- Email: [email protected]
The commercial performance of global airports
Transport Policy xxx (2017) 1–9
Contents lists available at ScienceDirect
Transport Policy journal homepage: www.elsevier.com/locate/tranpol
The commercial performance of global airports Franz Fuerst a, b, *, Sven Gross c a
University of Cambridge, Dept of Land Economy, 16-21 Silver St, Cambridge CB3 9EP, United Kingdom University of Melbourne, THRIVE Research Hub, Parkville, VIC 3010, Australia c Harz University of Applied Sciences, Friedrichstraße 57-59, 38855 Wernigerode, Germany b
A R T I C L E I N F O
A B S T R A C T
Keywords: Commercial revenues Non-aeronautical airport income Airport finances
Revenues from non-aeronautical business have received increasing attention from airports seeking to enhance their profitability. This study analyzes the commercial performance of global airports using a panel dataset of 75 airports in 30 countries. Applying pooled OLS, random effects and 3SLS estimation frameworks, we identify the main drivers of financial performance. Our results are in line with the existing literature. The share of international passengers, the size of the commercial area as well as airport size and the mix and intra-terminal location of retail space are found to be significant determinants. The latter finding suggests the presence of economies of scale in generating higher commercial revenue and ensuring profitability for an airport's non-aeronautical operations. Staterun and partially privatized airports appear to retain a significantly smaller share of concession sales than their privatized counterparts. Regarding the retail mix, a higher share of food and beverage outlets appears to increase the revenue share retained by the airport while a higher percentage of outlets located airside rather than landside depresses commercial revenues.
1. Introduction Airports operate in a highly cyclical business environment. The demand for moving people and cargo on airplanes is typically more volatile than equivalents for other industries or even within the transport sector. One possible strategy to reduce exposure to these demand shocks is for airports to expand their non-aeronautical activities and diversify their activities away from full reliance on the aeronautical sector. Another strategy is the installation of activities without a physical connection to the terminal, such as eCommerce projects, for example QR code walls and retail-specific apps at airports owned by the Fraport group (Fraport, 2013, 2015). The main argument in favor of the first strategy is that the nonaeronautical segment encompasses a more diverse range of customer groups. Apart from passengers, airport employees and visitors that are all closely connected to the aeronautical operations of an airport, the nonaeronautical commercial and retail facilities are additionally also frequented by local residents, employees and visitors of companies locating near the airport and other types of customers that are not airline passengers. Increasingly, airports also include convention centers and other entertainment, business and leisure facilities which are largely independent of fluctuations in air passenger volumes. Furthermore, the
expansion of the non-aeronautical business is also attractive because of its lower operational cost structure compared to the aeronautical business which may make it more profitable. In some cases, non-aeronautical revenues even account for the largest share of the profits of an airport operator. Increasingly, major real estate companies cover airport real estate in their analysis and investments as a distinct property sub-type and credit rating agencies take the diversification of revenue sources into consideration in their financial ratings of airports (Deloitte, 2009). The importance of airport commercial activities and the airport retail sector in particular has received considerable attention over the last decade, particularly as liberalization and deregulation of the aviation industry has forced airports worldwide to find new sources of income to replace the aviation revenues loss because of the increasing market power of airlines such as Ryanair, EasyJet or other Low Cost Carriers or also to third party ground service operators. Despite the obvious importance of commercial activities in light of these recent developments, research on the commercial performance of airports and its determinants is still scarce. This paper presents the first comprehensive academic study of global commercial airport performance. There exists a growing body of empirical studies of the non-aviation business, particularly on individual business activities ranging from parking space management and parking revenue (e.g. Qin and Olaru, 2013) to
* Corresponding author. University of Cambridge, Dept of Land Economy, 16-21 Silver St, Cambridge, CB3 9EP, United Kingdom. E-mail address: [email protected] (F. Fuerst). http://dx.doi.org/10.1016/j.tranpol.2017.08.005 Received 2 August 2016; Received in revised form 29 July 2017; Accepted 15 August 2017 Available online xxxx 0967-070X/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Fuerst, F., Gross, S., The commercial performance of global airports, Transport Policy (2017), http://dx.doi.org/ 10.1016/j.tranpol.2017.08.005
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
both difficult to achieve and difficult to measure in a retail context. This limitation is particularly true for the present analysis which does not include measures of production costs due to unavailability of suitable data and instead proxies size effects with available measures such as size of the commercial area and total number of passengers of an airport. An overview of studies on commercial revenues at airports, their main drivers, and their main conclusions is provided in Table 1. These are typically single-country studies, particularly in the United States and Spain, drawing on figures reported by the airports directly and, in a few cases, on consumer surveys. Recent econometric studies conclude that the size of an airport and the number of passengers, average flight times and accordingly the fraction of passengers on international flights and the amount of retail space are the most important drivers of airport commercial revenues. Additionally, Castillo-Manzano's (2010) analysis implies that the share of frequent flyers is a distinct and important as this group may spend more at airport retail outlets. However, leisure travelers tend to spend more money at the airport than business travellers due to longer average dwell times which have a significant impact on an individual's expenses. It is considered that the waiting time prior to embarking and the inclination of travelers to spend more money if they are on a holiday trip explain the decision to consume food/beverages and to make a non-food purchase. Once the decision has been made to spend money, the amount has been shown to increase in line with a passenger's waiting time. Castillo-Manzano (2010) reports that the likelihood of food/beverages being consumed and a purchase being made increases by 31 and 19 percent respectively if the waiting time exceeds 3 h while the amount spent increases by almost 41 percent. The demographic and social characteristics of passengers are found to be important predictors of spending behavior (Castillo-Manzano, 2010):
commercial real estate to the development of airports towards so called airport cities (e.g. Aldridge et al., 2001; Appold and Kasarda, 2013 Poungias, 2009; Reiss, 2007; Tacke, 2008). Several studies focus on the retail sector, for example the characteristics of airport retail, its development and potential for improvement (Giljohann-Farkas, 2008), contractual aspects (e.g. Freathy and O'Connell, 1999; Kim and Shin, 2001) and customer structures (e.g. Geuens et al., 2004). However, the majority of these studies is of descriptive character. Econometric studies in this topic area predominantly deal with retail drivers, in particular drivers of retail revenues in general (Volkova, 2009), determinants of sales (Appold and Kasarda, 2006) and food/beverage consumption and other purchases (Castillo-Manzano, 2010), shopping intentions (Lu, 2014), waiting time (Torres et al., 2005) or time pressure and impulse buying tendencies (Lin and Chen, 2013). The present study expands on these studies by incorporating a large range of potential explanatory factors that impact on revenues from retail and other commercial non-aviation activities of airports. Our analysis includes both standard factors used in previous analyses of this topic such as the number and characteristics of passengers as well as a variable that has not been tested before in this context: the terms of ownership. The results from this comprehensive quantitative analysis performed on a unique panel dataset of global airports contribute to the state of knowledge on the changing business strategies of airports. In particular, it sheds light on a number of key strategic considerations for airports, for example if expanding the commercial area inside a terminal is likely to reap financial rewards and whether it is more profitable to expand the area on the landside or airside area of a terminal. More broadly, we seek to contribute to the study of economies of scale in the non-aviation operations of airports by investigating the effect of different airport size measures on revenues and yields. Finally, a subset of the airports in this study is analyzed for the effect of privatization to establish if it has any bearing on the bargaining power of the airport when negotiating the fraction of sales to be retained by the airport. The following section reviews the existing literature in more depth to set the scene for the specification of our empirical panel data model of airport commercial revenues and feasibility.
families with children have a higher likelihood to make a purchase or to consume food/beverages, but this factor is also the greatest curb on the amount spent, elderly people have a less likelihood to make a purchase or to consume food/beverages, homemaker purchase rarely at an airport, people in a group consume more food/beverages on average than if they are alone, passenger being accompanied to an airport eat/drink something with the accompanying individual.
