ارتباط های متغیر با زمان بین رسیدهای گردشگری و رشد اقتصادی در یک اقتصاد کوچک باز

The causal link between tourism receipts and GDP has recently become the major focus of some recent studies in tourism economics. Results obtained in these studies about the causal link appear to be sensitive with respect to the countries analyzed, sample period and methodology employed. Considering the sensitivity of the causal link, we use the rolling window and time-varying coefficients estimation methods to analyze the Granger causality based on Vector Error Correction Model (VECM). When applied to Turkey for the 1963–2006 periods, this methodology enables us to overcome differences in the outcome of the tests performed in other studies for tourism receipts and GDP. The findings of this paper are as follows: results from the full sample within the VECM model indicate that there is no Granger causality between the series, while the findings from the time-varying coefficients model based on the state-space model and rolling window technique show that GDP has no predictive power for tourism receipts; however, tourism receipts have a positive–predictive content for GDP following early 1980s.

There is no doubt that the export-based development strategy is of great importance and the main source of economic growth and development. A brief look at the world economy, especially the post-1960s, shows the preference of export and export-based growth over import based growth. The causality relationship between export and economic growth have for a long time been the focus of many researchers (Balassa, 1988 and Ghatak et al., 1997). While the evidence obtained in these studies differ, the general consensus is that there is a positive cause and effect relationship between export and economic growth; that is, the more the exports, the higher the economic growth (Shan and Sun, 1998). Also known as the smoke-stack industry, tourism and its economic impacts have for a long time been the focus of many researchers in both areas. Taking the relation between export and economic growth into account, it will not be inaccurate to hold that the premise that tourism will lead to economic growth stems from this relationship (Vanegas and Croes, 2003). Tourism, chiefly a labour-intensive sector, is in the section of international services under current accounts of the balance of payments. For this reason, tourism receipts can be said to have an export effect since the nature of tourism receipts are of foreign exchange nature. In other words, tourism is regarded as an intangible export item (Theobald, 2001). The demand for goods and services in the country visited is commensurate with the increase in the number of tourist arrivals. If the country visited has the resources to meet the increasing demand as a result of the number of tourists visiting the country, the spending of the tourists will remain in the country visited. Foreign exchange surpluses created by tourism activities will have a positive contribution to the balance of payments, which is why it is widely acknowledged that tourism in the long run might create economic growth, as is the case in the hypothesis of export-focus growth (Balaguer and Jorda, 2002 and Croes, 2006). The growth in the tourism sector has a positive impact on the current account, creates employment and induces an increase in GDP, thus having a desired effect on the economy concerned (Brohman, 1996). The related literature hosts some research into the issue with different conclusions. For example it was argued that a negative impact on growth and loss in economic welfare is very likely to ensue under a monopolistic administration of tourism operations (Hazari and Ng, 1993). Another study was carried out by Hazari and Ng (1993) in order to investigate the relation among tourism, capital accumulation, consumption per person and terms of trade. The evidence obtained suggests that tourism has a positive impact on the long-term growth of countries (Hazari and Sgro, 1995). In another study, Modeste (1995) holds that tourism contributes to the economic growth and the growth in tourism sector brings forth shrinkage in the agriculture sector. Balaguer and Jorda (2002) in their paper investigate the direction of relationship between tourism and economic growth, using error correction model and found that the causality goes from tourism to growth in the long run. Moreover, Dritsakis (2004) for Greece and Durbarry (2004) for Mauritius find bidirectional causality between tourism development and economic growth using error correction model. Kim et al. (2006), using Granger causality test, find out that in Taiwan for the period of 1971–2003 the causality between tourism development and economic growth is bidirectional. On the other hand, using Granger causality test, Oh (2005), unlike the previous researchers mentioned in this paper, find a relation from only economic growth to tourism development for Korea. Lee and Chang (2008) investigate the casual relation between tourism development and economic growth for OECD and non-OECD countries (including those in Asia, Latin America and Sub-Sahara Africa) for the 1990–2002 period. Evidence obtained in their study indicates that there is unidirectional causality relationship from tourism development to economic growth in OECD countries and bidirectional relationship in non-OECD countries, but only a weak relationship in Asia. As for Turkey, while Gunduz and Hatemi-J (2005) find a unidirectional causality from tourism to economic growth using leveraged bootstrap causality tests for the period 1963–2002, Ongan and Demiroz (2005) suggest bidirectional causality between international tourism and economic growth in Turkey for the period of 1980Q1–2004Q2 and using the Granger causality tests. In a similar manner, Bahar (2006) obtains in his study that tourism has a positive effect on economic growth, and the cointegration test has proved that there is a bidirectional relationship between tourism and economic growth. On the other hand, unlike the previous researchers, Katircioglu (2009) investigates the tourism-led-growth (TLG) hypothesis in the case of Turkey by employing the bounds test and Johansen approach for cointegration using annual data from 1960 to 2006 and rejects the tourism-led growth hypothesis for the Turkish economy on the grounds that no cointegration was found and error correction mechanisms plus causality tests cannot be run for further steps in the long term. Thoma, 1994 and Swanson, 1998 examine the sensitivity of results from causality test using recursive and rolling window techniques to analyse Granger causality between the time series. Their findings indicate that changes over the sample period might yield some substantial influences over the causal relationships considered and for this reason, taking this into account; the links between tourism receipts and economic growth might follow the same patterns of sensitivity as put by Thoma, 1994 and Swanson, 1998. Since it might be quite likely that causality relationships over the sample period (in this case of tourism receipts and economic growth) changed over time, the rolling window techniques were implemented in order to motivate the time-varying causality patterns. Building on the evidence from the rolling window estimation, the study uses time-varying coefficient (TVC) model to estimate the varying links between the tourism receipts and GDP. The TVC estimation confronts the unknown functional form problem, specification errors, and spurious relationships (Hall et al., 2009 and Hall et al., 2010). The main motivation of this type of method used in this paper is that non-constancy of causality might create some problems in econometric terms in the application of standard Granger causality tests as well as the concerns for economic theory and policy analysis. Time dependency of causality patterns could be put down to many reasons such as the changes in monetary policy and large shocks to the economy. In order to rise above these problems, we use Vector Error Correction Model (VECM) with time-varying parameters. We also determine the sign impact between the series over the sample period. What is new in this paper is that it could be the first to examine the relationship between tourism receipts and economic growth (real GDP) using autoregressive (VECM) models with time-varying parameters in tourism literature. This paper analyses the time-varying linkage between the real tourism receipts and real GDP series using the annual data from 1963 to 2006 for Turkey. The findings obtained indicate the following results: results from the full sample within the VECM model show that there is no Granger causality between the series. On the other hand, the findings obtained from the state-space and rolling window techniques show that GDP has no predictive power for tourism receipts, while tourism receipts have a positive–predictive content for GDP following the early 1980s. The paper is organized as follows: Section 2 explains the methodology employed and Section 3 provides empirical results and the last section presents the conclusion.

