Assessing the Economic Sustainability of Airlines in the U.S. Through Labor Efficiency

Abstract

This study applies data envelopment analysis (DEA) to evaluate the economic sustainability of U.S. airlines by examining labor efficiency as a pivotal component of cost management and long-term sustainability. Focusing on five key employee categories—pilots, flight attendants, ground staff, maintenance staff, and management—the analysis uses data from the MIT Airline Data Project spanning 2007 to 2020 to calculate relative efficiency scores for fifteen major airlines. The findings reveal significant disparities in labor efficiency across different airline sectors, particularly highlighting challenges in managing cost-intensive roles, such as ground, maintenance, and management staff. Notably, Southwest Airlines consistently demonstrates strong economic sustainability through its efficient labor practices, while carriers including United, jetBlue, Alaska, and Hawaiian Airlines exhibited marked inefficiencies in 2020, indicating a critical need for operational improvements. This research contributes to the field of airline management by linking labor efficiency metrics with broader economic sustainability objectives, thereby offering strategic insights for enhancing cost-effectiveness and ensuring the long-term financial health of the industry.

1. Introduction

In the U.S. airline industry, where a diverse mix of carriers competes in a dynamic market, economic sustainability has emerged as a critical objective. In the face of market fluctuations and challenges such as the COVID-19 pandemic, airlines are increasingly prioritizing strategies that ensure long-term financial resilience. A key component of this sustainability is labor efficiency, which directly influences operating costs and overall economic performance. Notably, labor expenses rose to 45.6% of total operating costs in the second quarter of 2020, up from 32.3% in 2019 (A4A, 2020; BTS, 2021) [1,2], highlighting the need for more cost-effective workforce management. In an effort to enhance capacity and sustainability, U.S. airlines reported a 2.5% increase in full-time employees in January 2020 compared to the previous year, and a 12.6% increase over four years (BTS, 2020a) [3]. Research indicates that efficient labor allocation plays a vital role in not only improving operational efficiency but also in achieving sustainable economic growth (Li and Cui, 2018) [4]. With challenges such as pilot and ground staff shortages, high training costs, and the imperative to optimize workload management, airlines must develop innovative strategies that balance labor costs with enhanced employee performance to enhance their competitive edge and long-term economic sustainability (Taneja, 2020) [5].

According to BTS data, from December 2019, American Airlines led full-service carriers (FSCs) with 95,518 full-time employees, followed by United Airlines (79,745) and Delta Airlines (79,841). Among low-cost carriers (LCCs), Southwest had the highest full-time employee count (59,629), followed by jetBlue (16,648). In contrast, Allegiant had the fewest full-time employees (3908), followed by Frontier (4905) and Hawaiian (6011) (BTS, 2020b) [6]. Figure 1 provides a breakdown of airline employment in 2020, revealing that ground staff accounted for 35%, followed by flight attendants (23%), pilots (17%), management staff (15%), and maintenance staff (10%). Ground staff handles passenger, cargo, and aircraft responsibilities, while maintenance staff conducts technical evaluations and aircraft maintenance before and after flights, and management staff oversees managerial activities.

The COVID-19 pandemic in early 2020 had substantial impacts on the airline industry. The surge in COVID-19 cases, along with states’ stay-home orders and travel restrictions, had reduced air travel demand significantly. According to BTS, U.S. airlines reported substantial losses in three consecutive quarters in 2020, with a loss of $4.1 billion in the first quarter, $8.9 billion in the second quarter, and $9.5 billion in the third quarter in domestic operations (BTS, 2020c) [8]. So, how did the pandemic affect airline employee recruitment? Figure 2 compares full-time employees among the ten current airlines in the U.S. between December 2019 and October 2020. It appears that major carriers, including American, United, and Delta, were the ones that made a significant reduction in their full-time employment amid COVID-19. LCCs show a slight reduction in their full-time employment, as seen with Southwest, jetBlue, and Alaska. Allegiant, Frontier, Spirit, and Hawaiian almost kept the same full-time employment.

Academic literature on airline efficiency has often overlooked labor efficiency. Labor is usually treated as one input variable, with a lack of focus on the impacts of wages and benefits. Employee-type-specific labor efficiency, essential due to varied roles in flight operations, remains underexplored. Studies categorizing employees often neglect key factors like salaries, wages, benefits, and workloads (Alam and Sickles, 1998 [9]; Coelli et al., 1999) [10]. Additionally, there is a gap in understanding how airline labor efficiency evolves over time and how inefficient airlines can regain efficiency in a competitive market. Given recent mergers and the impact of COVID-19, assessing changing efficiency scores could provide insights into airline situations and strategies for sustainable success. The detailed literature review and research gap are presented in Section 2.

This paper aims to assess the labor efficiency of U.S. airlines longitudinally from 2007 to 2020 within the framework of economic sustainability. Focusing on five distinct categories of airline employees—pilots, flight attendants, ground staff, maintenance staff, and management staff—this research provides a comprehensive analysis of how efficient labor practices can drive long-term economic performance in the U.S. airline industry. This study contributes to the existing literature by establishing a benchmark for sustainable labor efficiency, identifying airlines that excel or lag in integrating efficient labor practices, and cross-comparing efficiency trends across different airline sectors. To achieve these objectives, a series of efficiency models were constructed using the data envelopment analysis (DEA) method. These models compute annual efficiency scores for all U.S. airlines across the five employee categories, offering a holistic view of how labor efficiency influences economic sustainability over the 14-year study period.