2. State of research There is a distinct gap in the literature on financial performance and economies of scale of airports, particularly with regard to nonaeronautical operations. Pels et al. (2003) perform an analysis of European airports and find that they are on average inefficient and operate on constant returns to scale when producing air transport movements and on increasing returns to scale when producing air passenger movements. However, their analysis is restricted to aeronautical operations only. Similarly, Martin and Voltes-Dorta (2010) perform an analysis of 41 global airports and find significant economies of scale. In theory, airports should be able to increase their profitability in the commercial sector with increasing size as the marginal cost of managing the commercial operations such as retail outlets and other facilities should fall. Fixed cost elements can be spread over a larger base of commercial operations. Further benefits may accrue from access to specialized labor and equipment which may not be available to smaller establishments. There may be additional size effects for airports which go beyond the definition of economies of scale in the narrow sense of the term. For example, some high-revenue facilities, for example an airport-based convention center, may only be feasible for airports above a certain size threshold. Retail operations may also be affected by airport size, in terms of the scope and depth of the retail products that are offered. However, Empirical evidence in the retail sector is rare. Arndt and Olsen (1975) study the Norwegian retail sector and find only limited evidence of economies of scale. The authors point out a number of limitations to economies of scale in retail, the most important among these being that retail is not a closed system such as manufacturing and retailers have very limited control over production and customer flows. Thus, economies of scale may be
A further theme of empirical studies of non-aviation revenues is the impact of the location and quality of the airport retail outlets and the role of the regulatory environment of airports (Del Chiappa et al., 2016; Forsyth, 2004). In particular, an evaluation of single and dual till models is undertaken in some studies but the discussion as to which approach is preferable remains controversial (e.g. Evangelinos et al., 2011; Kratzsch and Sieg, 2011; Zhang and Zhang, 2010). While the information on single or dual till regulation was not available in our global study of airports, the distinction between the two models may have ramifications for the commercial success of an airport. The single-till approach takes all areas of the airport including the non-aviation sector into consideration in determining the remuneration levels in the regulated aviation sector. By contrast, the dual-till approach typically separates areas of an airport which are necessary for the actual provision of services in air transport from the revenues (or income) from the commercial activities in the nonaviation sector. Non-aviation expenditure and costs incurred in this sector are then excluded from the regulation of charges in the aviation sector in the dual-till model. In a related study, Bel and Fageda (2009) find that the level of charges that airlines have to pay airports depends on the ownership structure (state-run, partially privatized, fully privatized) but their study does not include commercial revenues. A further strand of the literature is concerned with physical and network characteristics of an airport and their impact on commercial performance. For instance, Van Dender (2007) reports that concessions revenues per departing passenger are lower at hub airports in the United 2
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 1 Overview of selected studies on commercial revenues at airports and their main drivers (þ positive influence, - negative influence). Authors
Place
Data and Methodology
Main conclusions
Torres et al. (2005)
Spain
a non-parametric approach, 997 passengers at one airport
Appold and Kasarda (2006)
USA
OLS regression, 75 airports
Tovar and Martín-Cejas (2009)
Spain
parametric input Distance Func-tions (DEA), 26 airports
Graham (2014)
Worldwide
comparative assessment of data, general information about large number of airports
“The Moodie Report (2009)”
Worldwide
among others correlation analysis, 45 airports
Volkova (2009)
13 airports from EU countries
panel data econometric analysis, 13 airports
Castillo-Manzano (2010)
Spain
bivariate probit model, over 20,000 passengers, 7 airports
Fuerst et al. (2011)
45 airports from EU countries
regression analysis
vacationers spend more than business travelers (þ) length of stay prior to boarding effects the consumption in commercial area (þ) level of consumption is independent of the waiting time () number of passengers have the single largest effect on retail sales (þ) average distance flown has an effect on food and beverage sales (þ) average distance of traveling and amount of retail space have a positive effect on sales per passenger (þ) type of airport: international gateway status, major tourist destination and clearance time () passenger traffic not significant for sales per passenger () proportion of international passengers that circulating through the airport terminal (þ) airports with well-developed commercial activities are more efficient than those that focus on aeronautical revenues leisure charter passengers are good shoppers (þ) young leisure passengers who travel several times a year are high spenders (þ) LCC passengers are good users of food and beverage (þ) foreigners who have high duties and taxes buy more tax-free products (þ) women are more likely to fall into the category “shopping lovers” (þ) transfer passengers are less likely to make use of facilities and if they do, tend to spend less () international and leisure passengers spend more because of longer dwell time and availability of duty free (þ) large airports have more diverse, higher yielding commercial activities and to attract strong brands (þ) share of domestic passengers inconclusive (neutral) increasing floor space indirectly generates better returns but not higher sales per passenger (±) low-cost carrier passengers are not necessarily low spenders extra-EU passengers increase retail revenue per square meter at hub airports (þ) number of short stay parking places, check-in facilities and the number of employees contribute to the retail revenue (þ) retail revenue per square meter grows significantly once critical mass of space is reached (þ) bars/restaurants influence the retail revenue and generate externalities (þ) extra-EU passengers have no effect on retail revenue at regional airports (neutral) waiting time prior to embarking, being on vacation, being a frequent flyer and traveling with children (þ) number of group members and accompanied passengers explain the consumption of food/beverages (þ) passengers from outside Eurozone more likely to consume food/beverages, albeit just loose change (þ) passengers flying to international European destinations (þ) elderly people purchase less and spend less on food/beverages () passenger's arrival at the airport on a courtesy bus by the hotel () business passengers not likely to make last-minute purchases at the airport () passenger using a LCC have a lower likelihood to purchase food/beverages () main drivers of commercial revenues per passenger include the number of passengers passing through the airport, the ratio of commercial to total revenues, national income, the share of domestic and leisure travelers and the number of flights (þ) a large amount of retail space per passenger is generally associated with lower commercial revenues per square meter confirming decreasing marginal revenue effects () business travelers have a negative influence on commercial revenues per passenger ()
Source: own compilation
well as the willingness to pay for reductions in reduced ground travel time. Moreno and Muller (2003) demonstrate empirically for Sao Paolo that airport choice in multi-airport urban regions is influenced by these local accessibility considerations.
States. The size of an airport, typically proxied by passenger numbers and aircraft movements, the regional or local income of the area in which the airport is located as well as the size and configuration of the retail surface area are found to be key drivers of commercial revenue. Moreover, the present analysis includes additional factors which have not or rarely been investigated in studies on airport revenues, notably the terms of ownership (state-run, partially privatized, fully privatized). Recent empirical evidence (Adler et al., 2015) suggests that a move towards incentive regulation of airports is associated with increased productive efficiency. Finally, a number of studies consider the implications of airport accessibility and location in the urban or regional context. While outside of the scope of the present global study of airports, the distance, travel time or travel cost to the Central Business District (CBD) and other destinations has been considered in empirical case studies (Koster et al., 2011; Kouwenhoven, 2008; Reynolds-Feighan and McLay, 2006). These include trip timing decisions of travelers going to and from the airport as
3. Data and analytical framework Research into airport commercial revenues and profitability has been held back to date by severe constraints in the availability and quality of relevant data. The principal source of information for the present analysis is the Moodie Report which contains revenue and other financial performance data for a large number of global airports. The Moodie Report has been published since 2001. However, the format of the reporting was changed from the 2011 report onwards which no longer allows tracking of performance data for individual airports. Extracting the data from the Moodie Reports and supplementing the airport-level data with economic 3
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 2 Definitions and sources of variables. Variable
Definitions
Source
Commercial revenues
Income from retail and commercial activities (duty free, news and gifts, specialty retailing, food and beverage and currency exchange) Percentage of airport commercial revenue divided by gross commercial sales made by the concessions. It shows the percentage of total sales that is retained by the airport/airport operator. Gross domestic product per capita of the country where the respected airport is located in purchasing power parities The area, measured in square metres, devoted to commercial activities, it excludes passageways, cartilage areas or common seating areas in food court as well as facilities outside of passenger terminals Part of the terminal building after check-in, security, customs and passport control. This airside portion is typically only accessible to passengers. One form of commercial outlets, e.g. full-service and fast food restaurants Passengers departing or arriving at an airport Passengers traveling to an international destination (regardless of their nationality) Airport is partially privatized or owned by the government
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Commercial yield
GDP per capita Commercial area
Outlets airside Outlets food & beverage Passengers International travelers Partially privatized and staterun
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Datastream Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09 Moodies Study 2000/01, 2002/03, 2006/07, 2008/09 ICAOdata ICAOdata Phone/email/internet surveys of individual airports
of the retail offer, proxied by the percentage of shops not classified under the basic categories food and beverage, duty free and gift shops. A particularly strong relationship is detected between commercial revenues per passenger and the size of commercial floorspace. These potential agglomeration economies and economies of scale/scope will be tested more formally in the ensuing regression analysis. The commercial yield exhibits a similar, albeit weaker association with size and variety measures. The analytical strategy for estimating the impact of various drivers on the commercial performance of airports includes three performance metrics (commercial yield, total commercial revenues and commercial revenues per passenger) evaluated with three different estimation techniques (pooled OLS, random effects panel and 3 stage least squares). The basic pooled OLS single equation estimation of the commercial yield takes the following log-log form:
indicators yielded a panel dataset of global airport indicators for the period 1998–2009 (The Moodie Report, 2001, 2003, 2007 and 2009). Table 2 gives an overview of the variables employed in our panel data analysis. It is important to note that several airport operators do not provide a breakdown of revenues by individual airports. For example, the revenues reported for the British BAA Group comprise the three London airports Heathrow, Gatwick and Stansted. However, we were able to infer the volume of revenues for individual airports by using airport-level information provided by these operators such as fee and cost structures, the number of passengers etc. All monetary values were then converted into SDR (Special Drawing Rights), an international foreigm-exchange reserve asset created by the IMF to ensure comparability using the interbank exchange rate at the call date of the respective annual statements. Supplementary research and data quality control were conducted via telephone and email surveys with airport operators. The database used in this analysis includes information on 75 airports over a period of 11 years, albeit as an unbalanced panel with a considerable number of gaps due to limited data availability. Fig. 2 in the appendix displays the locations of the airports in this study. The selection of these airports was driven by data availability in the sources listed in Table 2. It cannot be ruled out that regression estimates of the entire population of global airports may differ from the ones reported below but, as far as we are aware, the current database represents the largest global database of its kind ever used in an empirical study. Table 3 examines different key characteristics of the dataset used in this analysis. As expected, standard deviations are relatively high for most variables, reflecting the broad range of values across the world. A comparison of the percentage for the food & beverage outlets in the four regions reveals that these outlets are more important in the Americas and Europe than in Asia Pacific and Africa. However, the average commercial area at airports in Africa, North America and Europe is similar, while Asia Pacific and South America have less commercial space. Interestingly, North America has the lowest commercial yield (the fraction of concession sales retained by the airport). This may be attributable to higher levels of competition among airports in the North American market. As airport size may be an important driver of commercial revenue, we examine the distribution by number of passengers and – as expected – the highest volume is found in the Americas. Finally, GDP per capita reveals a ranking as expected, North America before Europe, followed by Asia Pacific, South America and Africa. Next, we perform an exploratory analysis of the key financial performance metrics. Fig. 1 shows scatterplots of revenues per passenger and commercial yields with other variables. The first measure correlates significantly with size measures, i.e. the total area of commercial floorspace in an airport and the total number of retail outlets. A significant relationship is also found between revenue per passenger and the variety
ln CYi ¼ β0 þ β1 ln CATi þ β2 ln FBi þ β3 ln RTi þ β4 ln PIi þ β5 ln AIRi þ β6 ln CRPi þ β7 PRIVi þ β8 STATi þ εi
(1)
where CY is the commercial yield, CAT is the total commercial floorspace, FB is the fraction of food and beverage outlets, RT is total revenue in SDR, PI is the fraction of international passengers, AIR is the fraction of outlets that are located airside, CRP is the commercial revenue per passenger and PRIV and STAT are dummy variables indicating whether an airport is fully privatized, partially privatized or state-run. The estimation is conducted both as a pooled OLS and as a random-effects panel data analysis to ensure that the results are robust to modelling technique. Fixed effects results were also estimated to conduct a Hausman test with the results confirming the use of a random effects model. Analogous to Equation (1), we then estimate the equations for commercial revenues and commercial revenues per passenger. An obvious concern of our single-equation estimation is that the commercial performance indicators used as explanatory variables may be endogenous with the dependent variables. For example, commercial yields and revenues as well as revenues per passenger may be codetermined. Hence, we employ a more advanced 3-stage least squares (3SLS) technique following Zellner and Theil (1962). The 3SLS estimation models the endogenous variables in a system of equations, obtaining instrumental variable estimates conditional on the covariances of the error terms in each individual equation. Due to the endogeneity problem, the dependent variables will be correlated with the error terms but the 3SLS estimation takes this problem into account by taking the remaining explanatory variables as exogenous for obtaining unbiased estimates of the three endogenous variables (Greene, 2000). The specification of our 3SLS system is as follows:
4
Transport Policy xxx (2017) 1–9
ln CYi ¼ β9 þ β10 ln CATi þ β11 ln FBi þ β12 ln RTi þ β13 ln PIi 3 3656 1347 34 20,321 14,430 57 26,676 12,427 45 36,950 7240 3 5355 2251 144 31,277 13,803
GDP/capita (SDR)
F. Fuerst, S. Gross
þ β14 ln AIRi þ β15 ln CRPi þ φi
(2)
ln CRi ¼ β16 þ β17 ln CATi þ β18 ln FBi þ β19 ln Pi þ β20 ln PIi (3)
ln CRPi ¼ β23 þ β24 ln CATi þ β25 ln FBi þ β26 ln Pi þ β27 ln PIi 3 44.00% 2.59% 34 32.15% 32.02% 57 84.53% 18.54% 45 18.00% 25.29% 3 58.00% 3.06% 144 49.10% 34.68%
þ β28 ln GDPi þ β29 ln AIRi þ ξi
(4)
Apart from the variables employed in Equation (1), P is the total number of passengers and GDP is the GDP per capita of the airport's location as a proxy for economic development of the surrounding region.
