Assessing the efficiency of airports considering health and safety issues: A pilot study for Turkey

JAIRM, 2022 – 12(2), Online ISSN: 2014-4806 – Print ISSN: 2014-4865 https://doi.org/10.3926/jairm.288 Assessing the efficiency of airports considering health and safety issues: A pilot study for Turkey Burcu Ustael 1, Berna Ulutas 2 1Eskisehir Osmangazi University (Turkey) 2Department of Industrial Engineering, Faculty of Engineering and Architecture, Eskisehir Osmangazi University (Turkey) [email protected], [email protected] Received October, 2022 Accepted November, 2022 Abstract Purpose: Airport management is influenced by several related critical decisions on financial and technical operations. Optimum utilization of resources, including assets and personnel, is critical to achieve better service quality for the passengers and an efficient airport. Due to its importance, this paper aims to assess the performance of airports based on security, safety, and possible work-related health problems by considering the uncertain and unclear number of passengers and their baggage. Design/methodology: The efficiency of 30 airports in Turkey is evaluated with a basic Data Envelopment Analysis (DEA) model with two inputs and five outputs. Then, the model is redefined with an undesired output. High, medium, and low seasons are defined in the second DEA model to estimate the number of passenger baggage. The third model utilizes the principles of Fuzzy DEA (F-DEA) that aims to handle the uncertainty for the undesired output data. Findings: The results of three models confirm that the number and weight of baggage and consequently health and safety issues in airports should not be overlooked when optimizing airport efficiency. Utilizing the fuzzy theory has the potential to help managers to improve the operational efficiency of airports when dealing with an uncertain number of passengers and estimating the workload of baggage handlers. -49- Journal of Airline and Airport Management 12(2), 49-68 Research limitations/implications: No permission was given to make interviews with the ground handling personnel and gather real-life data to analyze task durations and workers’ body movements. Practical implications: Inputs, outputs, and undesired output defined in this study can be used to assess the airports in any other country. Social implications: The importance of health and safety issues for passengers, airport personnel, baggage handlers, and the residents who live close to the airports is considered. Originality/value: This study contributes to the airport performance assessment literature by considering the uncertain and dynamic data related to health and safety issues. This pioneering study, up-to-best knowledge, is the first to assess the airports in Turkey by DEA with the defined undesired output (baggage handler workload) and also utilizing the fuzzy model for the uncertain data. Keywords: Data Envelopment Analysis (DEA), Fuzzy DEA (F-DEA), airport, baggage handling, efficiency, health and safety To cite this article: Ustael, B., & Ulutas, B. (2022). Assessing the efficiency of airports considering health and safety issues: A pilot study for Turkey. Journal of Airline and Airport Management, 12(2), 49-68. https://doi.org/10.3926/jairm.288 1. Introduction Providing reliable and quick transportation, the use of airlines has gained more importance within the last century. Airports involve high expenditures related with traffic control, terminals, and runways to provide the aircraft operations, cargo handling and passenger movements. Therefore, analyzing the efficiency of airports to utilize the resources and infrastructure in air transport industry is very important. A passenger may expect low cost and high-quality services when choosing the air transport. On the other hand, airport personnel may expect a good salary and healthy working environment. The prerequisites for airport health and safety are defined as international standards. To meet the expectations of passengers and personnel the operations related with aircraft (passenger embarkation and disembarkation, aircraft maintenance, catering, cleaning etc.) and baggage handling (ground handling, security, check-in desk, etc.) must be planned in detail. Li et al. (2022) state that Covid-19 has influenced the regulations and operations at airports. The air service quality has been investigated with the keywords: access (ground transportation, parking), check-in service, security, wayfinding (signs, directions, flight information, mobility), arrival (passport control, customs, arrival services, baggage claim), facilities (food, beverage, washrooms, shopping, wi-fi access, waiting area, environment (accessibility, air quality, noise, aesthetic, personnel, and service. It is concluded that the ranking for access, wayfinding, facilities, and environment differ before and after Covid-19. Tabares (2021) focuses on the health issue and proposes screening locations in the airport terminal building to enable screening of passengers, crews, and airport workers. Sivakumar (2022) points the low frequency but high severity risks at airport operations. Based on an aviation fuel leak risk example, environmental risk (contamination of soil and water), occupational risk (vomiting, nausea, and slips), and operational risk (aircraft fire) are defined and the proposed integrated risk assessment is stated to be helpful in identifying