namely the yield curve spread or term structure of interest rates, as a proxy for expected macroeconomic prospects. It is noteworthy to mention here that this study refers to the “term structure of interest rates”, “yield curve” and “spread” interchangeably throughout the paper.

It is maintained here that the use of leading indicator is capable of revealing important new insights on the link between tourism and economic growth based on two premises. On the one hand, if one anticipates that tourism demand will yield positive effects for a destination economy, then these prospects should be reflected first in yield curve spreads prior to their appearance in the real economy. On the other hand, the paper opines that the tourism sector primarily responds to the anticipated, rather than current, economic conditions. The economic literature has convincingly shown that the yield curve spread is capable of successfully predicting output growth, and thus act as the most desired leading indicator (see, inter alia, Estrella & Mishkin, 1998; Hamilton & Kim, 2002; Rudebusch & Williams, 2009; Christiansen, 2013). From a theoretical standpoint, the usefulness of the term structure of interest rates (or yield curve spread) as a leading economic indicator can be explained by expectations theory,¹ the liquidity premium theory² or the theory of intertemporal consumption,³ among others (see Wheelock and Wohar (2009) for an overview of these theories).

Second, the bulk of the previous studies have used static economic frameworks, which do not allow for the potential dynamic character of the aforementioned relationship. Only a handful of studies have recently concentrated their attention on the time-varying relationship between tourism and economic growth, using frameworks such as the Diebold and Yilmaz spillover index, multivariate GARCH models and rolling-window Granger causality (Antonakakis, Dragouni, & Filis, 2015; Dragouni, Filis, Gavriilidis, & Santamaria, 2016; Lean & Tang, 2010; Tang & Tan, 2013). The study contributes to this limited number of studies by employing the Dynamic Conditional Correlation (DCC) model of Engle (2002) to assess the time-varying relationship between tourism growth and the term structure of interest rates as a proxy for expected economic conditions.

Finally, in the spirit of de Oliveira Santos (2009), Gounopoulos, Petmezas, and Santamaria (2012) and Chatziantoniou, Degiannakis, Eeckels, and Filis (2016), the present study considers both aggregated and disaggregated tourism demand data to accommodate for any origin-specific effects. It also investigates the effects of the 9/11 terrorist attacks in the US and Global Financial Crisis (GFC) on tourism demand, the term structure of interest rates and the said relationship. Moreover, the Quantitative Easing (QE) programme by the European Central Bank (ECB) on Spanish yield curve spreads is also considered, so to establish whether it has an impact on this relationship.

The results suggest that the tourism-economic growth relationship, based on expected macroeconomic conditions, is time varying and volatile both in sign and magnitude. Furthermore, the time-varying correlations do not reveal any notable differences among the origin countries of the Spanish tourism, with the only exception being Germany's tourist arrivals that exhibit a constant negative relationship with the Spanish term structure of interest rates. More importantly, with the exception of Germany, the evidence suggests that positive correlations arise when there is a regime shift in the Spanish economy (either entering into a recession or boom phase). This is suggestive of the fact that the regime change seems to influence the behaviour of tourist arrivals. By contrast, negative correlations tend to prevail during periods that the Spanish economy is at a more permanent state (either in a recession or in economic growth). Finally, the 9/11 attack has a significant impact on the said relationship that is country-specific in terms of signs and magnitude, a feature that is repeated for the recent GFC and the QE programme.

The rest of the paper is structured as follows. Section 2 presents the econometric framework and Section 3 describes the data of the study. Section 4 analyses the empirical findings, before Section 5 to conclude the study.