3 12.40 4.46 34 17.40 14.20 57 17.10 29.00 45 20.60 23.50 3 0.66 0.14 144 17.90 23.90
4. Results & discussion
3 20.83% 5.97% 32 16.95% 7.92% 52 13.98% 7.41% 45 9.68% 9.65% 3 36.86% 4.87% 135 13.32% 9.00%
3 1.24 0.31 32 1.34 1.61 54 1.95 1.14 45 0.49 0.96 3 1.49 0.34 137 1.27 1.30
3 13,471 1845 33 8949 9367 55 10,474 22,046 45 12,060 9415 3 7160 2264 141 10,351 15,681
3 22.10% 4.80% 32 22.00% 10.53% 52 25.35% 9.12% 44 43.05% 12.19% 3 21.00% 4.75% 136 28.00% 13.36%
The first proposition to be tested empirically is whether the total volume of airport commercial revenues can be explained by the factors as hypothesized in the previous section. Note that a set of year dummy variables is used to control for unobserved time trends and longer term developments. The panel data estimation results of Models 1 and 2 demonstrate that larger airports (proxied by the number of passengers) have higher commercial revenues. While this result is unsurprising, it is interesting to note that even when the number of passengers is kept constant, we find a significant relationship between the square footage of commercial space in an airport and its commercial revenues, arguably because a larger total area of commercial floorspace in an airport tends to generate higher commercial income. The share of international passengers is also found to be a significant driver, implying that international travellers tend to have higher expenditure on commercial activities than domestic travellers. This finding is in line with most existing studies (with the exception of the recent study of German airports by Fasone et al., 2016). However, it is also possible that airports with a high percentage of international travellers have a larger proportion of high-revenue generating commercial activities such as high-end retail outlets selling international designer fashion, jewellery and watches which are predominantly found in international airports and to a lesser extent in airports serving the domestic market. Further analysis is required to elucidate the reasons for these findings. Since the airports in our data sample vary considerably in terms of their size, ranging from Riga Airport with just under 2 million passengers to Atlanta International Airport with approximately 90 million passengers we estimate the drivers of commercial revenue on a per passenger basis in a next step. Models 3 and 4 underline the importance of large commercial areas for generating higher revenues per passengers. A possible explanation for this may be found in economies of scale and economies of scope in larger commercial and retail operations. The retail outlets and other commercial activities may also create positive spillover effects and externalities for one another. Conversely, keeping the size of the commercial area constant, an increase in the total number of passengers appears to affect commercial revenues per passenger either insignificantly (Model 3) or even negatively (Model 4). Hence, revenues per passenger seem to be driven more by the supply of retail and other commercial outlets rather than by the number of passengers passing through the airport. This may also be taken as an indication of a shift in the role of an airport as a venue for commercial activities, entertainment and business rather than a pure infrastructure facility. This hypothesis is put forward most prominently by Kasarda (2006) and Appold and Kasarda (2006, 2013). While the estimation results in Models 1–4 show that total commercial revenue increases both with increasing commercial area and number of passengers, the negative coefficient of passengers on revenue per passenger hints at the fact that, after controlling for commercial square footage, an increase in the number of passengers might lead to adverse effects, for example through crowding, long wait times at
Total
South America
North America
Europe
Asia Pacific
N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. Africa
Table 3 Summary statistics.
Statistics
Commercial yield (in %)
Commer-cial revenue (SDR)
Commercial area (sq m)
Food & beverage out-lets on total outlets (in %)
Passenger total (Mio.)
International passengers (%)
þ β21 ln GDPi þ μi
5
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Fig. 1. Scatterplot of commercial airport performance indicators. *p < 0.05.
A negative association is found with the level of total revenues, i.e. larger airports with higher total revenues appear to have lower commercial yields. Viewed in conjunction with the coefficient on commercial area, this might be suggestive of a space productivity effect. In other words, higher revenues in a given area of commercial space may entail a lower yield demanded by the airport. Less productive commercial spaces may have to pay a higher share of their revenues to the airport to meet the required rates. However, there could also be other reasons for this finding. For example, if the airport has a large passenger base, it can earn sufficient revenue from the aviation side to cover the cost and is not dependent on extracting a high share of commercial revenues from concessions and other commercial operations. The percentage of international passengers is positive and significant in Model 5 but not in Model 6. The percentage of outlets located airside (as opposed to landside) is insignificant for commercial yields in both estimations. Revenue per passenger is inversely related to commercial yields which suggests that airports demand a minimum share of turnover from concessions that does not increase linearly with higher revenue per passenger. This means that concessions with high per passenger revenues
the check-out, excessive noise etc, thereby reducing the willingness of passengers to frequent shops and make purchases. These effects may present diseconomies of scale, particularly if airport size fails to keep up with an increase in the number of passengers. However, more in-depth information, for example on the ratio of purchasing to non-purchasing passengers would be necessary to test this proposition empirically. It is also noteworthy that the economic development status as measured by GDP per capita of the country in which an airport is located is shown to have predictive power for the airport's commercial revenues both in terms of total volumes and on a per passenger basis. In the next step, we test the drivers of airport commercial yields (Models 5–7 of Table 4). While the yield as defined in our dataset is not a direct measure of the rate of return or profitability of airport commercial space in general, it affords us an indication of the bargaining power of an airport and, indirectly, of the potential productivity and profitability of a commercial space. We also find again that airports with a higher amount of commercial areas and floorspace tend to have higher commercial yields, underlining the positive impact of size and, potentially, higher bargaining power of larger airports. 6
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 4 Estimation results (pooled OLS and random effects). Dependent V
(2) Commercial revenues (SDR million)
(1) Commercial revenues (SDR million)
(4) Commercial revenues (SDR) per passenger
(3) Commercial revenues (SDR) per passenger
(5) Commercial yield (in %)
(6) Commercial yield (in %)
(7) Commercial yield (in %)
Technique
Pooled OLS
Random Effects
Pooled OLS
Random Effects
Pooled OLS
Random Effects
Random Effects
Commercial area total (log) %Outlets food &beverage (log) Revenue total (log) %International passenger (log) %Outlets airside (log) Commercial revenue per passenger Passenger total (log) GDP/capita (log) Staterun airport Partially privatized airport Constant
0.323** (3.20)
0.465*** (4.69)
0.323** (3.20)
0.465*** (4.69)
0.0590** (3.18)
0.0542** (2.78)
0.00809 (0.39)
0.122 (1.00)
0.0170 (0.14)
0.122 (1.00)
0.0170 (0.14)
0.0813*** (-4.23)
0.0751*** (-3.90)
0.00747 (0.25)
0.260*** (6.71)
0.195*** (5.02)
0.260*** (6.71)
0.195*** (5.02)
0.0325* (-2.16) 0.0187* (2.35)
0.0310 (1.92) 0.0105 (1.31)
0.00676 (0.34) 0.0525** (3.01)
0.389*** (-4.36)
0.292** (-3.12)
0.389*** (-4.36)
0.292** (-3.12)
0.0144 (0.84) 0.0480* (-2.38)
0.0195 (1.09) 0.0315 (1.54)
0.0537 (1.78) 0.0613* (-2.18)
0.874*** (9.98) 0.245*** (4.26)
0.715*** (7.80) 0.203** (2.68)
0.126 (1.44) 0.245*** (4.26)
0.285** (-3.10) 0.203** (2.68) 0.114* (-2.18) 0.101* (-2.10)
1.975 (1.63)
0.655 (0.48)
1.975 (1.63)
0.655 (0.48)
0.253 (1.72)
0.234 (1.38)
0.162 (0.59)
Time controls (annual)
Yes
Yes
Yes
Yes
Yes
Yes
No
N R2 adj. R2 R2 within R2 between AIC Hausman test prob BIC
121 0.890 0.874
121 0.87
121 0.641 0.590
121 0.60
123 0.406 0.375
123 0.47
40 0.29
0.23 0.51
0.49 0.31
0.65 0.89 130.9
0.32 0.60 289.4
130.9 0.99
175.6
0.99
0.12 269.7
175.6
t statistics in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.
Table 5 Estimation results (3-stage least squares).
Commercial revenue per passenger
0.0427* (-2.57)
(10) Commercial revenue (SDR) per passenger
(9) Commercial revenue (SDR million)
(8) Commercial yield (in %) Commercial area total (log)
0.323*** (3.43)
%Outlets food &beverage (log)
0.122 (1.07)
Commercial area total (log) %Outlets food & beverage (log) Passenger total (log) %International passenger (log) GDP/capita (log) %Outlets airside (log)
0.927*** (3.94) 0.291* (-1.07) 0.467** (2.29) 0.292** (3.24)
Commercial area total (log) %Outlets food & beverage (log)
0.0602*** (3.49) 0.0590*** (-3.13)
Passenger total (log) %International passenger (log)
0.874*** (10.71) 0.260*** (7.20)
Revenue total (log) %International passenger (log) %Outlets airside (log)
0.0408* (-2.76) 0.0214** (3.09 0.0161 (1.00)
GDP/capita (log) %Outlets airside (log)
0.245*** (4.58) 0.389*** (-4.68)
Time controls (annual)
Yes
Yes
Yes
N R2
120 0.49
120 0.89
120 0.50
0.113 (1.32) 0.149 (0.72)
t statistics in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.
that the size of the commercial area is a powerful determinant for all three financial performance metrics. Additionally, the fraction of international passengers is a significant predictor of commercial revenue and commercial revenue per passenger. A high percentage of food and beverage outlets in the terminal is associated with lower yields and lower commercial revenue per passenger. As compared to other more expensive items sold in airport outlets, the food and restaurants business normally incurs a lower profit margin which may, in turn, lower the percentage of sales revenues retained by the airport.