complex risks and managing them. The efficiency of an airport depends on several related factors. By increasing the security, possible terrorist attacks can be minimized, by optimizing the number of emergency equipment, the damage caused from any -50- Journal of Airline and Airport Management 12(2), 49-68 disaster like fire can be minimized. On the other hand, by providing healthy working conditions and tracking the workload of workers may help to minimize work related musculoskeletal diseases (WMSD) and absenteeism. Several reports confirm the dramatically decrease in the number of qualified ground handling personnel after the Covid-19 pandemic. The baggage handlers take an important role to achieve efficient airport operations and high service quality. Because any problem in ground handling may cause a delay for a connecting flight that leads to high penalty costs and require rescheduling. Due to its importance, airport ground service management that is related to safety, timing, and efficiency is considered in this study. The literature on airport efficiency incorporates several important issues, such as: airport efficiency (their determinants, methods, benchmarking studies), service quality (passenger's perceptions and satisfaction, level of service assessment, simulation models of airport operations), safety performance, security issues, economicfinancial aspects (impact of non-aeronautical revenues, sustainability), and environmental issues (undesirable outputs of the airport processes (Bezerra & Gomes, 2018). This study aims to assess the efficiency of airports based on the security, safety, and possible work-related health problems by considering the uncertain and unclear number of passengers and their baggage. There are several contributions of this research paper. The input and output variables in DEA are defined as to evaluate security, safety, and health issues at airports. A model that is based on fuzzy set theory is proposed to handle the imprecise data. The airports in Turkey are considered and the efficiency scores are discussed based on the efficiency scores. In Section 2, a summary of DEA literature for airport performance assessment and F-DEA is provided. Then, the airports selected for this study, the details of the input, output, undesired output, and DEA models are explained. In Section 3, the results from the DEA models are provided. Discussions are given in Section 4. The final section concludes the study and gives suggestions for future research. 2. Methodology 2.1. Literature review for airport assessment with DEA DEA is one of the performance analysis techniques to evaluate the relative efficiency of homogenous decisionmaking units (DMU) such as schools, hospitals, airports, etc. There are a notable number of studies concerning airport efficiency. In the review study, the common performance indicators for airports are summarized as economic, operational, and environmental performance Graham (2005). Lai et al. (2012) defined the most popular input variables as airport service factors (number of employees, size of terminal area, number of runways, number of gates, size of apron, number of check-in desks, length of runway, number of parking spots, number of collection belts, and number of aprons) and financial factors (operational cost, capital cost, labor cost, and amount of capital stock). The output variables were grouped under airport service factors (number of passengers, aircraft movement, amount of cargo) and financial factors (amount of non-aeronautical revenue, amount of aeronautical revenue, amount of operational revenue, amount of nonoperational revenue). Iyer and Jain (2019) considered the studies published during 2009-2017 that focus on the performance of airports and the input and output variables, used at least in three papers, were summarized. The inputs were defined as capital assets, capital invested, labor cost, material cost, operating cost, soft costs, annual capacity of terminal, apron area, apron stands, baggage collection belts, boarding gates, check-in counters, dynamic apron capacity, full time equivalent employees, maximum throughput capacity, runway area, runway length, runways, scheduled routes, terminal area, and total airport area. Outputs were defined as aeronautical revenue, commercial revenue, ATM, cargo, mail, passengers, and workload unit. The papers that utilized DEA, Bootstrap DEA, Network DEA, DEA discrete regression and TOBIT to assess performance of airports in several countries can be summarized as; Turkey (Kocak, 2011), Italy (Curi et al., 2011), Greece (Tsekeris, 2011), Brasil (Wanke, 2012), France (Barros, 2013), Spain (Coto-Millan et al., 2014), Poland (Augustyniak et al., 2015), Spain and Turkey (Ulku, 2015), USA (Zou et al., 2015), Spain (Coto-Millan, 2016), East Asia (Liu, 2016), Pakistan (Ennen and Batool, 2018), Turkey (Keskin and Koksal, 2019), Germany (Stichhauerova and Pelloneova, 2019), and Europe and Asia-Pacific regions (Chaouk et al., 2020). -51- Journal of Airline and Airport Management 12(2), 49-68 The desirable output variable in a DEA model requires to be increased and undesired output to be decreased. Dyckhoff and Allen (2001) introduce three main approaches to handle the undesirable output variable(s) in a DEA model. 1. The undesired output can be considered as being desirable by using the reciprocal of the undesirable output as a DEA output variable. 