There are several important delimitations in this study. First, it is focused on the U.S. domestic air travel market only, i.e., the revenue from international flights was excluded. This delimitation is to ensure the comparability of the results since only employment details for domestic operations are included in the data. Furthermore, only the preliminary impact of COVID-19 on airline labor efficiency can be examined, but not the post-pandemic period, as data are only available until 2020. Nonetheless, the trend of efficiency scores in the past 14 years, along with recommended improvements for inefficient airlines in 2020, allows us to evaluate the challenges those airlines would have to face in similar disruptive events in the future.

This paper is structured as follows. The next section reviews relevant literature on airline efficiency and discusses research gaps. Then, the methodology is explained, including DEA models, variables, and data collection. Subsequently, the results of the DEA models are presented along with the comparison of labor efficiency scores across airlines over 14 years. Finally, discussions and conclusions on the implications of this study are provided.

2. Literature Review

2.1. Review of Previous Studies

Airline efficiency, a crucial factor in the competitive airline industry, is often assessed through diverse literature perspectives. Many studies incorporate labor as an input variable, with labor costs being a significant portion of airlines’ total operating expenses.

Table 1. Airline efficiency literature.

Authors	Input Variables	Output Variables	Airlines
Caves, 1984 [30]	Labor cost, fuel price, flight equipment, ground property, equipment (GPE), and materials	Revenue passenger miles (RPM), RPM of charter service, revenue ton miles (RTM) of mail, and RTM of all other freight	15 U.S. airlines from 1970 and 1981
Schmidt and Sickles, 1984 [31]	Capital, labor, energy, and materials	Capacity ton miles (CTM)	12 U.S. airlines from 1970 to 1978
Gillen et al., 1990 [32]	Labor, fuel, capital, and materials	Scheduled revenue passenger kilometers, scheduled revenue freight ton kilometers, non-scheduled revenue ton kilometers, load factor, stage length, and average number of seats per departure	7 Canadian airlines from 1964 to 1981
Bauer, 1990 [33]	Labor, capital (flight equipment and landing fees), energy cost (quantity of fuel), and materials cost (advertising, insurance, commissions, and passenger meals)	Revenue passenger ton miles and revenue cargo ton miles	12 U.S. airlines between 1970 and 1981
Cornwell et al., 1990 [34]	Capital, labor, energy, and materials	Available ton miles, average stage length, and service quality	8 U.S. airlines from 1970 to 1981
Schefczyk, 1993 [35]	Available ton kilometers, facilities, affiliated companies, current asset, labor, aircraft fuel, and commission to agents	Revenue passenger kilometers, non-passenger revenue (cargo), and other revenue	15 airlines from 1989 to 1992
Good, Nadiri, Roller, and Sickles, 1993 [36]	Labor, energy/materials, and aircraft fleet	Passenger revenues, revenue ton kilometers (RTK), and incidental services	Four European airlines and eight U.S. airlines
Baltagi et al., 1995 [37]	Labor, fuel, and materials,	Points served, average stage length, and technical change (load factor and fuel efficiency)	24 U.S. airlines from 1971 to 1986
Good et al., 1995 [38]	Labor cost, fuel cost, materials cost, and flight equipment cost	Revenue passenger ton kilometers, revenue cargo ton kilometers, and incidental services revenue	16 European and U.S. airlines from 1976 to 86
Ray and Mukherjee, 1996 [39]	Labor, fuel, materials, flight equipment, and ground equipment	Load factor, points served, and average stage length	21 U.S. airlines from 1983 to 1984
Alam and Sickles, 1998 [9]	Flight capital (number of planes), labor (pilots, flight attendants, mechanics, and ground handlers), and gallons of aircraft fuel and materials (supplies, outside services, and non-flight capital)	Revenue ton miles (RTM), aircraft stage length, load factor, average size of the carrier’s aircraft, and percentage of a carrier’s fleet	11 U.S. airlines from 1970 to 1990
Coelli et al., 1999 [10]	Labor input (cockpit crew and flight attendant) and capital input (sum of the maximum take-off weights of all aircraft multiplied by the number of operation days)	Passenger and cargo ton kilometers available, average stage length, average load factor, and average aircraft size	32 airlines from 1977 to 1990 (from different regions)
Forsyth, 2001 [40]	Employees, fuel, capital expenditure deflated by capital equipment price, and other materials	Revenue passenger kilometers (RPK) and total revenue deflated by price index	Australian domestic airlines
Sickles et al., 2002 [24]	Flight capital (number of planes), labor (pilots, flight attendants, mechanics, and passenger and aircraft handlers), energy (gallons of aircraft fuel), and materials (supplies, outside services, and non-flight capital)	Revenue ton miles (RTM), average stage length, average load factor, and average aircraft size	16 European airlines from 1977 to 1990
Oum, Fu, and Yu, 2005 [11]	Labor, fuel, materials, flight equipment, and ground property and equipment	Scheduled passenger service, scheduled freight service, mail service, non-scheduled passenger and freight services, and incidental services output	10 Airlines in North America from 1990–2000
Chiou and Chen, 2006 [12]	Fuel cost, personnel cost, and aircraft cost	Number of flights, seat-mile, passenger mile, and embarkation passengers	15 Taiwanese air routes in 2001
Färe, Grosskopt and Sickles, 2007 [41]	Multilateral labor index, fuel, fleet index adjusted for aircraft size and age, and material input	Revenue passenger miles (RPM) (scheduled) and revenue ton miles (RTM) (cargo and nonscheduled)	13 U.S. airlines from 1979–1994
Barbot et al., 2008 [25]	Labor (number of core business workers), fleet (number of operated aircraft), and fuel (in gallons consumed)	Available seat kilometers (ASK), revenue passenger kilometers (RPK), and revenue ton kilometers (RTK)	49 airlines in 2005
Greer, 2008 [42]	Full-time equivalent employees (FTEs) and fuel (in gallons consumed)	Seating capacity and available seat miles (ASM)	8 U.S. airlines in 2000 and 2004
Bhadra, 2009 [43]	Jet fuel, full time employees (FTEs), ratio of flight stage miles to trip stage miles, utilization of aircraft, number of seats per aircraft, and number of aircraft	Available seat miles (ASM)	13 U.S. airlines from 1985 to 2006
Greer, 2009 [44]	Labor, fuel, and fleet-wide seating capacity	Available seat miles (ASM)	17 Major U.S. airlines from 1999 to 2008
Barros and Peypoch, 2009 [13]	Number of employees, operational cost, and number of planes	Operating revenue passenger kilometers and EBIT	27 European airlines from 2000 to 2005
Assaf, 2009 [45]	Total operational cost, labor cost, aircraft fuel and oil expenses, number of planes, and load factor	Total operating revenues	12 U.S. airlines from 2006 to 2007
Hong and Zhang, 2010 [46]	Labor and capital	Total revenue, revenue passenger kilometers (RPK), and freight revenue ton kilometers (RTK)	29 airlines from 1998 to 2002
