Temporal dynamics of air passenger flow and economic performance: insight from turkish airport hubs

Abstract

The aviation sector has grown rapidly in recent years. While the aviation sector used to generate revenue from passengers and cargo, it has now begun to generate revenue from non-aviation commercial sectors as well. Examples of these commercial revenues include duty-free shops, food and beverage revenues, lounges, and airport hotels, among many other segments. Consequently, airports have begun to contribute to the country's economic growth. The literature on the aviation sector has focused primarily on gross domestic product (GDP) and passenger numbers, examining the relationship between these two. Most studies have found a positive relationship between GDP and passenger numbers, concluding that increases in one influence the other. Furthermore, when investigating the causes of demand in aviation, variables other than passenger numbers have been used. For example, the relationships between population, distance to alternative airports, tourism, migration, and exchanges have been examined. However, limited research has been conducted on airport revenues, particularly non aviation revenues. For example, factors affecting product purchases and food and beverage consumption at airport facilities have been analyzed. Furthermore, tests have been conducted to reveal the relationship between the terminal's built environment and retail revenue. This study examines the relationships between commercial revenue segments at Atatürk Airport (AHL) and Esenboğa Airport (ESB), two major airports in Türkiye. The scope of the study includes commercial revenue data from restaurant&bar, cafe&bakery, buffet, fast food, vending machine, staff cafeteria, lounge, and airport hotel. The analysis focuses on monthly data covering the period 2015-2019This thesis aims to fill a gap in the literature by investigating the relationship between non aviation commercial revenues. Prior to the analysis, revenue figures were transformed using the natural logarithm for conform to a normal distribution. The Augmented Dickey-Fuller (ADF) unit root test was applied to determine the stationarity of the time series. A Vector Autoregressive (VAR) model was used to analyze the dynamic interactions between revenue segments. The ADF test revealed that only buffet and fast food revenues for AHL did not have a unit root. The first difference was calculated and found that restaurant&bar, vending machine, and hotel revenues were stationary. The second difference determined that cafe & bakery, and lounge revenues were still non-stationary. To ensure a reliable data set, lounge revenues and cafe & bakery revenues were excluded. The same test was performed for the ESB, and no unit root was found for cafe & bakery revenues. The first difference determined that staff cafeteria, vending machine, and restaurant & bar revenues were stationary. Granger causality tests were conducted to determine the relationships between different revenue segments. The Granger causality test results indicate strong and directional relationships between commercial revenue segments at Atatürk Airport. In particular significant unidirectional causality was found between hotel, restaurant & bar, buffet, fast food, and vending machine revenues. For example, while past hotel revenues strongly predict future fast food and buffet revenues. Similarly, restaurant & bar revenues predict future buffet, vending, and hotel revenues. Fast food revenues also strongly predict future staff cafeteria revenues. These findings indicate that some revenue segments in the Atatürk Airport (AHL) have a significant driving influence on other segments. Mutual causal relationships are also evident in the AHL. A bidirectional relationship was found, particularly between hotel and vending machine revenues. Similarly, a strong correlation was found between fast food revenue and restaurant & bar revenues. These results suggest that some segments in the AHL influence each other and that their revenue dynamics are more integrated. However, the Granger causality test results indicate that there is no statistically significant causal relationship between personnel revenues and the restaurant, vending, buffet, and hotel segments. The acceptance of the null hypothesis in both directions suggests that these revenue streams do not Granger-cause each other. The results for the ESB were different from those for the AHL, but weaker relationships were found. In the ESB, only past café revenues predicted future staff cafeteria revenues, and a significant relationship was found between staff cafeteria revenues and vending machine revenues. However, none of these relationships were bidirectional. Accordingly, consumption in the ESB was found to be more independent and differentiated from one another. Following this test, an Impulse-Response test was applied to the revenue segments. Strong impulse-response patterns were found for the AHL. Fast food and restaurant revenues were negatively affected in the short term by shocks to buffet revenues. When examining the impact of sudden shocks in the fast food segment on other revenue segments, it was found that the buffet segment was negatively affected in the short and medium term. However, no direct relationship was found with hotel and vending machine revenues. Sudden shocks in restaurant&bar revenues positively affect buffet and fast food revenues in the long term, while vending machine revenues are negatively affected in the short term. The effect on hotel and staff cafeteria revenues is largely neutral. Sudden shocks in hotel revenues have a slight negative impact on the buffet and restaurant & bar segments. Shocks in vending machine revenues have been observed to affect hotel and restaurant&bar revenues with small fluctuations. However, these fluctuations are generally close to zero. While they negatively affect buffet revenues in the short and medium term, the long-term impact of vending machines remains insignificant. It has been concluded that vending machines operate independently of other segments. Sudden shocks in staff cafeteria revenues have a near-zero impact on other segments in the long term. Impulse response tests were also conducted for the ESB. Sudden shocks in the cafe revenue segment negatively affect vending machine revenues in the short term but do not significantly affect other revenues in the long term. Sudden changes in restaurant& bar revenues cause small fluctuations in other revenue segments. These effects approach zero in the short term, and no lasting impact is observed in the long term. Shocks in staff meal revenues appear to have a slight negative impact on other revenue types. It initially had a negative impact, but it quickly stabilized. Changes in staff cafeteria revenues appear to have no significant impact on airport commercial sales. Based on the results, Atatürk Airport (AHL) appears to have a more integrated and cohesive revenue structure. Therefore, an integrated commercial area management system could be adopted for Atatürk Airport. This system ensures that commercial areas are planned holistically, rather than in isolation. As a strategy, they should be managed in a coordinated manner under a single management center. A concept-based management system could be developed for Esenboğa Airport (ESB). The results at ESB indicate that commercial areas work independently and in isolation. This suggests that each commercial revenue segment has different passenger bases. Therefore, each revenue segment should be planned as an independent market.