retain a larger fraction of their revenues, thus potentially increasing their profitability even more. In a final variation on model specifications, we test whether the operating structure of an airport has a significant impact on the size of the commercial yield. Model 7 includes binary variables for state run, partially and fully privatized airports. The results indicate that indeed privatized airports exhibit higher yields than their staterun or partially privatized counterparts but an important caveat of these findings is the much smaller sample size for which this information was available. Turning to the 3SLS estimation in Table 5, we model each of the dependent variables as endogenous in a system of equations to correct for any endogeneity bias introduced by the presence of financial metrics that are jointly and simultaneously determined. The results from this estimation are broadly similar to the single-equation estimation. We find
5. Conclusions This study set out to test the factors affecting the commercial performance of global airports. To this end, we analyzed a panel dataset of 7
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
The analysis showed some, albeit inconsistent evidence of a negative effect of the share of food and beverage outlets on revenue per passenger and commercial yields (4 of the 10 model estimations confirm this effect). This may lead airports to conclude that this segment is best reduced and substituted with higher-value outlets in a strategy to increase per passenger revenue. However, this strategy is not necessarily warranted. Instead, it should be considered how passengers can be introduced to such offers or how these products can be made palatable in times of increasingly rare free catering services on board. Regionally sourced, seasonal, fresh and health-oriented offers are some currenttrends in gastronomy. Regarding airside/landside placement of shopping facilities, one possible approach may be to make more innovative offers accessible to passengers while they are waiting for boarding. Larger und pleasantly designed recreational areas, shopping and dining offers have been implemented at many airports along with technological enhancements, such as apps, mobile payment, online platforms, QR-Code-ShoppingWalls, touch screen monitors. A more in-depth study of the linkages between airport commercial revenues, local, national and global economies is required to ascertain these risk-mitigating effects. The present study was also limited by gaps in data coverage both in terms of the number of years available as well as the geographical distribution of the airports. For example, it was not possible to incorporate several of the most important global airports in our analysis because of a lack of data. Further research is also needed to corroborate the reported findings for a larger sample of global airports and a more thorough investigation of the financial performance metrics of commercial activities such as information on cash flows, return on equity, return on assets, stock market returns and liquidity measures would be desirable. The present analysis provides a foundation for these more fine-grained and expanded studies if and when the required datasets become available.
75 airports in 30 countries. Applying pooled OLS, random effects and 3SLS estimation techniques, the main drivers of financial performance were identified. Our results are broadly in line with the existing literature. The share of international passengers, the size of the commercial area as well as airport size and the mix and intra-terminal location of retail space are found to be significant determinants. The latter finding the presence of economies of scale in generating higher commercial revenue and ensuring profitability for an airport's non-aeronautical operations. Staterun and partially privatized airports appear to retain a significantly smaller share of concession sales than their privatized counterparts. Regarding the retail mix, a higher share of food and beverage outlets appears to decrease the revenue share retained by the airport while a higher percentage of outlets located airside rather than landside depresses commercial revenues. The corollary of these findings for airports seeking to enhance their financial performance in the non-aeronautical sector is that careful planning of the size and mix of retail and other commercial areas and their positioning in the airside/landside structure may be as important as the decision to privatise or remain state-run. Regarding airport management strategy, the results indicate that airports should seek measures to retain airlines with a large fraction of international passengers and, consequently, the higher spending level of these international passengers. On the other hand, it should be considered how domestic travelers can be encouraged to increase dwelling times and expenses. Customer loyalty programs that are already implemented in some airports are a possible strategy. In addition, the visitors may receive current promotions via Smartphone by entering the participating stores (Jegminat, 2015). Other examples are the CPH Advantage program at Copenhagen airport, the BAA WorldPoints of the British Airport Authority, the ViaMilano programme at Malpensa Airport or the Changi Rewards programme at Singapore's Changi airport.
Appendix
Fig. 2. Location of airports in this study.
Appold, S.J., Kasarda, J.D., 2013. The Airport City phenomenon: evidence from large US airports. Urban Stud. 6, 1239–1259. Arndt, J., Olsen, L., 1975. A research note on economies of scale in retailing. Swed. J. Econ. 207–221. Bel, G., Fageda, X., 2009. Privatization, Regulation and Airport Pricing: an Empirical Analysis for Europe. EUI Working Papers RSCAS 2009/27. Florence/Italy. Castillo-Manzano, J.I., 2010. Determinants of commercial revenues at airports: lessons learned from Spanish regional airports. Tour. Manag. 6, 788–796.
References Adler, N., Forsyth, P., Mueller, J., Niemeier, H.M., 2015. An economic assessment of airport incentive regulation. Transp. Policy 41, 5–15. Aldridge, K., Carreno, M., Ison, S., Rye, T., Straker, I., 2001. Car parking management at airports: a special case? Transp. Policy 13, 511–521. Appold, S.J., Kasarda, J.D., 2006. The appropriate scale of US airport retail activities. J. Air Transp. Manag. 12, 277–287.
8
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9 Lu, J.L., 2014. Investigating factors that influence passengers' shopping intentions at airports – evidence from Taiwan. J. Air Transp. Manag. 35, 72–77. Martin, J.C., Voltes-Dorta, A., 2010. International airports: economies of scale and marginal costs. J. Transp. Res. Forum 47 (1). Moreno, M., Muller, C., 2003. The influence of ground accessibility on airport choice in Sao Paulo metropolitan area. In: Proceedings 19th Dresden Conference of Traffic and Transportation Sciences, 22.-23.09.2003. Dresden. Pels, E., Nijkamp, P., Rietveld, P., 2003. Inefficiencies and scale economies of European airport operations. Transp. Res. Part E Logist. Transp. Rev. 39 (5), 341–361. Poungias, P., 2009. Airport city developments: an airport investor's perspective. J. Airpt. Manag. 4, 14–22. Qin, J., Olaru, D., 2013. System Dynamcis based simulation for airport revenue analysis. In: Conference Proceedings 31st International Conference of the System Dynamics Society, vol. 4. Curran Associates, Inc., Red Hook, NY, pp. 1–23. Reiss, B., 2007. Maximising non-aviation revenue for airports: developing airport cities to optimise real estate and capitalise on land development opportunities. J. Airpt. Manag. 3, 284–293. Reynolds-Feighan, A., McLay, P., 2006. Accessibility and attractiveness of European airports: a simple small community perspective. J. Air Transp. Manag. 12, 313–323. Tacke, G., 2008. Price optimization for airport parking. In: Proceedings 11th Hamburg Aviation Conference 13.-15.02.2008 (CD-rom). Hamburg. The Moodie Report, 2001. The Airport Commercial Revenue Study – 2000/2001. Brentford. The Moodie Report, 2003. The Airport Commercial Revenue Study – 2002/2003. Brentford. The Moodie Report, 2007. The Airport Retail Study – Fourth Edition 2006/07. Brentford. The Moodie Report, 2009. The Airport Commercial Revenues Study – Fifth Edition 2008/ 09. Brentford. Torres, E., Domínguez, J.S., Valdes, L., Aza, R., 2005. Passenger waiting time in an airport and expenditure carried out in the commercial area. J. Air Transp. Manag. 6, 363–367. Tovar, B., Martín-Cejas, R.R., 2009. Are outsourcing and non-aeronautical revenues important drivers in the efficiency of Spanish airports? J. Air Transp. Manag. 5, 217–220. Van Dender, K., 2007. Determinants of fares and operating revenues at US airports. J. Urban Econ. 62, 317–336. Volkova, N., 2009. Determinants of Retail Revenue for Today's Airports. GAP Project Working Paper, Berlin. Zellner, A., Theil, H., 1962. Three stage least squares: simultaneous estimate of simultaneous equations. Econometrica 29, 54–78. Zhang, A., Zhang, Y., 2010. Airport capacity and congestion pricing with both aeronautical and commercial operations. Transp. Res. Part B Methodol. 44 (3), 404–413.
Del Chiappa, G., Martin, J.C., Roman, C., 2016. Service quality of airports' food and beverage retailers. A fuzzy approach. J. Air Transp. Manag. 53, 105–113. Deloitte, 2009. Europ€ aische Flughafenbetreiber – Hohe Profitabilit€atsniveaus sind nur eine Seite der Medaille, p. 2009. Evangelinos, C., Püschel, R., Goldhahn, S., 2011. Inverting the Regulatory Rules? Optimizing Airport Regulation to Account for Commercial Revenues. Diskussionsbeitr€ age aus dem Institut für Wirtschaft und Verkehr 01/2011. Fasone, V., Kofler, L., Scuderi, R., 2016. Business performance of airports: non-aviation revenues and their determinants. J. Air Transp. Manag. 53, 35–45. Forsyth, P., 2004. Locational and monopoly rents at airports: creating them and shifting them. J. Air Transp. Manag. 10, 51–60. Fraport, A.G., 2013. Heinemann Duty Free und Fraport machen mobil, „Virtual Shopping“ erstmals am Frankfurter Airport. Retrieved from, 20.10.2016. http://www.fraport. de/de/presse/newsroom/service-news/2013/heinemann-duty-free-und-fraportmachen-mobil.html. Fraport, A.G., 2015. Online Shopping at Frankfurt Airport: Products from Airport Shops Available via the FRA App and Website. Retrieved from, 20.10.2016. http://www. fraport.com/en/press-center/reports–-publications-/archive/2015/online-shoppingat-fra-2015.html. Freathy, P., O'Connell, F., 1999. A typology of European airport retailing. Serv. Ind J. 3, 119–134. Fuerst, F., Gross, S., Klose, U., 2011. The sky is the limit? The determinants and constraints of European airports commercial revenues. J. Air Transp. Manag. 17, 278–283. Geuens, M., Vantomme, D., Brengman, M., 2004. Developing a typology of airport shoppers. Tour. Manag. 25, 615–622. Giljohann-Farkas, K., 2008. Airport Retailing, DEGI Research. December 2008. Graham, A., 2014. Managing Airports – an International Perspective, fourth ed. New York. Greene, W.H., 2000. Econometric Analysis, fourth ed. Upper Saddler River, New Jersey. Jegminat, G., 2015. Flughafen Frankfurt – Online-Portal für Terminal-Shops. Retrieved from, 25.11.2016. http://www.fvw.de/flughafen-frankfurt-online-portal-fuerterminal-shops/393/150887/18663. Kasarda, J.D., 2006. The rise of the aerotropolis. Next Am. City 10, 35–37. Kim, H.-B., Shin, J.-H., 2001. A contextual investigation of the operation and management of airport concessions. Tour. Manag. 22, 149–155. Koster, P., Kroes, E., Verhoef, E., 2011. Travel time variability and airport accessibility. Transp. Res. Part B Methodol. 10, 1545–1559. Kouwenhoven, M., 2008. The Role of Accessibility in Passengers' Choice of Airports. Discussion Paper No. 2008-14. The Hague/Netherlands. Kratzsch, U., Sieg, G., 2011. Non-aviation revenues and their implications for airport regulation. Transp. Res. Part E 47, 755–763. Lin, Y.-H., Chen, C.-F., 2013. Passengers' shopping motivations and commercial activities at airports – the moderating effects of time pressure and impulse buying tendency. Tour. Manag. 36, 426–434.