Ouellette et al., 2010 [47]	Labor, fuel, and materials	Passenger–freight ton kilometers	7 Canadian airlines from 1960 to 1999
Brits, 2010 [48]	Employees, fuel, capital, and materials	Passenger ton kilometers (PTK), freight ton kilometers (FTK), and postage ton kilometers (postage TK)	South African airlines (years unidentified)
Zhu, 2011 [49]	Cost per ASM, labor cost, fuel expense, and gallons of fuel used	Load factor, fleet size, revenue passenger miles (RPM), and passenger revenue	21 U.S. airlines from 2007 to 2008
Merkert and Hensher, 2011 [50]	Labor cost (FTE), available ton kilometers (ATK), full-time equivalent (FTE) price, and ATK price	Revenue passenger kilometers (RPK) and revenue ton kilometers (RTK)	58 international airlines from 2007 to 2009
Wang et al., 2011 [51]	Number of employees, fuel expense, and number of aircraft in service	Available seat miles (ASM), revenue passenger miles (RPM), and non-passenger revenue	22 U.S. carriers and 8 non-U.S. airlines in 2006
Gramani, 2012 [52]	Aircraft fuel, wages, salaries, benefits, and cost per available seat mile (CASM)	Revenue passenger mile (RPM) and flight revenue	2 Brazilian and 2 American airlines 1997–2006
Assaf and Josiassen, 2012 [53]	Labor, capital, fuel, and other operating inputs (salaries, benefits, and capital cost)	Revenue passenger kilometers (RPKs), incidental revenues (nonairline revenues), and purchasing power parity (PPP) index	31 airlines in Europe, U.S., and Canada
Barros et al., 2013 [54]	Total cost, number of employees, and number of gallons	Total revenue, revenue passenger mile (RPM), load factor	11 U.S. airlines from 1998 to 2010
Barros and Couto, 2013 [55]	Number of employees, operational cost, and number of seats available	Revenue by passenger kilometers, and revenue cargo ton of freight carried	23 European airlines from 2000 to 2011
Rai, 2013 [56]	Number of planes, number of employees, and gallons of fuel consumed	Revenue passenger miles (RPM), number of departures, number of passengers, and available ton miles (ATM)	10 U.S. airlines from 1986 to 1995
Lee and Worthington, 2014 [57]	Kilometers flown, number of employees, total assets, ownership, departures, and weighted load factor	Available ton kilometers (ATK)	42 airlines in Asia, Europe, and North America in 2006
Arjomandi and Seufert, 2014 [18]	Number of full-time equivalent employees and number of flying hours divided by average daily revenue hours	The number of tons available for revenue load (pax, cargo, and mail) on each flight multiplied by the flight distance, and CO2 emissions	48 of the world’s major full-service and low-cost carriers from six different regions over the period 2007–2010
Cao, Lv, and Zhang, 2014 [58]	Labor, fuel, and number of aircraft	Total flights and revenue–ton kilometers (RTM) of passengers and freight	China’s airlines from 2005–2009
Lozano and Gutiérrez, 2014 [59]	Fuel cost, non-current assets, wages and salaries, other operating costs, and selling costs	Available seat kilometers (ASK), available ton kilometers (ATK) (intermediate outputs), revenue passenger kilometers (RPK), and revenue ton kilometers (RTK)	16 European airlines in 2007
Wu and Liao, 2014 [28]	Material, energy, capital, and passenger, labor	Operating revenue, return on asset (ROA), return on investment (ROI), and net income (NI)	38 international airlines in 2010
Tavassoli, Faramarzi, and Saen, 2014 [29]	Number of passenger planes, labor, and number of cargo planes	Passenger–plane kilometers and cargo–plane kilometers (intermediate outputs); passenger kilometers and ton kilometers (final outputs)	11 domestic airlines in Iran in 2010
Lu et al., 2012 [26]	Employees (FTEs), fuel consumed, seating capacity, flight equipment, maintenance expense, and ground property and equipment	Revenue passenger miles (RPMs) and non-passenger revenue (NPR)	30 U.S. airlines in 2010
Cui and Li, 2015 [17]	Number of employees, capital stock, and tons of aviation kerosene	Revenue ton kilometers (RTK), revenue passenger kilometers (RPK), total business income, and CO2 emissions	11 airlines in Asia, Europe, and North America from 2008–2012
Marti et al., 2015 [60]	Labor costs, assets, and supplies	Operating revenue	28 airlines in Europe in 2010
Barros and Wanke, 2015 [61]	Number of employees, number of aircraft, and operating costs	Revenue ton kilometers (RTK) and revenue passenger kilometers (RPK)	29 African airlines from 2010 to 2013
Choi, Lee, and Olson, 2015 [62]	Employees number and available seat miles (ASM)	Service quality index, revenue passenger miles (RPM), and operating revenue	12 U.S. airlines from 2008 to 2011
Cui et al., 2016 [19]	Number of employees and aviation kerosene	Revenue ton kilometers (RTK), revenue passenger kilometers (RPK), and total business income (TBI).	21 airlines in Asia, Europe, and North America
Kottas and Madas, 2018 [14]	Number of employees, operating costs, and number of operated aircraft,	Total operating revenue, revenue passenger kilometers (RPK), and revenue ton kilometers (RTK)	6 U.S. airlines, 9 European airlines, and 15 Asia Pacific airlines from 2010–2016
Sakthidharan and Sivaraman, 2018 [63]	Available ton kilometers (ATK), cost per available seat kilometers (CASK), fuel per ASK, CASK ex-fuel, maintenance cost per ASK, ownership per ASK, and number of employees	Revenue passenger kilometers (RPK) and freight ton kilometers (FTK)	5 Indian airlines from 2013–2014
Li and Cui 2018 [4]	Number of employees, aviation kerosene, fleet size, and sale costs	Available seat kilometers (ASK), revenue passenger kilometers (RPK), and total revenue	29 global airlines from 2008 to 2015
Cui and Li, 2018 [20]	Number of employees and aviation kerosene	Total revenue, greenhouse gases emissions, dynamic factors, and fleet size	29 global airlines from 2009 to 2015
Hermoso et al., 2019 [64]	Number of employees, total assets, destinations, degree, eigencentrality, tweets, publication, and number of videos	Sales, passengers, liked Twitter, liked Facebook, and views	43 European airlines in 2014
Mhlanga, 2019 [15]	Number of employees, available seat kilometers, operating expense, and employee expenditure	Revenue passenger kilometers (RPK) and operating revenue	10 African airline from 2012–2016
Heydari et al., 2020 [16]	Number of employees, available seat kilometers (ASK), available ton kilometers (ATK), and number of scheduled flights	Passenger kilometers performed (PKP) and ton kilometers performed (TKP)	124 Iranian airlines in 2014
Yen et al., 2022 [27]	Flight frequency, aircraft size, and fuel consumption	Passenger load factor and revenue per available seat kilometer (RASK)	Two Taiwanese airlines operating 112 international routes
Saini et al., 2023 [21]	Fuel consumption, number of employees, and operating expenses	Revenue passenger kilometers (RPK), available seat kilometers (ASK), and environmental performance indicators (CO2 emissions)	13 global airlines from 2013–2015
Voltes-Dorta et al., 2024 [23]	Operational costs, fuel consumption, and labor costs	Revenue passenger kilometers (RPK) and environmental and social sustainability scores	34 global airlines from 2019 and 2022
Yang et al. (2024) [22]	Fuel consumption, labor hours, and number of aircraft	Revenue passenger kilometers (RPK), customer satisfaction scores, and CO2 emissions	Iranian airlines in 2022