2. Depending on the operational scale of the DMUs, the undesired output can be considered as an input variable in CCR or BCC DEA models. 3. For BCC and additive DEA models, it is possible to make a value translation for each DMU by adding to the reciprocal additive transformation of the undesired output a positive scalar, big enough, so that the final values are positive. To assess the efficiency of 22 airports in Turkey, Ulutas (2018) defined the green gas emission data as an undesirable output that was considered as an input variable in the DEA model. 2.2. Literature review for F-DEA and airport assessment applications The traditional DEA approach uses the efficiency frontier generated by inputs and outputs of the DMUs to calculate the efficiency and can be named as a precise data-based approach. The F-DEA was first introduced by Sengupta (1992) and used for cases where the observed values for input and output data of DMUs are imprecise and uncertain. Kao and Liu (2000) presented F-DEA model under Variable Return to Scale (VRS) and then converted the fuzzy model into a deterministic model using the α-cut approach. F-DEA is categorized by Emrouznejad et al. (2014) as tolerance, α-level based, fuzzy ranking, possibility, fuzzy arithmetic, and fuzzy random/type-2 fuzzy set approaches. Wanke et al. (2016) used F-DEA model for the weights related to the inputs and outputs of DMUs for a airport efficiency study. In a recent paper, Guner et al. (2021) focused on sustainability and proposed Fuzzy Double-Frontier Network DEA (FDFNDEA) assess the relationship between outputs (desirable and undesirable) related to use of infrastructure, fuel consumption, and movements. 2.3. Defining decision making units The Turkish State Airports Authority Directorate General (TSAADG) is a public-enterprise company connected to the Ministry of Transport. There are currently 56 airports operating in Turkey. In this study, 30 airports with domestic and international flights are defined as DMUs. Based on the annual passenger numbers, the airports are ranked from highest to lowest. The domestic airports that have lower than 200000 annual passengers are not included in the study. On the other hand, Istanbul Ataturk, Ankara Esenboga, Izmir Adnan Menderes, and Antalya Airports with very high number of passengers per year are not considered to enable the homogeneity of the DMUs. Related statistics are not available for Istanbul Airport, currently the largest airport in Turkey that begun to operate on 29 October 2020. Zonguldak Caycuma, Gazipasa Alanya, Zafer, and Aydin Cildir airports are operated by private organizations regulated by State Airports Authority Directorate General, Istanbul Sabiha Gökçen Airport are operated by a private organization regulated by Defence Industry Department, Eskisehir Hasan Polatkan Airport is operated by Eskisehir Technical University. Therefore, these airports were not defined as DMUs. 2.4. Defining inputs and outputs The domestic and international passenger data for the DMUs in concern are obtained from the TSAADG statistical annuals that is the only reliable source. Defining proper inputs and outputs are critical for assessing the performance of DMUs. When defining the input, output, and undesired output variables, accessing available and trusted data is important. Being different from other airport assessment with DEA, this study utilizes the data related with passengers, airport personnel, baggage handlers, and even people living near the airport. Main concern is to evaluate possible effect of health and safety issues to the airport efficiency. -52- Journal of Airline and Airport Management 12(2), 49-68 The Communication Navigation Surveillance (CNS) Services Safety Management System in Turkish airports was put into use on September 1, 2014. The safety policy in based on the “Guidelines for the Use of Safety Management Systems by Air Navigation Service Providers (SHT 65-03)” and the requirements specified by the Air Navigation Service Provider’s Safety Management, Safety Responsibility, Safety Priority, and Air Traffic Management (ATM) Services’ Safety Objectives. The staff who work at the Turkish airports are authorized by the Air Traffic Safety Electronics Personnel (ATSEP) license approved by the General Directorate of Civil Aviation to establish the safety. The typical inputs that are considered are the number of personnel and terminal area. Outputs and undesired output defined for this study is given in Table 1. Inputs Outputs Undesired outputs I.1.Number of personnel (person/year) I.2.Terminal area (m2) O.1.Revenue (1.000 TL/year) O.2.Number of x-ray O.3.Number of check-in counters O.4.Number of fire equipment O.5.Distance to the city center (km) UO.1.Load per baggage handler (kg/ baggage handler) Table 1. The inputs, outputs, and undesired output defined in this study I.1. Number of personnel (person/year) The data for operational and ground service personnel who work at the airport is named as the number of personnel. The airports in concern have domestic and international terminals. To enable high service quality, optimum number of personnel for domestic and international terminals should be assigned. I.2. Terminal area (m2) The terminal area is the place where passengers spend time waiting for the departure. Several DEA studies related with airport performance pay attention to identify the ideal terminal area. Larger terminal area may cause