9
Contents lists available at ScienceDirect
Transport Policy journal homepage: www.elsevier.com/locate/tranpol
The commercial performance of global airports Franz Fuerst a, b, *, Sven Gross c a
University of Cambridge, Dept of Land Economy, 16-21 Silver St, Cambridge CB3 9EP, United Kingdom University of Melbourne, THRIVE Research Hub, Parkville, VIC 3010, Australia c Harz University of Applied Sciences, Friedrichstraße 57-59, 38855 Wernigerode, Germany b
A R T I C L E I N F O
A B S T R A C T
Keywords: Commercial revenues Non-aeronautical airport income Airport finances
Revenues from non-aeronautical business have received increasing attention from airports seeking to enhance their profitability. This study analyzes the commercial performance of global airports using a panel dataset of 75 airports in 30 countries. Applying pooled OLS, random effects and 3SLS estimation frameworks, we identify the main drivers of financial performance. Our results are in line with the existing literature. The share of international passengers, the size of the commercial area as well as airport size and the mix and intra-terminal location of retail space are found to be significant determinants. The latter finding suggests the presence of economies of scale in generating higher commercial revenue and ensuring profitability for an airport's non-aeronautical operations. Staterun and partially privatized airports appear to retain a significantly smaller share of concession sales than their privatized counterparts. Regarding the retail mix, a higher share of food and beverage outlets appears to increase the revenue share retained by the airport while a higher percentage of outlets located airside rather than landside depresses commercial revenues.
1. Introduction Airports operate in a highly cyclical business environment. The demand for moving people and cargo on airplanes is typically more volatile than equivalents for other industries or even within the transport sector. One possible strategy to reduce exposure to these demand shocks is for airports to expand their non-aeronautical activities and diversify their activities away from full reliance on the aeronautical sector. Another strategy is the installation of activities without a physical connection to the terminal, such as eCommerce projects, for example QR code walls and retail-specific apps at airports owned by the Fraport group (Fraport, 2013, 2015). The main argument in favor of the first strategy is that the nonaeronautical segment encompasses a more diverse range of customer groups. Apart from passengers, airport employees and visitors that are all closely connected to the aeronautical operations of an airport, the nonaeronautical commercial and retail facilities are additionally also frequented by local residents, employees and visitors of companies locating near the airport and other types of customers that are not airline passengers. Increasingly, airports also include convention centers and other entertainment, business and leisure facilities which are largely independent of fluctuations in air passenger volumes. Furthermore, the
expansion of the non-aeronautical business is also attractive because of its lower operational cost structure compared to the aeronautical business which may make it more profitable. In some cases, non-aeronautical revenues even account for the largest share of the profits of an airport operator. Increasingly, major real estate companies cover airport real estate in their analysis and investments as a distinct property sub-type and credit rating agencies take the diversification of revenue sources into consideration in their financial ratings of airports (Deloitte, 2009). The importance of airport commercial activities and the airport retail sector in particular has received considerable attention over the last decade, particularly as liberalization and deregulation of the aviation industry has forced airports worldwide to find new sources of income to replace the aviation revenues loss because of the increasing market power of airlines such as Ryanair, EasyJet or other Low Cost Carriers or also to third party ground service operators. Despite the obvious importance of commercial activities in light of these recent developments, research on the commercial performance of airports and its determinants is still scarce. This paper presents the first comprehensive academic study of global commercial airport performance. There exists a growing body of empirical studies of the non-aviation business, particularly on individual business activities ranging from parking space management and parking revenue (e.g. Qin and Olaru, 2013) to
* Corresponding author. University of Cambridge, Dept of Land Economy, 16-21 Silver St, Cambridge, CB3 9EP, United Kingdom. E-mail address: [email protected] (F. Fuerst). http://dx.doi.org/10.1016/j.tranpol.2017.08.005 Received 2 August 2016; Received in revised form 29 July 2017; Accepted 15 August 2017 Available online xxxx 0967-070X/© 2017 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Fuerst, F., Gross, S., The commercial performance of global airports, Transport Policy (2017), http://dx.doi.org/ 10.1016/j.tranpol.2017.08.005
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
both difficult to achieve and difficult to measure in a retail context. This limitation is particularly true for the present analysis which does not include measures of production costs due to unavailability of suitable data and instead proxies size effects with available measures such as size of the commercial area and total number of passengers of an airport. An overview of studies on commercial revenues at airports, their main drivers, and their main conclusions is provided in Table 1. These are typically single-country studies, particularly in the United States and Spain, drawing on figures reported by the airports directly and, in a few cases, on consumer surveys. Recent econometric studies conclude that the size of an airport and the number of passengers, average flight times and accordingly the fraction of passengers on international flights and the amount of retail space are the most important drivers of airport commercial revenues. Additionally, Castillo-Manzano's (2010) analysis implies that the share of frequent flyers is a distinct and important as this group may spend more at airport retail outlets. However, leisure travelers tend to spend more money at the airport than business travellers due to longer average dwell times which have a significant impact on an individual's expenses. It is considered that the waiting time prior to embarking and the inclination of travelers to spend more money if they are on a holiday trip explain the decision to consume food/beverages and to make a non-food purchase. Once the decision has been made to spend money, the amount has been shown to increase in line with a passenger's waiting time. Castillo-Manzano (2010) reports that the likelihood of food/beverages being consumed and a purchase being made increases by 31 and 19 percent respectively if the waiting time exceeds 3 h while the amount spent increases by almost 41 percent. The demographic and social characteristics of passengers are found to be important predictors of spending behavior (Castillo-Manzano, 2010):
commercial real estate to the development of airports towards so called airport cities (e.g. Aldridge et al., 2001; Appold and Kasarda, 2013 Poungias, 2009; Reiss, 2007; Tacke, 2008). Several studies focus on the retail sector, for example the characteristics of airport retail, its development and potential for improvement (Giljohann-Farkas, 2008), contractual aspects (e.g. Freathy and O'Connell, 1999; Kim and Shin, 2001) and customer structures (e.g. Geuens et al., 2004). However, the majority of these studies is of descriptive character. Econometric studies in this topic area predominantly deal with retail drivers, in particular drivers of retail revenues in general (Volkova, 2009), determinants of sales (Appold and Kasarda, 2006) and food/beverage consumption and other purchases (Castillo-Manzano, 2010), shopping intentions (Lu, 2014), waiting time (Torres et al., 2005) or time pressure and impulse buying tendencies (Lin and Chen, 2013). The present study expands on these studies by incorporating a large range of potential explanatory factors that impact on revenues from retail and other commercial non-aviation activities of airports. Our analysis includes both standard factors used in previous analyses of this topic such as the number and characteristics of passengers as well as a variable that has not been tested before in this context: the terms of ownership. The results from this comprehensive quantitative analysis performed on a unique panel dataset of global airports contribute to the state of knowledge on the changing business strategies of airports. In particular, it sheds light on a number of key strategic considerations for airports, for example if expanding the commercial area inside a terminal is likely to reap financial rewards and whether it is more profitable to expand the area on the landside or airside area of a terminal. More broadly, we seek to contribute to the study of economies of scale in the non-aviation operations of airports by investigating the effect of different airport size measures on revenues and yields. Finally, a subset of the airports in this study is analyzed for the effect of privatization to establish if it has any bearing on the bargaining power of the airport when negotiating the fraction of sales to be retained by the airport. The following section reviews the existing literature in more depth to set the scene for the specification of our empirical panel data model of airport commercial revenues and feasibility.
families with children have a higher likelihood to make a purchase or to consume food/beverages, but this factor is also the greatest curb on the amount spent, elderly people have a less likelihood to make a purchase or to consume food/beverages, homemaker purchase rarely at an airport, people in a group consume more food/beverages on average than if they are alone, passenger being accompanied to an airport eat/drink something with the accompanying individual.
2. State of research There is a distinct gap in the literature on financial performance and economies of scale of airports, particularly with regard to nonaeronautical operations. Pels et al. (2003) perform an analysis of European airports and find that they are on average inefficient and operate on constant returns to scale when producing air transport movements and on increasing returns to scale when producing air passenger movements. However, their analysis is restricted to aeronautical operations only. Similarly, Martin and Voltes-Dorta (2010) perform an analysis of 41 global airports and find significant economies of scale. In theory, airports should be able to increase their profitability in the commercial sector with increasing size as the marginal cost of managing the commercial operations such as retail outlets and other facilities should fall. Fixed cost elements can be spread over a larger base of commercial operations. Further benefits may accrue from access to specialized labor and equipment which may not be available to smaller establishments. There may be additional size effects for airports which go beyond the definition of economies of scale in the narrow sense of the term. For example, some high-revenue facilities, for example an airport-based convention center, may only be feasible for airports above a certain size threshold. Retail operations may also be affected by airport size, in terms of the scope and depth of the retail products that are offered. However, Empirical evidence in the retail sector is rare. Arndt and Olsen (1975) study the Norwegian retail sector and find only limited evidence of economies of scale. The authors point out a number of limitations to economies of scale in retail, the most important among these being that retail is not a closed system such as manufacturing and retailers have very limited control over production and customer flows. Thus, economies of scale may be
A further theme of empirical studies of non-aviation revenues is the impact of the location and quality of the airport retail outlets and the role of the regulatory environment of airports (Del Chiappa et al., 2016; Forsyth, 2004). In particular, an evaluation of single and dual till models is undertaken in some studies but the discussion as to which approach is preferable remains controversial (e.g. Evangelinos et al., 2011; Kratzsch and Sieg, 2011; Zhang and Zhang, 2010). While the information on single or dual till regulation was not available in our global study of airports, the distinction between the two models may have ramifications for the commercial success of an airport. The single-till approach takes all areas of the airport including the non-aviation sector into consideration in determining the remuneration levels in the regulated aviation sector. By contrast, the dual-till approach typically separates areas of an airport which are necessary for the actual provision of services in air transport from the revenues (or income) from the commercial activities in the nonaviation sector. Non-aviation expenditure and costs incurred in this sector are then excluded from the regulation of charges in the aviation sector in the dual-till model. In a related study, Bel and Fageda (2009) find that the level of charges that airlines have to pay airports depends on the ownership structure (state-run, partially privatized, fully privatized) but their study does not include commercial revenues. A further strand of the literature is concerned with physical and network characteristics of an airport and their impact on commercial performance. For instance, Van Dender (2007) reports that concessions revenues per departing passenger are lower at hub airports in the United 2
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 1 Overview of selected studies on commercial revenues at airports and their main drivers (þ positive influence, - negative influence). Authors
Place
Data and Methodology
Main conclusions
Torres et al. (2005)
Spain
a non-parametric approach, 997 passengers at one airport
Appold and Kasarda (2006)
USA
OLS regression, 75 airports
Tovar and Martín-Cejas (2009)
Spain
parametric input Distance Func-tions (DEA), 26 airports
Graham (2014)
Worldwide
comparative assessment of data, general information about large number of airports
“The Moodie Report (2009)”
Worldwide
among others correlation analysis, 45 airports
Volkova (2009)
13 airports from EU countries
panel data econometric analysis, 13 airports
Castillo-Manzano (2010)
Spain
bivariate probit model, over 20,000 passengers, 7 airports
Fuerst et al. (2011)
45 airports from EU countries
regression analysis
vacationers spend more than business travelers (þ) length of stay prior to boarding effects the consumption in commercial area (þ) level of consumption is independent of the waiting time () number of passengers have the single largest effect on retail sales (þ) average distance flown has an effect on food and beverage sales (þ) average distance of traveling and amount of retail space have a positive effect on sales per passenger (þ) type of airport: international gateway status, major tourist destination and clearance time () passenger traffic not significant for sales per passenger () proportion of international passengers that circulating through the airport terminal (þ) airports with well-developed commercial activities are more efficient than those that focus on aeronautical revenues leisure charter passengers are good shoppers (þ) young leisure passengers who travel several times a year are high spenders (þ) LCC passengers are good users of food and beverage (þ) foreigners who have high duties and taxes buy more tax-free products (þ) women are more likely to fall into the category “shopping lovers” (þ) transfer passengers are less likely to make use of facilities and if they do, tend to spend less () international and leisure passengers spend more because of longer dwell time and availability of duty free (þ) large airports have more diverse, higher yielding commercial activities and to attract strong brands (þ) share of domestic passengers inconclusive (neutral) increasing floor space indirectly generates better returns but not higher sales per passenger (±) low-cost carrier passengers are not necessarily low spenders extra-EU passengers increase retail revenue per square meter at hub airports (þ) number of short stay parking places, check-in facilities and the number of employees contribute to the retail revenue (þ) retail revenue per square meter grows significantly once critical mass of space is reached (þ) bars/restaurants influence the retail revenue and generate externalities (þ) extra-EU passengers have no effect on retail revenue at regional airports (neutral) waiting time prior to embarking, being on vacation, being a frequent flyer and traveling with children (þ) number of group members and accompanied passengers explain the consumption of food/beverages (þ) passengers from outside Eurozone more likely to consume food/beverages, albeit just loose change (þ) passengers flying to international European destinations (þ) elderly people purchase less and spend less on food/beverages () passenger's arrival at the airport on a courtesy bus by the hotel () business passengers not likely to make last-minute purchases at the airport () passenger using a LCC have a lower likelihood to purchase food/beverages () main drivers of commercial revenues per passenger include the number of passengers passing through the airport, the ratio of commercial to total revenues, national income, the share of domestic and leisure travelers and the number of flights (þ) a large amount of retail space per passenger is generally associated with lower commercial revenues per square meter confirming decreasing marginal revenue effects () business travelers have a negative influence on commercial revenues per passenger ()
Source: own compilation
well as the willingness to pay for reductions in reduced ground travel time. Moreno and Muller (2003) demonstrate empirically for Sao Paolo that airport choice in multi-airport urban regions is influenced by these local accessibility considerations.