presents the primary literature on airline efficiency using labor as one of the input variables, ranked by publication year. Labor is a vital input in operational efficiency (Oum et al., 2005 [11]; Chiou and Chen, 2006 [12]; Barros and Peypoch, 2009 [13]; Kottas and Madas, 2018 [14]; Li and Cui, 2018 [4]; Mhlanga, 2019 [15]; Heydari et al., 2020 [16]), environmental impact efficiency (Cui and Li, 2015 [17]; Arjomandi and Seufert, 2014 [18]; Cui et al., 2016 [19]; Cui and Li, 2018 [20]; Saini et al., 2023 [21]; Yang et al., 2024 [22]; Voltes-Dorta et al., 2024 [23]), and overall efficiency studies (Alam and Sickles, 1998 [9]; Coelli et al., 1999 [10]; Sickles et al., 2002 [24]; Barbot et al., 2008 [25]; Lu et al., 2012 [26]; Yen et al., 2022 [27]) conducted domestically and internationally (Wu and Liao, 2014 [28]; Tavassoli et al., 2014 [29]; Cui et al., 2016 [19]; Kottas and Madas, 2018 [14]). Despite its inclusion in efficiency models, labor is usually measured as an aggregated number, with only a few studies breaking it down by employee type (Alam and Sickles, 1998 [9]; Coelli et al., 1999 [10]).

Almost all airline efficiency studies employ DEA as a quantitative method, which helps compare relative efficiency across airlines in the same market (Kottas and Madas, 2018 [14]; Li and Cui, 2018 [4]; Heydari et al., 2020 [16]). DEA is a mathematical optimization model determining the efficient frontier or benchmark, with efficiency scores of 1 indicating efficiency. It is widely accepted for evaluating airline efficiency and comparing efficiency across airlines in the same market. In those efficiency models, input and output variables vary depending on the study’s focus, with common output variables including available seat miles (ASM), revenue passenger miles (RPM), load factor, passenger revenue, and total operating revenue for operational and overall efficiency models. Common input variables include labor costs, fuel costs, fleet size, number of flights, and various other factors depending on the study’s specific scope (Chiou and Chen, 2006 [12]; Barros and Peypoch, 2009 [13]; Kottas and Madas, 2018 [14]; Mhlanga, 2019 [15]; Heydari et al., 2020 [16]; Arjomandi and Seufert, 2014 [18]; Cui and Li, 2015 [17]; Cui et al., 2016 [19]; Cui and Li, 2018 [20]).