problems related to maintenance and security. On the other hand, considering current health issues related with Covid-19, the terminal area should be large enough to enable social distance between the passenger and airport personnel. O.1. Revenue (1.000 TL/year) Air transport management typically aims to maximize the revenue. The annual revenue is based on Civil Aviation Activities such as air traffic control, ground, and terminal services and other profits. O.2. Number of x-ray The x-rays in airports enable to check the baggages for any illegal or dangerous items before the airplane takes off. Due to the increasing security requirements, using the optimal number of x-rays are critical. O.3. The number of check-in counters Check-in counters in an airport are defined as the desks where passengers get their boarding cards and drop their baggage. During these operations, the long ques in the peak time periods or hours may cause congestion and -53- Journal of Airline and Airport Management 12(2), 49-68 dissatisfaction for the passengers. Therefore, the number of check-in counters are defined as an output variable that needs to be maximized. O.4.Number of fire equipment: In a possible catastrophic event, the number of fire equipment and related emergency response equipment (ambulance, rescue, etc.) are vital to preserve airport assets and save lives of the passengers, employees, and the baggage handlers who are at risk in the airport terminal building. Specifically, the number of equipment can also be critical during the fire extinguishing of any wild forest fire close to the airport. In this case, the residents can directly or indirectly be affected. O.5.Distance to the city center (km) The construction of an airport relies on a detailed long term investment plan. The site location of the airport is one of the decisions that is related with land costs, easy access, and several other factors. As the residence areas expand, the distance from the city center to the airport can change within decades. The maximum noise levels generated by aircraft was defined by a legislation and airport and airline operators develop local traffic regulations. Based on the airport noise map pilot project by the Scientific and Technological Research Council of Turkey Marmara Research Center (TUBITAK MAM) and Environment and Clean Production Institute, noise monitors have been established and data is being analyzed. However, no corrective measurement system is introduced yet. It is not possible to relocate an airport. However, there are studies related to rescheduling the flights in certain hours to reduce the effect of noise for the residents living close the airports during the plane take-off and landing. Therefore, the distance of the airport to city center are defined as one of the outputs. UO.1.Load per baggage handler (kg/baggage handler) The baggage handling operations are related with the number of passengers and the weight limitations that depend on the flight destination (international, domestic). The lost baggage or late transfer of baggage can lead to critical problems related with flight schedules in airports or high penalty costs for the airlines. The baggage screeners and handlers who are involved in the baggage handling operations at airports perform manual lifting and/or lowering tasks that can be considered as the main risk factor for work related musculoskeletal disorders (WMSDs). In this study, the load per baggage handler is defined as an undesired output that needs to be minimized. 2.5. Definition of DEA models This study aims to assess the efficiency of Turkish airports by using the available data from the TSAADG statistical annuals. It is apparent that the number of domestic and international passengers are not constant within time. The random changes may depend on season, travel promotions, or several other factors. The number and weight of baggage that needs to be handled to the aircraft is highly correlated with the number of passengers. Three DEA models are defined to identify the possible effect of health and safety issues to the airport efficiency. The efficiency scores of airports are separately obtained from domestic and international passenger data. Figure 1 summarizes the general structure of the study. -54- Journal of Airline and Airport Management 12(2), 49-68 Figure 1. The general structure of the study The DEA models, named as DEA-I, DEA-II, and DEA-III, considers two inputs (number of personnel and terminal area) and five outputs (number of check-in counters, annual income, number of x-ray equipment, number of firefighting equipment, and the distance of the airport to the city center). The undesirable output (load per baggage handler) is considered only in DEA-II and DEA-III. DEA-I model definition DEA-I is modeled based on two inputs and five outputs and the data are obtained from TSAADG statistic annuals. The descriptive statistics for international terminal variables are given as an example in Table 2. Variables Min Max Mean SD Inputs Number of personnel 81.00 165.20 392.00 79.87 Terminal area 3460.00 61078.67 1255540.00 226144.18 Outputs Number of x-ray 3.00 13.73 37.00 6.40 Number of check-in counters 3.00 16.13 87.00 14.63 Number of fire equipment 8.00 10.77 15.00 1.76 Distance to the city center 5.30 58.88 723.00 132.76 Revenue 3585.00 54783.57 323319.00 92439.50 Table 2. Descriptive statistics for the inputs and outputs for DEA-I