States. The size of an airport, typically proxied by passenger numbers and aircraft movements, the regional or local income of the area in which the airport is located as well as the size and configuration of the retail surface area are found to be key drivers of commercial revenue. Moreover, the present analysis includes additional factors which have not or rarely been investigated in studies on airport revenues, notably the terms of ownership (state-run, partially privatized, fully privatized). Recent empirical evidence (Adler et al., 2015) suggests that a move towards incentive regulation of airports is associated with increased productive efficiency. Finally, a number of studies consider the implications of airport accessibility and location in the urban or regional context. While outside of the scope of the present global study of airports, the distance, travel time or travel cost to the Central Business District (CBD) and other destinations has been considered in empirical case studies (Koster et al., 2011; Kouwenhoven, 2008; Reynolds-Feighan and McLay, 2006). These include trip timing decisions of travelers going to and from the airport as
3. Data and analytical framework Research into airport commercial revenues and profitability has been held back to date by severe constraints in the availability and quality of relevant data. The principal source of information for the present analysis is the Moodie Report which contains revenue and other financial performance data for a large number of global airports. The Moodie Report has been published since 2001. However, the format of the reporting was changed from the 2011 report onwards which no longer allows tracking of performance data for individual airports. Extracting the data from the Moodie Reports and supplementing the airport-level data with economic 3
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 2 Definitions and sources of variables. Variable
Definitions
Source
Commercial revenues
Income from retail and commercial activities (duty free, news and gifts, specialty retailing, food and beverage and currency exchange) Percentage of airport commercial revenue divided by gross commercial sales made by the concessions. It shows the percentage of total sales that is retained by the airport/airport operator. Gross domestic product per capita of the country where the respected airport is located in purchasing power parities The area, measured in square metres, devoted to commercial activities, it excludes passageways, cartilage areas or common seating areas in food court as well as facilities outside of passenger terminals Part of the terminal building after check-in, security, customs and passport control. This airside portion is typically only accessible to passengers. One form of commercial outlets, e.g. full-service and fast food restaurants Passengers departing or arriving at an airport Passengers traveling to an international destination (regardless of their nationality) Airport is partially privatized or owned by the government
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Commercial yield
GDP per capita Commercial area
Outlets airside Outlets food & beverage Passengers International travelers Partially privatized and staterun
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Datastream Moodies Study 2000/01, 2002/03, 2006/07, 2008/09
Moodies Study 2000/01, 2002/03, 2006/07, 2008/09 Moodies Study 2000/01, 2002/03, 2006/07, 2008/09 ICAOdata ICAOdata Phone/email/internet surveys of individual airports
of the retail offer, proxied by the percentage of shops not classified under the basic categories food and beverage, duty free and gift shops. A particularly strong relationship is detected between commercial revenues per passenger and the size of commercial floorspace. These potential agglomeration economies and economies of scale/scope will be tested more formally in the ensuing regression analysis. The commercial yield exhibits a similar, albeit weaker association with size and variety measures. The analytical strategy for estimating the impact of various drivers on the commercial performance of airports includes three performance metrics (commercial yield, total commercial revenues and commercial revenues per passenger) evaluated with three different estimation techniques (pooled OLS, random effects panel and 3 stage least squares). The basic pooled OLS single equation estimation of the commercial yield takes the following log-log form:
indicators yielded a panel dataset of global airport indicators for the period 1998–2009 (The Moodie Report, 2001, 2003, 2007 and 2009). Table 2 gives an overview of the variables employed in our panel data analysis. It is important to note that several airport operators do not provide a breakdown of revenues by individual airports. For example, the revenues reported for the British BAA Group comprise the three London airports Heathrow, Gatwick and Stansted. However, we were able to infer the volume of revenues for individual airports by using airport-level information provided by these operators such as fee and cost structures, the number of passengers etc. All monetary values were then converted into SDR (Special Drawing Rights), an international foreigm-exchange reserve asset created by the IMF to ensure comparability using the interbank exchange rate at the call date of the respective annual statements. Supplementary research and data quality control were conducted via telephone and email surveys with airport operators. The database used in this analysis includes information on 75 airports over a period of 11 years, albeit as an unbalanced panel with a considerable number of gaps due to limited data availability. Fig. 2 in the appendix displays the locations of the airports in this study. The selection of these airports was driven by data availability in the sources listed in Table 2. It cannot be ruled out that regression estimates of the entire population of global airports may differ from the ones reported below but, as far as we are aware, the current database represents the largest global database of its kind ever used in an empirical study. Table 3 examines different key characteristics of the dataset used in this analysis. As expected, standard deviations are relatively high for most variables, reflecting the broad range of values across the world. A comparison of the percentage for the food & beverage outlets in the four regions reveals that these outlets are more important in the Americas and Europe than in Asia Pacific and Africa. However, the average commercial area at airports in Africa, North America and Europe is similar, while Asia Pacific and South America have less commercial space. Interestingly, North America has the lowest commercial yield (the fraction of concession sales retained by the airport). This may be attributable to higher levels of competition among airports in the North American market. As airport size may be an important driver of commercial revenue, we examine the distribution by number of passengers and – as expected – the highest volume is found in the Americas. Finally, GDP per capita reveals a ranking as expected, North America before Europe, followed by Asia Pacific, South America and Africa. Next, we perform an exploratory analysis of the key financial performance metrics. Fig. 1 shows scatterplots of revenues per passenger and commercial yields with other variables. The first measure correlates significantly with size measures, i.e. the total area of commercial floorspace in an airport and the total number of retail outlets. A significant relationship is also found between revenue per passenger and the variety
ln CYi ¼ β0 þ β1 ln CATi þ β2 ln FBi þ β3 ln RTi þ β4 ln PIi þ β5 ln AIRi þ β6 ln CRPi þ β7 PRIVi þ β8 STATi þ εi
(1)
where CY is the commercial yield, CAT is the total commercial floorspace, FB is the fraction of food and beverage outlets, RT is total revenue in SDR, PI is the fraction of international passengers, AIR is the fraction of outlets that are located airside, CRP is the commercial revenue per passenger and PRIV and STAT are dummy variables indicating whether an airport is fully privatized, partially privatized or state-run. The estimation is conducted both as a pooled OLS and as a random-effects panel data analysis to ensure that the results are robust to modelling technique. Fixed effects results were also estimated to conduct a Hausman test with the results confirming the use of a random effects model. Analogous to Equation (1), we then estimate the equations for commercial revenues and commercial revenues per passenger. An obvious concern of our single-equation estimation is that the commercial performance indicators used as explanatory variables may be endogenous with the dependent variables. For example, commercial yields and revenues as well as revenues per passenger may be codetermined. Hence, we employ a more advanced 3-stage least squares (3SLS) technique following Zellner and Theil (1962). The 3SLS estimation models the endogenous variables in a system of equations, obtaining instrumental variable estimates conditional on the covariances of the error terms in each individual equation. Due to the endogeneity problem, the dependent variables will be correlated with the error terms but the 3SLS estimation takes this problem into account by taking the remaining explanatory variables as exogenous for obtaining unbiased estimates of the three endogenous variables (Greene, 2000). The specification of our 3SLS system is as follows:
4
Transport Policy xxx (2017) 1–9
ln CYi ¼ β9 þ β10 ln CATi þ β11 ln FBi þ β12 ln RTi þ β13 ln PIi 3 3656 1347 34 20,321 14,430 57 26,676 12,427 45 36,950 7240 3 5355 2251 144 31,277 13,803
GDP/capita (SDR)
F. Fuerst, S. Gross
þ β14 ln AIRi þ β15 ln CRPi þ φi
(2)
ln CRi ¼ β16 þ β17 ln CATi þ β18 ln FBi þ β19 ln Pi þ β20 ln PIi (3)
ln CRPi ¼ β23 þ β24 ln CATi þ β25 ln FBi þ β26 ln Pi þ β27 ln PIi 3 44.00% 2.59% 34 32.15% 32.02% 57 84.53% 18.54% 45 18.00% 25.29% 3 58.00% 3.06% 144 49.10% 34.68%
þ β28 ln GDPi þ β29 ln AIRi þ ξi
(4)
Apart from the variables employed in Equation (1), P is the total number of passengers and GDP is the GDP per capita of the airport's location as a proxy for economic development of the surrounding region.