2.2. Research Gaps and Significance of This Study

The existing literature on airline efficiency delivers valuable evaluations of airline efficiency on both international and domestic scales. However, it lacks an exclusive focus on labor efficiency. In these studies, labor is just one of several input variables within the DEA model, making it challenging to distinctly assess how efficiently airlines employ their labor workforce to achieve desired outputs. Consequently, the ability to evaluate and compare airlines’ labor efficiency or inefficiency and pinpoint necessary improvements is hindered by the potential distortion of labor’s effects by other inputs.

Moreover, most airline efficiency studies represented labor as an aggregated figure without categorization, such as wages, benefits, and workloads. Even a study like Li and Cui (2018) [4], which claimed to focus on optimal employee allocation, employed the aggregated number of employees alongside other inputs like aviation kerosene, fleet size, and sales costs. Very few studies, such as Alam and Sickles (1998) [9] or Coelli et al. (1999) [10], attempted to break labor down into categories like pilots, flight attendants, mechanics, or ground handlers but overlooked factors such as salaries, benefits, and workloads. Additionally, management staff are consistently omitted from these analyses. Finally, the existing literature appears to prioritize measuring and comparing efficiency across airlines rather than addressing the need for improvement among inefficient airlines. Airlines may transition from being efficient one year to less efficient the next, reflecting the dynamic nature of the market. Understanding the necessary improvements for such airlines is crucial for identifying weaknesses and developing strategies to regain a competitive edge.

These gaps in the airline efficiency literature are especially significant given the pivotal role of the labor workforce in airline operations. Historical trends show that many airlines have vanished from the market due to mergers or acquisitions. Most recently, the COVID-19 pandemic has severely impacted the airline industry, resulting in substantial workforce reductions, particularly among major carriers. Evaluating and comparing airline labor efficiency over time can shed light on these circumstances and help explain why certain airlines have disappeared. Moreover, assessing labor efficiency in 2020 offers valuable insights into the challenges faced by airlines during the pandemic or a similar disruptive event. Inefficient airlines can gain insight into necessary improvements and whether they can regain their competitive edge. Even efficient airlines may encounter significant post-pandemic challenges due to shifts in air travel demand, necessitating strategic adjustments to maintain their competitive position. Given that labor costs account for the largest proportion of total operating expenses, it is reasonable to anticipate significant changes in the labor workforce for many airlines.

This paper seeks to address these gaps by evaluating and comparing labor efficiency among U.S. airlines longitudinally from 2007 to 2020. The assessment focuses on five primary categories of airline employees: pilots, flight attendants, ground staff, maintenance staff, and management staff. Input variables encompass the count of employees, salaries, benefits expenses, and workload for each employee category. Non-employee input variables are intentionally excluded to preserve the integrity of labor’s impact on airline efficiency. DEA models were developed for each employee category separately, providing a deeper understanding of how airlines have managed their labor workforces over the past 14 years. The longitudinal analysis offers valuable insights into the overall efficiency of all U.S. airlines, particularly those involved in mergers or acquisitions. Given the significant challenges posed by the pandemic, inefficient airlines in 2020 were further assessed to identify recommended improvements.

3. Methodology

3.1. Data Envelopment Analysis (DEA) Models

The data envelopment analysis (DEA) method is the chosen approach to calculate the relative efficiency of airlines, establish a reference efficiency benchmark (known as the efficient frontier), and pinpoint areas where airlines can enhance their performance. In DEA, efficiency is defined as the total weighted output divided by the total weighted inputs. DEA is a powerful technique rooted in linear programming, and its application revolves around the examination of input and output variables for specific decision-making units (DMUs) (Charnes, Cooper, and Rhodes, 1978 [65]; Cooper et al., 2011 [66]).

DEA operates as a comprehensive multifactor analysis model, providing insights into the relative efficiencies of DMUs. Relative efficiency in DEA refers to the performance of a DMU, such as an airline, compared to the best performers in a given group, also called the “efficient frontier”, using the same types of inputs to produce similar outputs. This evaluation is at the heart of DEA’s methodology, providing a foundation for assessing how well an airline utilizes its inputs to produce outputs effectively. The essence of DEA’s functioning lies in determining the most efficient combination of inputs and outputs for each airline under examination, thereby offering a roadmap for enhancing efficiency (Charnes et al., 1978 [65]; Cooper et al., 2011 [66]). The relative efficiency score ranges from 0 to 1 (or 0% to 100%). A DMU is considered efficient if its efficiency score is equal to 1, indicating that it operates at the frontier’s level of efficiency. If the score is less than 1, the DMU is deemed inefficient.

Let us examine an example of twenty airlines operating in the same market. Hypothetical input variables (resources used) include the number of employees, gallons of jet fuel used, and total operating costs. Outputs (results produced) are revenue passenger miles (RPM), the number of flights operated, and the customer satisfaction score. Suppose Airline 1 uses fewer employees and fuel but still flies more passengers, has higher customer satisfaction, and experiences fewer delays. DEA might identify it as efficient, assigning it a relative efficiency score of 1.0 (100%). Now, consider Airline 2. It uses more fuel and labor per RPM, and its customer satisfaction score is lower. DEA might rate its efficiency at 0.85, indicating that it could reduce 15% of its inputs or improve outputs by 15% to match the performance of the efficient airlines.

DEA is flexible and adaptable, capable of operating as an input minimization model or an output maximization model, depending on the specific analytical requirements and objectives. This adaptability ensures that DEA can accommodate a wide range of scenarios and industries, making it a valuable tool for performance assessment and enhancement. In essence, DEA empowers analysts to measure airline efficiency, establish benchmarks, and uncover opportunities for improvement by optimizing the allocation of resources.