DEA-II model definition DEA-II is modeled based on two inputs, five outputs, and one undesired output. The undesired output data is generated considering the passenger data for each month and assumptions on the baggage weight limits. The regular baggage limit for domestic flights is considered as 15 kg and international flights as 25 kg. It is assumed that no passenger has overweight baggage (over 32 kg). To represent possible passenger intensity, the time periods (seasons) are defined as high (June, July, August, September), average (October, November, December, January), and low (February, March, April, May). -55- DMUs Data Model Results 56 airports Inputs Outputs DEA-I Efficiency scores for domestic flights -no data -private Inputs Outputs Undesirable output Defining seasons DEA-II Efficiency scores for international flights 30 airports Inputs Outputs Undesirable output Defining fuzzy numbers DEA-III Journal of Airline and Airport Management 12(2), 49-68 Nine different scenarios are defined in DEA-II model to estimate the baggage handlers workload that is considered as undesired output. Table 3 summarizes the multipliers to generate different scenarios. For example, S1 utilizes the total number of passengers for an airport in high season and assumes that 10% of the passenger check-in their baggage. Likewise, in S5, 60% of the passenger travelling in an average season check-in their baggage. The workload of a baggage handler is expected to be higher when 90% of the passenger check-in baggage in any season. Scenario no Season definition The multiplier for baggage check-in S1 High 0.10 S2 High 0.60 S3 High 0.90 S4 Average 0.10 S5 Average 0.60 S6 Average 0.90 S7 Low 0.10 S8 Low 0.60 S9 Low 0.90 Table 3. Scenario definitions based on passenger check-in baggage estimations To estimate the data for undesired output in DEA-II model, the number of passengers for each airport (DMU) in each month is obtained from statistical annual. First, the total number of passengers are calculated based on the high, average, and low season definition. Then, the multipliers are used to estimate number of check-in baggage. It is assumed that maximum weight of a baggage can be 25 kg for an international flight. Possible load per baggage handler (UO) is calculated by multiplying the estimated check-in baggage number with 25. An example for DMU1 utilizing the international passenger data is provided in Table 4. June July Aug. Sept. 70380 82879 85809 71088 Oct. Nov. Dec. Jan. 63404 54457 51805 50181 Feb. Mar. April. May 45014 53590 66550 56600 Higher season 310156 Average season 219847 Low season 221754 X 0.10 X 0.60 X 0.90 X 0.10 X 0.60 X 0.90 X 0.10 X 0.60 X 0.90 S1 S2 S3 S4 S5 S6 S7 S8 S9 31015.6 186093.6 279140.4 219140.7 131908.2 197862.3 22175.4 133052.4 199578.6 UO-S1 UO-S2 UO-S3 UO-S4 UO-S5 UO-S6 UO-S7 UO-S8 UO-S9 193847.5 1163085 1744627.5 137404.3 7824426.2 1236639.3 138596.2 831577.5 1247366.2 Table 4. An example of undesired output data estimation for DMU1 with international passenger data For each DMU in concern, the calculation steps (represented in Table 4) is repeated. When considering the domestic passenger data, S1 value is multiplied by 15. In this study, since it is aimed to minimize the load per baggage handler, the undesirable output was considered as an input in the DEA-II model. -56- Journal of Airline and Airport Management 12(2), 49-68 DEA-III model definition DEA-III is modeled based on two inputs, five outputs, and one undesired output. This model is different from DEA-II in the data estimation for undesired output. The uncertainty in the number of passenger baggage and their weight is considered by a fuzzy DEA model. The uncertain data are assumed as fuzzy parameters expressed by a triangular fuzzy membership function because exactly estimation for the number and weight of each passenger baggage may not be possible. To handle this uncertainty, the number of passengers travelling from an airport is considered and the average of passengers (considering data for each month) are calculated. Then, the data are modeled as triangular fuzzy numbers. An example of triangular membership function is represented in Figure 2. Figure 2. Representation of triangular fuzzy function By considering the domestic passenger data for DMU1, first the average (a2) number of passengers is calculated. The maximal (a3) and minimal (a1) values are determined by an offset of 20% from the mean values, to represent the UO value (2313466.267, 2891832.833, 3470199.4). This calculation is done for each DMU in concern. 3. Results The performance assessment for the DMUs in concern are calculated by using R studio deaR package. The DEA-I results are obtained from CCR DEA model and summarized in Table 5. Based on domestic terminal data, Adana, Adiyaman, Bingol, Diyarbakir, Kapadokya, Kayseri, Mardin, M.Dalaman, M.Milas-Bodrum, Mus, Ordu-Giresun, S.Urfa-Gap, and Trabzon are the efficient airports that corresponds to the 13 out of 30 airports. The results for the international terminal data is quite similar where 14 airports are identified as efficient (Adana, Adiyaman, Bingol, Diyarbakir, Igdir, Kayseri, Mardin, M.Dalaman, M.Milas-Bodrum, Mus, Ordu-Giresun, Sivas, S.Urfa-Gap, and Trabzon). DEA-II results for domestic flights are obtained by considering the same, inputs, and outputs defined in DEA-I model. Table 6 illustrates the results for DEA-II obtained from domestic flight data. The