3 12.40 4.46 34 17.40 14.20 57 17.10 29.00 45 20.60 23.50 3 0.66 0.14 144 17.90 23.90
4. Results & discussion
3 20.83% 5.97% 32 16.95% 7.92% 52 13.98% 7.41% 45 9.68% 9.65% 3 36.86% 4.87% 135 13.32% 9.00%
3 1.24 0.31 32 1.34 1.61 54 1.95 1.14 45 0.49 0.96 3 1.49 0.34 137 1.27 1.30
3 13,471 1845 33 8949 9367 55 10,474 22,046 45 12,060 9415 3 7160 2264 141 10,351 15,681
3 22.10% 4.80% 32 22.00% 10.53% 52 25.35% 9.12% 44 43.05% 12.19% 3 21.00% 4.75% 136 28.00% 13.36%
The first proposition to be tested empirically is whether the total volume of airport commercial revenues can be explained by the factors as hypothesized in the previous section. Note that a set of year dummy variables is used to control for unobserved time trends and longer term developments. The panel data estimation results of Models 1 and 2 demonstrate that larger airports (proxied by the number of passengers) have higher commercial revenues. While this result is unsurprising, it is interesting to note that even when the number of passengers is kept constant, we find a significant relationship between the square footage of commercial space in an airport and its commercial revenues, arguably because a larger total area of commercial floorspace in an airport tends to generate higher commercial income. The share of international passengers is also found to be a significant driver, implying that international travellers tend to have higher expenditure on commercial activities than domestic travellers. This finding is in line with most existing studies (with the exception of the recent study of German airports by Fasone et al., 2016). However, it is also possible that airports with a high percentage of international travellers have a larger proportion of high-revenue generating commercial activities such as high-end retail outlets selling international designer fashion, jewellery and watches which are predominantly found in international airports and to a lesser extent in airports serving the domestic market. Further analysis is required to elucidate the reasons for these findings. Since the airports in our data sample vary considerably in terms of their size, ranging from Riga Airport with just under 2 million passengers to Atlanta International Airport with approximately 90 million passengers we estimate the drivers of commercial revenue on a per passenger basis in a next step. Models 3 and 4 underline the importance of large commercial areas for generating higher revenues per passengers. A possible explanation for this may be found in economies of scale and economies of scope in larger commercial and retail operations. The retail outlets and other commercial activities may also create positive spillover effects and externalities for one another. Conversely, keeping the size of the commercial area constant, an increase in the total number of passengers appears to affect commercial revenues per passenger either insignificantly (Model 3) or even negatively (Model 4). Hence, revenues per passenger seem to be driven more by the supply of retail and other commercial outlets rather than by the number of passengers passing through the airport. This may also be taken as an indication of a shift in the role of an airport as a venue for commercial activities, entertainment and business rather than a pure infrastructure facility. This hypothesis is put forward most prominently by Kasarda (2006) and Appold and Kasarda (2006, 2013). While the estimation results in Models 1–4 show that total commercial revenue increases both with increasing commercial area and number of passengers, the negative coefficient of passengers on revenue per passenger hints at the fact that, after controlling for commercial square footage, an increase in the number of passengers might lead to adverse effects, for example through crowding, long wait times at
Total
South America
North America
Europe
Asia Pacific
N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. N Median Std. dev. Africa
Table 3 Summary statistics.
Statistics
Commercial yield (in %)
Commer-cial revenue (SDR)
Commercial area (sq m)
Food & beverage out-lets on total outlets (in %)
Passenger total (Mio.)
International passengers (%)
þ β21 ln GDPi þ μi
5
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Fig. 1. Scatterplot of commercial airport performance indicators. *p < 0.05.
A negative association is found with the level of total revenues, i.e. larger airports with higher total revenues appear to have lower commercial yields. Viewed in conjunction with the coefficient on commercial area, this might be suggestive of a space productivity effect. In other words, higher revenues in a given area of commercial space may entail a lower yield demanded by the airport. Less productive commercial spaces may have to pay a higher share of their revenues to the airport to meet the required rates. However, there could also be other reasons for this finding. For example, if the airport has a large passenger base, it can earn sufficient revenue from the aviation side to cover the cost and is not dependent on extracting a high share of commercial revenues from concessions and other commercial operations. The percentage of international passengers is positive and significant in Model 5 but not in Model 6. The percentage of outlets located airside (as opposed to landside) is insignificant for commercial yields in both estimations. Revenue per passenger is inversely related to commercial yields which suggests that airports demand a minimum share of turnover from concessions that does not increase linearly with higher revenue per passenger. This means that concessions with high per passenger revenues
the check-out, excessive noise etc, thereby reducing the willingness of passengers to frequent shops and make purchases. These effects may present diseconomies of scale, particularly if airport size fails to keep up with an increase in the number of passengers. However, more in-depth information, for example on the ratio of purchasing to non-purchasing passengers would be necessary to test this proposition empirically. It is also noteworthy that the economic development status as measured by GDP per capita of the country in which an airport is located is shown to have predictive power for the airport's commercial revenues both in terms of total volumes and on a per passenger basis. In the next step, we test the drivers of airport commercial yields (Models 5–7 of Table 4). While the yield as defined in our dataset is not a direct measure of the rate of return or profitability of airport commercial space in general, it affords us an indication of the bargaining power of an airport and, indirectly, of the potential productivity and profitability of a commercial space. We also find again that airports with a higher amount of commercial areas and floorspace tend to have higher commercial yields, underlining the positive impact of size and, potentially, higher bargaining power of larger airports. 6
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
Table 4 Estimation results (pooled OLS and random effects). Dependent V
(2) Commercial revenues (SDR million)
(1) Commercial revenues (SDR million)
(4) Commercial revenues (SDR) per passenger
(3) Commercial revenues (SDR) per passenger
(5) Commercial yield (in %)
(6) Commercial yield (in %)
(7) Commercial yield (in %)
Technique
Pooled OLS
Random Effects
Pooled OLS
Random Effects
Pooled OLS
Random Effects
Random Effects
Commercial area total (log) %Outlets food &beverage (log) Revenue total (log) %International passenger (log) %Outlets airside (log) Commercial revenue per passenger Passenger total (log) GDP/capita (log) Staterun airport Partially privatized airport Constant
0.323** (3.20)
0.465*** (4.69)
0.323** (3.20)
0.465*** (4.69)
0.0590** (3.18)
0.0542** (2.78)
0.00809 (0.39)
0.122 (1.00)
0.0170 (0.14)
0.122 (1.00)
0.0170 (0.14)
0.0813*** (-4.23)
0.0751*** (-3.90)
0.00747 (0.25)
0.260*** (6.71)
0.195*** (5.02)
0.260*** (6.71)
0.195*** (5.02)
0.0325* (-2.16) 0.0187* (2.35)
0.0310 (1.92) 0.0105 (1.31)
0.00676 (0.34) 0.0525** (3.01)
0.389*** (-4.36)
0.292** (-3.12)
0.389*** (-4.36)
0.292** (-3.12)
0.0144 (0.84) 0.0480* (-2.38)
0.0195 (1.09) 0.0315 (1.54)
0.0537 (1.78) 0.0613* (-2.18)
0.874*** (9.98) 0.245*** (4.26)
0.715*** (7.80) 0.203** (2.68)
0.126 (1.44) 0.245*** (4.26)
0.285** (-3.10) 0.203** (2.68) 0.114* (-2.18) 0.101* (-2.10)
1.975 (1.63)
0.655 (0.48)
1.975 (1.63)
0.655 (0.48)
0.253 (1.72)
0.234 (1.38)
0.162 (0.59)
Time controls (annual)
Yes
Yes
Yes
Yes
Yes
Yes
No
N R2 adj. R2 R2 within R2 between AIC Hausman test prob BIC
121 0.890 0.874
121 0.87
121 0.641 0.590
121 0.60
123 0.406 0.375
123 0.47
40 0.29
0.23 0.51
0.49 0.31
0.65 0.89 130.9
0.32 0.60 289.4
130.9 0.99
175.6
0.99
0.12 269.7
175.6
t statistics in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.
Table 5 Estimation results (3-stage least squares).
Commercial revenue per passenger
0.0427* (-2.57)
(10) Commercial revenue (SDR) per passenger
(9) Commercial revenue (SDR million)
(8) Commercial yield (in %) Commercial area total (log)
0.323*** (3.43)
%Outlets food &beverage (log)
0.122 (1.07)
Commercial area total (log) %Outlets food & beverage (log) Passenger total (log) %International passenger (log) GDP/capita (log) %Outlets airside (log)
0.927*** (3.94) 0.291* (-1.07) 0.467** (2.29) 0.292** (3.24)
Commercial area total (log) %Outlets food & beverage (log)
0.0602*** (3.49) 0.0590*** (-3.13)
Passenger total (log) %International passenger (log)
0.874*** (10.71) 0.260*** (7.20)
Revenue total (log) %International passenger (log) %Outlets airside (log)
0.0408* (-2.76) 0.0214** (3.09 0.0161 (1.00)
GDP/capita (log) %Outlets airside (log)
0.245*** (4.58) 0.389*** (-4.68)
Time controls (annual)
Yes
Yes
Yes
N R2
120 0.49
120 0.89
120 0.50
0.113 (1.32) 0.149 (0.72)
t statistics in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001.
that the size of the commercial area is a powerful determinant for all three financial performance metrics. Additionally, the fraction of international passengers is a significant predictor of commercial revenue and commercial revenue per passenger. A high percentage of food and beverage outlets in the terminal is associated with lower yields and lower commercial revenue per passenger. As compared to other more expensive items sold in airport outlets, the food and restaurants business normally incurs a lower profit margin which may, in turn, lower the percentage of sales revenues retained by the airport.