In this paper, five primary DEA models were constructed, one for each type of airline employee: pilots, flight attendants, ground staff, maintenance staff, and management. The input variables differ among these models, as presented in

Table 2. Input and output variables.

	Input Variables	Output Variables
Model 1: pilot efficiency	Pilots—annual wages and salaries Pilots—average block hours per month Pilots—average pension and benefit package Pilots—total cockpit cost per block hours Pilots—total employee equivalents	Available seat miles (ASM) Revenue passenger miles (RPM) Total passenger revenue
Model 2: flight attendant efficiency	Flight attendants—average annual wages Flight attendants—average block hours Flight attendants—total employee equivalents	Available seat miles (ASM) Revenue passenger miles (RPM) Total passenger revenue
Model 3: ground staff efficiency	Ground staff—average annual wages Ground staff—employees per aircraft Ground staff—total employee equivalents	Available seat miles (ASM) Revenue passenger miles (RPM) Total passenger revenue
Model 4: maintenance staff efficiency	Maintenance staff—average annual wages Maintenance staff—percent of maintenance Maintenance staff—total employee equivalents	Available seat miles (ASM) Revenue passenger miles (RPM) Total passenger revenue
Model 5: management staff efficiency	Management staff—average annual wages Management staff—total employee equivalents	Available seat miles (ASM) Revenue passenger miles (RPM) Total passenger revenue

. It is important to note that certain input variables are only available for specific types of employees due to data availability.

Output variables include available seat miles (ASM), revenue passenger miles (RPM), and total passenger revenue. These outputs are consistent with existing literature (Good et al., 1993 [36]; Greer, 2008 [42]; Greer, 2009 [44]; Zhu, 2011 [49]; Wang et al., 2011 [51]; Lozano and Gutiérrez, 2014 [59]; Li and Cui, 2018 [4]; Heydari et al., 2020 [16]; Losa et al., 2020 [67]). Load factor is not used to avoid duplication since it is the ASM divided by RPM. In addition, total passenger revenue is selected instead of total operating revenue since it reflects the revenue from air passengers and captures better airline operations and success in air travel, whereas the total operating revenue may include revenues from other investment sources.

Most input variables, such as wages, salaries, pension, benefit package, and total employee equivalents, are self-explanatory. The following presents definitions of some key input and output variables:

• Block hour: number of hours between the aircraft door closing time at the departure of the flight and the aircraft door opening time at the arrival gate following its landing;
• Percent of maintenance: percent of maintenance expenses outsourced;
• Passenger revenue: revenue received by the airline from the carriage of passengers in scheduled operations;
• Revenue passenger miles (RPM): number of miles traveled by paying passengers, calculated by multiplying the number of paying passengers by the distance traveled;
• Available seat miles (ASM): a measure of passenger carrying capacity, calculated by multiplying the number of seats available by the distance traveled.

In this study, the output maximization DEA model is selected. More specifically, the model determines the maximum level of outputs the units have achieved, given the level of inputs used by them. Charnes et al. (1978)’s [65] model used the constant returns to scale (CRS). Banker, Charnes, and Cooper (1984) [68] extended Charnes et al.’s model by adding variable returns to scale (VRS). Since proportional changes for input and output variables cannot be ensured in airline efficiency based on labor, the VRS was used (Banker et al., 1984) [68]. In this paper, the output-oriented BCC model is selected for modeling, as presented by the following formulas:

$$\mathcal{E}*=Max \mathcal{E}$$

subject to

$$\sum _i^n\sum _j^m{w}_i{x}_{ij}\le x_i$$

$$\sum _i^n\sum _k^l{w}_i{y}_{iv}\ge \mathcal{E}{y}_i$$

$$\sum _i^n{w}_i=1$$

$$w_i\ge 0$$

where

• $\mathcal{E}*$ = efficiency score;
• i = number of DMUs; i = 1, 2, …, n; n = 15 airlines (the number is lower for several years due to the airline merging);
• j = number of inputs; j = 1, 2, …, m (Model 1: m = 5; Model 2: m = 3; Model 3: m = 3; Model 4: m = 3; and Model 5: m = 2);
• k = number of outputs; k = 1, 2, …, l; l = 3 for all models (ASM, RPM, and total passenger revenue);
• $w_i$ = weight applied for inputs and outputs for DMU (i);
• $x_{ij}$ = input j for DMU (i);
• $y_{ik}$= output k for DMU (i).

Explanation of the BCC formula:

• Objective: to maximize efficiency ($\mathcal{E}$) for a specific DMU under evaluation;
• Constraint 1 (input constraint): this ensures the weighted combination of all other DMUs does not use more inputs than the DMU being evaluated;
• Constraint 2 (output constraint): this indicates the composite DMU’s outputs should be at least as good as the actual DMU’s outputs multiplied by E;
• Constraint 3 (convexity constraint): this enforces variable returns to scale by making sure that the weights form a convex combination (like creating a blend of other DMUs without inflating the scale). This is what differentiates BCC from the earlier CCR model (which assumes constant returns to scale);
• Constraint 4 (non-negativity): must not have a “negative” contribution from a DMU in the benchmark. All weights must be positive or zero.

The use of weights in DEA helps mitigate the scale differences in variables when assessing the efficiency of DMUs. By assigning appropriate weights to the inputs and outputs in a DEA model, more importance could be given to certain variables, and the impact of scale could be reduced. Furthermore, data normalization was performed, which ensures that all input and output variables are on a comparable scale. This step involves transforming the data so that each variable falls within a common range between 0 and 100).