last column represents the average efficiency scores for the nine scenarios. Among the 30 domestic airports, 16 airports are identified as efficient. Five domestic airports (Elazig, Erzurum, K.Maras, Konya, and Sivas) have different efficiency scores for the scenarios in concern. This represents that the number of passengers is not consistent in certain seasons. Erzurum is identified as the domestic airport with the lowest average efficiency score (0.610). -57- 1 0.5 0 a1a2a3 Journal of Airline and Airport Management 12(2), 49-68 Domestic International DMUs DEA-I DEA-II (avg) DEA-III (best case α=1) Average DEA-I DEA-II (avg) DEA-III (best case α=1) Average Adana 1 1 1 11 1 1 1 Adiyaman 1 1 1 11 1 1 1 Balikesir 0.851 1 1 0.950 0.865 1 1 0.955 Batman 0.943 0.950 0.951 1 1 0.984 Bingol 1 1 1 11 1 1 1 Bursa 0.705 1 1 0.902 0.806 0.828 1 0.878 Denizli 0.967 1 1 0.989 0.997 0.997 0.998 0.977 Diyarbakir 1 1 1 11 1 1 1 Elazig 0.630 0.644 0.650 0.641 0.648 0.700 0.732 0.693 Erzincan 0.874 0.878 0.881 0.878 0.903 .0981 1 0.961 Erzurum 0.557 0.610 0.862 0.676 0.572 0.890 1 0.821 Gaziantep 0.613 0.614 0.665 0.631 0.669 0.777 0.807 0.751 Hatay 0.903 0.903 0.905 0.904 0.906 0.912 0.941 0.920 Igdir 0.956 1 1 0.985 1 1 1 1 K.Maras 0.670 0.850 0.897 0.848 0.844 0.991 1 0.872 Kapadokya 1 1 1 10.949 0.974 1 0.974 Kars 0.834 0.835 0.874 0.848 0.844 0.994 1 0.946 Kayseri 1 1 1 11 1 1 1 Konya 0.664 0.676 .0714 0.685 0.673 0.688 0.682 0.681 Malatya 0.831 0.832 0.916 0.860 0.798 0.917 1 0.905 Mardin 1 1 1 11 1 1 1 M.Dalaman 1 1 1 11 1 1 1 M.Milas- Bodrum 1 1 1 11 1 1 1 Mus 1 1 1 11 1 1 1 Ordu-Giresun 1 1 1 11 1 1 1 Samsun 0.855 0.855 1 0.903 0.800 0.877 1 0.892 Sivas 0.800 0.817 0.824 0.814 1 0.831 0.871 0.901 S.Urfa-Gap 1 1 1 11 1 1 1 Trabzon 1 0.891 1 0.964 1 1 1 1 Van 0.792 0.792 0.891 0.825 0.799 0.988 1 0.929 Efficient DMU% 40 53.33 60 40 53.33 80 Table 10. Overall efficiency results for the airports in concern -64- Journal of Airline and Airport Management 12(2), 49-68 The results obtained in this study is in line with Fernandez et al. (2018) that state tourist-oriented airports to achieve higher efficiency levels than non-touristic ones. The undesired output variable was defined as load per baggage handler. Tapley and Riley (2005) confirm the WMSDs related tasks for the baggage handlers as heavy lifting, awkward and restricted postures (in the small cargo areas), and time pressure. Oxley et al. (2009) provide the information that baggage handlers can lift 5-10 bags per minute during loading/unloading to the airplane. Brauer et al. (2013) state that the average weight of the baggage under the shut area are 15 kg. However, in several airports, loads up to 32 kg are also accepted. Therefore, ground personnel may lift 4-5 tons in average up to 10 tons per day. Compared to its importance, there are only limited number of studies that focus on baggage handling operations at airports. Meersman et al. (2011) assesses baggage handling in Brussels Airports and related costs. The aim is to determine the optimum number of cargo handling provider. Koblauch (2016) defines the lower back load for baggage handlers and develops a general tool to assess the specific lumbar compression. Patriarca et al. (2016) attract attention to the delay caused by baggage handling and assess performance of baggage transfer. Kim et al. (2017) aims to balance the check-in and baggage loading-unloading activities by using an algorithm and window reservation technique. Monteiro and Santos (2017) measure the loads of baggage handlers at a Brazilian airport that provide ergonomically solutions to reduce risk factors. Møller et al. (2018) search the association of exposure to WMSD shoulder loading among baggage handlers with sub acromial shoulder disorder. 5. Conclusion and Suggestions for Future Research This study contributes the airport efficiency evaluation literature by considering the safety, security, and health issues. This pioneering study, up-to-best knowledge, is the first to assess the airports in Turkey by DEA with the defined undesired output (load per baggage handler) and also considering a fuzzy model to handle the uncertain data. In DEA-II model, the seasons are defined as high, average, and low. The passenger variation can be represented by considering different months. The undesired output variable in DEA-III is modeled as triangular fuzzy numbers. Other appropriate fuzzy models can be used based on the defined input and outputs. Future studies may use one or all of the three models, defined inputs, outputs, and undesired output defined in this study to assess the airports in any other country. The limitation of this study is not considering the Airport Baggage Handling System, because the technology was not available for the airports in concerns and data were not available. The investment and operational costs of such systems can be included in a DEA model to assess the impact on airport efficiency. It is not always possible to identify the problems in the work environment, analyze task durations, and worker’s body movements. Using digital human modelling software can help to generate various scenarios and identify the risky handling operations. The potential lower back pain for baggage handlers, as studied in Koblauch (2016), can be modelled by using data from the airports. On the hand, to minimize accidents and improve safety and health standards for airport ground handling, the safety culture measures (i.e., pressure for job completion, employee’s fatigue intensity) as stated in Musa and Isha (2021) can be included in a DEA model. Data for one year is considered in this study. In future studies, it is possible to utilize Malmquist solution approach when data from several years are considered. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. -65- Journal of Airline and Airport Management 12(2), 49-68 Funding The authors received no financial support for the research, authorship, and/or publication of this article. References Augustyniak, W., López-Torres, L., & Kalinowski, S. (2015). Performance of Polish regional airports after accessing the European Union: Does liberalisation impact on airports’ efficiency?. Journal of Air Transport Management, 43, 11-19. https://doi.org/10.1016/j.jairtraman.2015.01.001 Barros, C. P., Liang, Q. B., & Peypoch, N. (2013). The efficiency of French regional airports: An inverse B- convex analysis. International Journal of Production Economics, 141(2), 668-674. https://doi.org/10.1016/j.ijpe.2012.10.007 Bezerra, G. C. L. & Gomes, C. F. (2018). Performance measurement practices in airports: Multidimensionality and utilization patterns. Journal of Air Transport Management, 70, 113-125. https://doi.org/10.1016/j.jairtraman.2018.05.006 Brauer, C., Bern, S. H., Alkjær, T., Bonde, J. P., Helweg-Larsen, K., Koblauch, H., & Mikkelsen, S. (2013). Ergonomic exposure assessed by production statistics. Occupational and Environmental Medicine, 70(Suppl 1): A100.3-A101. https://doi.org/10.1136/oemed-2013-101717.295 Chaouk, M., Pagliari, R., & Moxon, R. (2020). The impact of national macro-environment exogenous variables on airport efficiency. Journal of Air Transport Management, 82, 101740. https://doi.org/10.1016/j.jairtraman.2019.101740 Coto-Millan, P, Inglada, V., Fernandez, X. L., Inglada-Perez, L., & Pesquera, M. A. (2016). The “effect procargo” on technical and scale efficiency at airports: The case of Spanish airports (2009-2011). Utilities Policy, 39, 29-35. https://doi.org/10.1016/j.jup.2016.01.004 Coto-Millán, P., Casares-Hontañón, P., Inglada, V., Agüeros, M., Pesquera, M. Á., & Badiola, A. (2014). Small is beautiful? The impact of economic crisis, low cost carriers, & size on efficiency in Spanish airports (2009– 2011). Journal of Air Transport Management, 40, 34-41. https://doi.org/10.1016/j.jairtraman.2014.05.006 Curi, C., Gitto, S., & Mancuso, P. (2011). New evidence on the efficiency of Italian airports: A bootstrapped DEA analysis. Socio-Economic Planning Sciences, 45(2), 84-93. https://doi.org/10.1016/j.seps.2010.11.002 Dyckhoff, H., & Allen, K. (2001). Measuring ecological efficiency with Data Envelopment Analysis (DEA). European Journal of Operational Research, 132(2), 312-325. https://doi.org/10.1016/S0377-2217(00)00154-5 Emrouznejad, A., Tavana, M., & Hatami-Marbini, A. (2014). The state of the art in fuzzy data envelopment analysis. A. Emrouznejad, M. Tavana (Eds.). Studies in fuzziness and soft computing: Performance measurement with fuzzy data envelopment analysis. Springer, New York, 1-45. https://doi.org/10.1007/978-3-642-41372-8_1 Ennen, D. & Batool, I. (2018). Airport efficiency in Pakistan-A Data Envelopment Analysis with weight restrictions. Journal of Air Transport Management, 69, 205-212. https://doi.org/10.1016/j.jairtraman.2018.02.007 Fernández, X. L., Coto-Millán, P., & Díaz-Medina, B. (2018). The impact of tourism on airport efficiency: The Spanish case. Utilities Policy, 55, 52-58. https://doi.org/10.1016/j.jup.2018.09.002 Graham, A. (2005). Airport benchmarking: a review of the current situation. Benchmarking: An International Journal, 12(2), 99-111. https://doi.org/10.1108/14635770510593059 Guner, S., Cebeci, H. I., Antunes, J. J. M. A., & Wanke P. F., (2021). Sustainable efficiency drivers in Eurasian airports: Fuzzy NDEA approach based on Shannon’s entropy. Journal of Air Transport Management, 92, 102039. https://doi.org/10.1016/j.jairtraman.2021.102039 Iyer, K. C. & Jain, S. (2019). Performance measurement of airports using data envelopment analysis: A review of methods and findings. Journal of Air Transport Management, 81, 101707 . https://doi.org/10.1016/j.jairtraman.2019.101707 Kao, C., & Liu, S. T. (2000). Fuzzy efficiency measures in data envelopment analysis. Fuzzy Sets and Systems, 113(3), 427-437 . https://doi.org/10.1016/S0165-0114(98)00137-7 -66- Journal of Airline and Airport Management 12(2), 49-68 Keskin, B. & Koksal, C. D. (2019). A hybrid AHP/DEA-AR model for measuring and comparing the efficiency of airports. International Journal of Productivity and Performance Management, 68(3), 524-541. https://doi.org/10.1108/IJPPM-02-2018-0043 Kim, G., Kim, J., & Chae, J. (2017). Balancing the baggage handling performance of a check-in area shared by multiple airlines. Journal of Air Transport Management, 58, 31-49. https://doi.org/10.1016/j.jairtraman.2016.08.017 Koblauch, H. (2016). Low back load in airport baggage handlers, Danish Medical Journal. Apr, 63(4): B5233. Kocak, H. (2011). Efficiency examination of Turkish airports with DEA approach. International Business Research, 4(2), 204. https://doi.org/10.5539/ibr.v4n2p204 Lai, P. L., Potter, A., & Beynon, M. (2012). The Development of Benchmarking Techniques in Airport Performance Evaluation Research. Transportation Journal, 51(3) (Summer 2012), 305-337. https://doi.org/10.5325/transportationj.51.3.0305 