retain a larger fraction of their revenues, thus potentially increasing their profitability even more. In a final variation on model specifications, we test whether the operating structure of an airport has a significant impact on the size of the commercial yield. Model 7 includes binary variables for state run, partially and fully privatized airports. The results indicate that indeed privatized airports exhibit higher yields than their staterun or partially privatized counterparts but an important caveat of these findings is the much smaller sample size for which this information was available. Turning to the 3SLS estimation in Table 5, we model each of the dependent variables as endogenous in a system of equations to correct for any endogeneity bias introduced by the presence of financial metrics that are jointly and simultaneously determined. The results from this estimation are broadly similar to the single-equation estimation. We find
5. Conclusions This study set out to test the factors affecting the commercial performance of global airports. To this end, we analyzed a panel dataset of 7
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9
The analysis showed some, albeit inconsistent evidence of a negative effect of the share of food and beverage outlets on revenue per passenger and commercial yields (4 of the 10 model estimations confirm this effect). This may lead airports to conclude that this segment is best reduced and substituted with higher-value outlets in a strategy to increase per passenger revenue. However, this strategy is not necessarily warranted. Instead, it should be considered how passengers can be introduced to such offers or how these products can be made palatable in times of increasingly rare free catering services on board. Regionally sourced, seasonal, fresh and health-oriented offers are some currenttrends in gastronomy. Regarding airside/landside placement of shopping facilities, one possible approach may be to make more innovative offers accessible to passengers while they are waiting for boarding. Larger und pleasantly designed recreational areas, shopping and dining offers have been implemented at many airports along with technological enhancements, such as apps, mobile payment, online platforms, QR-Code-ShoppingWalls, touch screen monitors. A more in-depth study of the linkages between airport commercial revenues, local, national and global economies is required to ascertain these risk-mitigating effects. The present study was also limited by gaps in data coverage both in terms of the number of years available as well as the geographical distribution of the airports. For example, it was not possible to incorporate several of the most important global airports in our analysis because of a lack of data. Further research is also needed to corroborate the reported findings for a larger sample of global airports and a more thorough investigation of the financial performance metrics of commercial activities such as information on cash flows, return on equity, return on assets, stock market returns and liquidity measures would be desirable. The present analysis provides a foundation for these more fine-grained and expanded studies if and when the required datasets become available.
75 airports in 30 countries. Applying pooled OLS, random effects and 3SLS estimation techniques, the main drivers of financial performance were identified. Our results are broadly in line with the existing literature. The share of international passengers, the size of the commercial area as well as airport size and the mix and intra-terminal location of retail space are found to be significant determinants. The latter finding the presence of economies of scale in generating higher commercial revenue and ensuring profitability for an airport's non-aeronautical operations. Staterun and partially privatized airports appear to retain a significantly smaller share of concession sales than their privatized counterparts. Regarding the retail mix, a higher share of food and beverage outlets appears to decrease the revenue share retained by the airport while a higher percentage of outlets located airside rather than landside depresses commercial revenues. The corollary of these findings for airports seeking to enhance their financial performance in the non-aeronautical sector is that careful planning of the size and mix of retail and other commercial areas and their positioning in the airside/landside structure may be as important as the decision to privatise or remain state-run. Regarding airport management strategy, the results indicate that airports should seek measures to retain airlines with a large fraction of international passengers and, consequently, the higher spending level of these international passengers. On the other hand, it should be considered how domestic travelers can be encouraged to increase dwelling times and expenses. Customer loyalty programs that are already implemented in some airports are a possible strategy. In addition, the visitors may receive current promotions via Smartphone by entering the participating stores (Jegminat, 2015). Other examples are the CPH Advantage program at Copenhagen airport, the BAA WorldPoints of the British Airport Authority, the ViaMilano programme at Malpensa Airport or the Changi Rewards programme at Singapore's Changi airport.
Appendix
Fig. 2. Location of airports in this study.
Appold, S.J., Kasarda, J.D., 2013. The Airport City phenomenon: evidence from large US airports. Urban Stud. 6, 1239–1259. Arndt, J., Olsen, L., 1975. A research note on economies of scale in retailing. Swed. J. Econ. 207–221. Bel, G., Fageda, X., 2009. Privatization, Regulation and Airport Pricing: an Empirical Analysis for Europe. EUI Working Papers RSCAS 2009/27. Florence/Italy. Castillo-Manzano, J.I., 2010. Determinants of commercial revenues at airports: lessons learned from Spanish regional airports. Tour. Manag. 6, 788–796.
References Adler, N., Forsyth, P., Mueller, J., Niemeier, H.M., 2015. An economic assessment of airport incentive regulation. Transp. Policy 41, 5–15. Aldridge, K., Carreno, M., Ison, S., Rye, T., Straker, I., 2001. Car parking management at airports: a special case? Transp. Policy 13, 511–521. Appold, S.J., Kasarda, J.D., 2006. The appropriate scale of US airport retail activities. J. Air Transp. Manag. 12, 277–287.
8
F. Fuerst, S. Gross
Transport Policy xxx (2017) 1–9 Lu, J.L., 2014. Investigating factors that influence passengers' shopping intentions at airports – evidence from Taiwan. J. Air Transp. Manag. 35, 72–77. Martin, J.C., Voltes-Dorta, A., 2010. International airports: economies of scale and marginal costs. J. Transp. Res. Forum 47 (1). Moreno, M., Muller, C., 2003. The influence of ground accessibility on airport choice in Sao Paulo metropolitan area. In: Proceedings 19th Dresden Conference of Traffic and Transportation Sciences, 22.-23.09.2003. Dresden. Pels, E., Nijkamp, P., Rietveld, P., 2003. Inefficiencies and scale economies of European airport operations. Transp. Res. Part E Logist. Transp. Rev. 39 (5), 341–361. Poungias, P., 2009. Airport city developments: an airport investor's perspective. J. Airpt. Manag. 4, 14–22. Qin, J., Olaru, D., 2013. System Dynamcis based simulation for airport revenue analysis. In: Conference Proceedings 31st International Conference of the System Dynamics Society, vol. 4. Curran Associates, Inc., Red Hook, NY, pp. 1–23. Reiss, B., 2007. Maximising non-aviation revenue for airports: developing airport cities to optimise real estate and capitalise on land development opportunities. J. Airpt. Manag. 3, 284–293. Reynolds-Feighan, A., McLay, P., 2006. Accessibility and attractiveness of European airports: a simple small community perspective. J. Air Transp. Manag. 12, 313–323. Tacke, G., 2008. Price optimization for airport parking. In: Proceedings 11th Hamburg Aviation Conference 13.-15.02.2008 (CD-rom). Hamburg. The Moodie Report, 2001. The Airport Commercial Revenue Study – 2000/2001. Brentford. The Moodie Report, 2003. The Airport Commercial Revenue Study – 2002/2003. Brentford. The Moodie Report, 2007. The Airport Retail Study – Fourth Edition 2006/07. Brentford. The Moodie Report, 2009. The Airport Commercial Revenues Study – Fifth Edition 2008/ 09. Brentford. Torres, E., Domínguez, J.S., Valdes, L., Aza, R., 2005. Passenger waiting time in an airport and expenditure carried out in the commercial area. J. Air Transp. Manag. 6, 363–367. Tovar, B., Martín-Cejas, R.R., 2009. Are outsourcing and non-aeronautical revenues important drivers in the efficiency of Spanish airports? J. Air Transp. Manag. 5, 217–220. Van Dender, K., 2007. Determinants of fares and operating revenues at US airports. J. Urban Econ. 62, 317–336. Volkova, N., 2009. Determinants of Retail Revenue for Today's Airports. GAP Project Working Paper, Berlin. Zellner, A., Theil, H., 1962. Three stage least squares: simultaneous estimate of simultaneous equations. Econometrica 29, 54–78. Zhang, A., Zhang, Y., 2010. Airport capacity and congestion pricing with both aeronautical and commercial operations. Transp. Res. Part B Methodol. 44 (3), 404–413.
Del Chiappa, G., Martin, J.C., Roman, C., 2016. Service quality of airports' food and beverage retailers. A fuzzy approach. J. Air Transp. Manag. 53, 105–113. Deloitte, 2009. Europ€ aische Flughafenbetreiber – Hohe Profitabilit€atsniveaus sind nur eine Seite der Medaille, p. 2009. Evangelinos, C., Püschel, R., Goldhahn, S., 2011. Inverting the Regulatory Rules? Optimizing Airport Regulation to Account for Commercial Revenues. Diskussionsbeitr€ age aus dem Institut für Wirtschaft und Verkehr 01/2011. Fasone, V., Kofler, L., Scuderi, R., 2016. Business performance of airports: non-aviation revenues and their determinants. J. Air Transp. Manag. 53, 35–45. Forsyth, P., 2004. Locational and monopoly rents at airports: creating them and shifting them. J. Air Transp. Manag. 10, 51–60. Fraport, A.G., 2013. Heinemann Duty Free und Fraport machen mobil, „Virtual Shopping“ erstmals am Frankfurter Airport. Retrieved from, 20.10.2016. http://www.fraport. de/de/presse/newsroom/service-news/2013/heinemann-duty-free-und-fraportmachen-mobil.html. Fraport, A.G., 2015. Online Shopping at Frankfurt Airport: Products from Airport Shops Available via the FRA App and Website. Retrieved from, 20.10.2016. http://www. fraport.com/en/press-center/reports–-publications-/archive/2015/online-shoppingat-fra-2015.html. Freathy, P., O'Connell, F., 1999. A typology of European airport retailing. Serv. Ind J. 3, 119–134. Fuerst, F., Gross, S., Klose, U., 2011. The sky is the limit? The determinants and constraints of European airports commercial revenues. J. Air Transp. Manag. 17, 278–283. Geuens, M., Vantomme, D., Brengman, M., 2004. Developing a typology of airport shoppers. Tour. Manag. 25, 615–622. Giljohann-Farkas, K., 2008. Airport Retailing, DEGI Research. December 2008. Graham, A., 2014. Managing Airports – an International Perspective, fourth ed. New York. Greene, W.H., 2000. Econometric Analysis, fourth ed. Upper Saddler River, New Jersey. Jegminat, G., 2015. Flughafen Frankfurt – Online-Portal für Terminal-Shops. Retrieved from, 25.11.2016. http://www.fvw.de/flughafen-frankfurt-online-portal-fuerterminal-shops/393/150887/18663. Kasarda, J.D., 2006. The rise of the aerotropolis. Next Am. City 10, 35–37. Kim, H.-B., Shin, J.-H., 2001. A contextual investigation of the operation and management of airport concessions. Tour. Manag. 22, 149–155. Koster, P., Kroes, E., Verhoef, E., 2011. Travel time variability and airport accessibility. Transp. Res. Part B Methodol. 10, 1545–1559. Kouwenhoven, M., 2008. The Role of Accessibility in Passengers' Choice of Airports. Discussion Paper No. 2008-14. The Hague/Netherlands. Kratzsch, U., Sieg, G., 2011. Non-aviation revenues and their implications for airport regulation. Transp. Res. Part E 47, 755–763. Lin, Y.-H., Chen, C.-F., 2013. Passengers' shopping motivations and commercial activities at airports – the moderating effects of time pressure and impulse buying tendency. Tour. Manag. 36, 426–434.
9