DEA models were built and run for each year from 2007 to 2020, resulting in efficiency scores for all the airlines in those fourteen years. Non-parametric ANOVA was conducted with those efficiency data to compare the efficiency scores among three sectors of U.S. airlines, FSCs, LCCs, and Others. Then, the yearly efficiency scores were assessed to show how airline efficiency had changed over time. For 2020, further analysis was conducted to identify improvements needed for airlines with efficiency scores lower than 1. In other words, the improvement results show what changes to specific output or input variables for those airlines to improve their efficiency scores. The information helps the airlines identify their weaknesses and improvements needed to enhance their strategies for labor workforce requirements and employee compensations.

3.2. Data Collection

Data were collected from the MIT Airline Data Project from 2007 to 2020 (MIT, 2022) [7]. The Airline Data Project (ADP), established by the MIT Global Airline Industry Program, focuses on understanding the opportunities, risks, and challenges for the airline industry. The ADP was developed based on several public resources, including U.S. Department of Transportation Form 41 (U.S. DOT Form 41) from the BTS and relevant filings to the Securities and Exchange Commission (SEC) (MIT, 2022 [7]). The latest update of the ADP was in December 2021 for the 2020 data. Accordingly, the post-pandemic effects on airline efficiency cannot be evaluated in this study. Future studies can use the data from 2021 and later to calculate the efficiency and compare it with the findings of this study.

Data validation is incorporated in all the phases of data collection, including compilation, processing and analysis, and quality evaluations of the statistical or financial information. The aircraft types used in the study include small narrowbody aircraft (150 seats and less, such as Boeing 737-700 and Airbus A319), large narrowbody aircraft (151 seats and more, such as Boeing 737-800/900/Max 8/Max 9, Boeing 757, Airbus A321/A320 NEO/A321NEO), and widebody aircraft (two aisle configuration), which allow us to eliminate incomparability across multiple carriers (MIT, 2022) [7]. These types of aircraft represent well the aircraft used by U.S. airlines for commercial aviation.

In this study, the operational and revenue data from the ADP project are collected for fifteen major U.S. airlines. Using the ICAO’s recommendations for LCCs, they are categorized into three sectors: FSCs (American Airlines, Continental Airlines, Delta Air Lines, Northwest Airlines, United Airlines, and U.S. Airways), LCCs (Southwest Airlines, jetBlue Airways, Allegiant Air, AirTran Airways, Frontier Airlines, Spirit Airlines, and Virgin America), and Others (Alaska Airlines and Hawaiian Airlines) (ICAO, 2020) [69]. Alaska Airlines and Haiwaian Airlines are listed under Others because they are not among ICAO’s recommended list for LCCs. The data are from 2007 to 2020. This dataset captures the data of all the airlines in the U.S. airline industry during this time period. Hence, it represents well the U.S. airline industry and all the airlines operating in this market. However, the mergers of several airlines over time resulted in the disappearance of some airlines in the comparison in specific years. The following information is noted regarding the mergers and the year (Bahloul, 2024) [70]:

• September 2005: U.S. Airways (U.S.) and America West (HP) merged and started to report jointly as U.S. Airways (U.S.). Accordingly, America West is not included in the analysis;
• December 2009: Delta (DL) and Northwest (NW) merged and started to report jointly as Delta (DL);
• January 2012: United (UA) and Continental (CO) merged and started to report jointly as United (UA);
• May 2011: Southwest (WN) and AirTran (FL) merged and started to report jointly as Southwest (WN);
• December 2013: American (AA) and U.S. Airways (U.S.) merged and started to report jointly as American (AA);
• December 2016: Alaska Airlines and Virgin America merged and started to report jointly as Alaska Airlines.

The information for input and output variables for five types of airline employees was extracted from this dataset from 2007 to 2020. The data were consolidated by variables and years in Excel spreadsheets. Except for missing values for non-existing airlines, there are no other missing values. The Excel files were imported to Frontier Analyst software (https://banxia.com/frontier/) for DEA modeling purposes.

4. Results

4.1. Demographic Information

Table 3. Total Operating Revenue (in billion USD).

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	22.83	23.70	19.90	22.15	23.96	24.83	25.76	27.14	41.08	40.42	42.20	44.53	45.76	17.34
Continental	14.11	15.03	12.36	14.01	16.17	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Delta	19.24	20.97	18.05	31.89	35.32	36.92	37.82	40.43	40.82	39.85	41.48	44.54	47.13	17.12
Northwest	12.73	14.10	10.86	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
United	20.05	20.24	16.36	19.68	21.16	37.16	38.29	38.90	37.86	36.56	37.74	41.30	43.26	15.36
U.S. Airways	12.05	12.46	10.78	12.20	13.34	14.12	14.94	15.75	0.00	0.00	0.00	0.00	0.00	0.00
America West	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
FSCs	101.02	106.49	88.31	99.93	109.94	113.02	116.80	122.22	119.76	116.83	121.41	130.37	136.15	49.81
Southwest	9.86	11.02	10.35	12.10	13.65	17.09	17.70	18.61	19.82	20.43	21.17	21.97	22.43	9.05
jetBlue	2.84	3.39	3.29	3.78	4.51	4.98	5.44	5.82	6.42	6.63	7.02	7.66	8.09	2.96
AirTran	2.31	2.55	2.34	2.62	2.94	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00
Frontier	1.33	1.37	1.11	1.32	1.66	1.43	1.35	1.57	1.60	1.71	1.91	2.16	2.51	1.25
Virgin America	0.02	0.37	0.55	0.72	1.04	1.33	1.42	1.49	1.53	1.66	1.64	0.00	0.00	0.00
Spirit	0.76	0.79	0.70	0.78	1.07	1.32	1.65	1.93	2.14	2.32	2.65	3.32	3.83	1.81
Allegiant	0.34	0.48	0.54	0.64	0.75	0.87	0.96	1.10	1.22	1.32	1.44	1.60	1.75	0.93
LCCs	17.46	19.98	18.88	21.96	25.62	27.02	28.53	30.52	32.73	34.08	35.83	36.70	38.61	15.99
Alaska	3.08	3.22	3.01	3.43	4.31	4.65	5.15	5.36	5.59	5.83	6.29	8.26	8.77	3.56
Hawaiian	0.98	1.21	1.18	1.31	1.65	1.96	2.16	2.31	2.31	2.44	2.69	2.83	2.83	0.84
Others	4.06	4.43	4.19	4.74	5.96	6.61	7.31	7.67	7.91	8.27	8.98	11.09	11.60	4.41
Total All Sectors	122.54	130.91	111.37	126.63	141.53	146.66	152.63	160.41	160.40	159.17	166.22	178.16	186.36	70.21