Li, L., Mao, Y., Wang, Y., & Ma, Z. (2022). How has airport service quality changed in the context of COVID-19: Adata-driven crowdsourcing approach based on sentiment analysis. Journal of Air Transport Management, 105, 102298. https://doi.org/10.1016/j.jairtraman.2022.102298 Liu, D. (2016). Measuring aeronautical service efficiency and commercial service efficiency of East Asia airport companies: An application of Network Data Envelopment Analysis. Journal of Air Transport Management, 52, 11- 22. https://doi.org/10.1016/j.jairtraman.2015.12.001 Meersman, H., Pauwels, T., Struyf, E., Van de Voorde, E., & Vanelslander, T. (2011). Ground handling in a changing market: The case of Brussels Airport. Research in Transportation Business & Management, 1(1), 128-135. https://doi.org/10.1016/j.rtbm.2011.06.003 Møller, S. P., Brauer, C., Mikkelsen, S., Alkjær, T., Koblauch, H., Pedersen, E. B., Simonsen, E. B., & Thygesen, L. C. (2018). Risk of subacromial shoulder disorder in airport baggage handlers: combining duration and intensity of musculoskeletal shoulder loads. Ergonomics, 61(4), 576-587. https://doi.org/10.1080/00140139.2017.1382721 Monteiro, L. F., Santos, J. W., Franca, V. V., Santos, V. M., & Alsina, O. L. S. (2017). Measurement of physical overload in the lumbar spine of baggage handlers at a Brazilian airport. Occupational Safety and Hygiene. ed. London: CRC Press, Taylor & Francis Group, 1, 87-92. https://doi.org/10.1201/9781315164809-17 Musa, M., & Isha, A. S. N. (2021). Holistic view of safety culture in aircraft ground handling: Integrating qualitative and quantitative methods with data triangulation, Journal of Air Transport Management, 92, 102019. https://doi.org/10.1016/j.jairtraman.2021.102019 Oxley, L., Riley D., & Tapley, S. (2009). Musculoskeletal ill-health risks for airport baggage handlers. London: Health and Safety Executive. http://www.hse.gov.uk/research/rrpdf/rr675.pdf Accessed 18th of November 2022 Patriarca, R., Di Gravio, G., & Costantino, F. (2016). Assessing performance variability of ground handlers to comply with airport quality standards. Journal of Air Transport Management, 57, 1-6. https://doi.org/10.1016/j.jairtraman.2016.06.009 Rosskam, E. (2007). Excess baggage. Leveling the load and changing the workplace. Amityville, N.Y: Baywood Publishing Company Inc. Sarkis, J. (2000). An analysis of the operational efficiency of major airports in the United States. Journal of Operation Management, 18, 335-351. https://doi.org/10.1016/S0272-6963(99)00032-7 Sengupta, J. K. (1992). A fuzzy systems approach in data envelopment analysis. Computers & Mathematics with Applications, 24(8), 259-266. https://doi.org/10.1016/0898-1221(92)90203-T Sivakumar, S. (2022). A novel Integrated Risk Management Method for Airport operations. Journal of Air Transport Management, 105, 102296. https://doi.org/10.1016/j.jairtraman.2022.102296 Stichhauerova, E. & Pelloneova, N. (2019). An Efficiency Assessment of Selected German Airports Using the DEA Model. Journal of Competitiveness, 11(1), 135-151. https://doi.org/10.7441/joc.2019.01.09 -67- Journal of Airline and Airport Management 12(2), 49-68 Tabares, D. A. (2021). An airport operations proposal for a pandemic-free air travel. Journal of Air Transport Management, 90, 101943. https://doi.org/10.1016/j.jairtraman.2020.101943 Tapley, S., & Riley, D. (2005). Baggage handling in narrow-bodied aircraft: identification and assessment of musculoskeletal injury risk factors. London: Health and Safety Executive. http://www.hse.gov.uk/airtransport/docs/baggage.pdf Accessed 18th of November 2022 The Turkish State Airports Authority Directorate General. https://www.dhmi.gov.tr Accessed 18th of November 2022. Tsekeris, T. (2011). Greek airports: Efficiency measurement and analysis of determinants. Journal of Air Transport Management, 17(2), 140-142. https://doi.org/10.1016/j.jairtraman.2010.06.002 Ulku, T. (2015). A comparative efficiency analysis of Spanish and Turkish airports. Journal of Air Transport Management, 46, 56-68. https://doi.org/10.1016/j.jairtraman.2015.03.014 Ulutas, B. H. (2018). Efficiency analysis of airports in Turkey considering environmental effects. Central European Review of Economics and Management, 2(1), 65-78. https://doi.org/10.29015/cerem.583 Wanke, P., Barros, C., & Nwaogbe, O. R. (2016). Assessing productive efficiency in Nigerian airports using fuzzy- DEA. Transport Policy, 49, 9-19 . https://doi.org/10.1016/j.tranpol.2016.03.012 Wanke, P. F. (2012). Capacity shortfall and efficiency determinants in Brazilian airports: Evidence from bootstrapped DEA estimates. Socio-Economic Planning Sciences, 46(3), 216-229. https://doi.org/10.1016/j.seps.2012.01.003 Zou, B., Kafle, N., Chang, Y. T., & Park, K. (2015). US airport financial reform and its implications for airport efficiency: An exploratory investigation. Journal of Air Transport Management, 47, 66-78. https://doi.org/10.1016/j.jairtraman.2015.05.002 Journal of Airline and Airport Management, 2022 - www.jairm.org Article's contents are provided on an Attribution-Non Commercial 4.0 Creative Commons International License. Readers are allowed to copy, distribute and communicate article's contents, provided the author's and Journal of Airline and Airport Management's names are included. It must not be used for commercial purposes. To see the complete license contents, please visit https://creativecommons.org/licenses/by-nc/4.0/. -68-