This paper investigates the time-varying causal links between the real tourism receipts and real GDP for Turkey by using state-space time-varying coefficients and rolling window methods for data for covering 1963–2006 period. The methodology employed in this paper allows us to overcome the econometric concerns that could appear in the application of Granger causality test. Changes in economic policies worldwide as well as political and social environment could influence the causality patterns in the data under consideration. For this reason, we prefer to use time-varying parameters VECM model. The empirical analyses yielded the following conclusions: firstly, the result obtained from the full sample show that there is no Granger causality between tourism receipts and GDP series. On the other hand, results obtained from the state-space time-varying coefficients and rolling window estimation methods to analyze the Granger causality based on VECM with time-varying parameters indicate that GDP does not have predictive power for tourism receipts, and that the tourism receipt positively Granger causes GDP after early the 1983s. In terms of the findings obtained from the analyses, it could be held that tourism receipts have a positive impact on the economic growth in Turkey. The result from the state-space time-varying coefficients and rolling window methods draw analogy with the results of Gunduz and Hatemi-J (2005) and Ongan and Demiroz (2005). Capital account liberalization and full convertibility allowed for capital inflows to Turkey, which has an impact on the finding that there is a casual relation from tourism receipts to GDP following the 1983 period. In this regard, it will not be wrong to argue that such dynamics as mentioned above have a positive impact in small open-economies such as the one Turkey has. Within this frame, it is projected by WTO that tourism will experience great developments and strides and tourism incomes worldwide might reach up to 2 trillion dollars. What's more, in parallel to the increasing welfare in particular in the developed countries, it is an undeniable fact that tourism sector will enlarge in space with the income allocated for travel and the increasing means of transport. For this reason, it is necessary that governments and actors in the tourism sector should prioritize the arrangements to further the sector and the policies and regulations to realize the potential projected to occur in the international tourism arena should be put into practice in a swift manner. Finally, this finding is important for policy makers as well as academicians in the field and shows that this issue still deserves further attention from researchers for comparison purposes, even for Turkey.