displays the total operating revenues for fifteen U.S. airlines between 2007 and 2020, while Figure 3 illustrates sector-specific revenue trends over this period. The data indicate the growth of the U.S. airline industry leading up to the COVID-19 pandemic. Before the pandemic, all three sectors (FSCs, LCCs, and Others) experienced steady revenue growth, with slight declines observed in 2009 and 2010 for FSCs. Notably, LCCs exhibited faster revenue growth, nearly doubling from 2007 to 2019. As for FSCs, American, Delta, and United emerged as the dominant players following a series of mergers in the last decade. Among LCCs, Southwest maintained its lead, followed by jetBlue. Smaller carriers like Frontier, Spirit, Allegiant, Alaska, and Hawaiian also showed notable growth. However, in 2020, the pandemic severely impacted the airline industry, resulting in a 62.3% overall revenue decrease compared to 2019. FSCs and LCCs experienced revenue declines of 63.4% and 58.6%, respectively.

Turning our focus to employment,

Table 4. Total Full-time Employee Equivalents.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	71,818	70,925	66,519	65,506	66,522	64,529	59,532	61,527	98,885	101,504	103,101	102,942	104,216	78,309
Continental	40,948	40,630	38,720	37,760	36,797	0	0	0	0	0	0	0	0	0
Delta	47,286	47,420	46,372	76,742	80,158	78,498	75,918	77,928	82,000	83,974	85,987	88,243	89,758	66,444
Northwest	29,619	29,124	29,828	0	0	0	0	0	0	0	0	0	0	0
United	55,160	51,536	46,587	46,060	46,491	87,966	87,405	84,472	84,228	88,814	85,644	86,641	90,116	69,277
USAirways	34,256	32,683	31,340	30,876	31,551	31,241	32,132	32,843	0	0	0	0	0	0
Subtotal—FSCs	279,087	272,318	259,366	256,944	261,519	262,234	254,987	256,770	265,113	274,292	274,732	277,826	284,090	214,030
Southwest	33,680	34,680	34,874	34,230	36,104	43,840	42,955	44,169	47,395	51,037	54,815	56,262	58,263	52,199
jetBlue	9713	10,177	10,583	11,211	11,749	12,345	12,778	13,387	14,432	15,611	16,998	17,681	18,397	16,005
AirTran	8304	8259	8220	8229	7751	0	0	0	0	0	0	0	0	0
Frontier	5219	4939	4774	4309	4286	4422	3384	3932	2979	3412	3844	4234	4964	5005
Virgin America	0	980	1421	1770	2115	2381	2481	2489	2656	2934	3206	0	0	0
Spirit	2145	2410	1901	2351	2501	3725	3764	4164	4833	5756	7104	8381	9006	9175
Allegiant	1133	1330	1533	1585	1571	1799	1978	1938	2546	3146	3433	3904	4140	3953
Subtotal—LCCs	60,194	62,775	63,306	63,685	66,077	68,512	67,340	70,079	74,841	81,896	89,400	90,462	94,770	86,337
Alaska	9680	9628	8912	8649	8917	9176	9489	10,471	11,614	12,225	13,896	17,520	17,919	16,643
Hawaiian	3315	0	3635	3802	4265	4863	5251	5368	5520	6160	6661	7291	7458	5080
Subtotal—Others	12,995	9628	12,547	12,451	13,182	14,039	14,740	15,839	17,134	18,385	20,557	24,811	25,377	21,723
Total All Sectors	352,276	344,721	335,219	333,080	340,778	344,785	337,067	342,688	357,088	374,573	384,689	393,099	404,237	322,090

presents the total full-time employee equivalents for these airlines, with Figure 4 providing sector-specific employment trends. Employment numbers generally parallel the revenue data. In 2009 and 2010, FSCs witnessed slight declines in total full-time employees, whereas LCCs and other sectors maintained more stable employment growth. In 2019, American, United, and Delta led FSCs in employment, while Southwest and jetBlue stood out among LCCs. Allegiant, with lower operating revenues, had the smallest workforce. In 2020, the pandemic’s impact on air travel demand prompted widespread workforce reductions. Overall employment decreased by 20.3%, with FSCs cutting their full-time staff by 24.7%. Surprisingly, LCCs reduced their workforce by only 8.9%.

4.2. Cross-Sector Efficiency Comparison Test Results

Frontier Analyst software was used to build and run DEA models using the output-oriented BCC model, yielding yearly efficiency scores for the fifteen airlines from 2007 to 2020. To statistically analyze differences in efficiency scores across three airline sectors (FSCs, LCCs, and Others), the Kruskal–Wallis one-way ANOVA, suitable for non-normally distributed data, was employed. Data examination confirmed similar distributions among the groups, which meets the test’s assumption. Table 3 summarizes the Kruskal–Wallis ANOVA results for five employee categories. Significant differences in efficiency scores were observed for flight attendants (H = 11.04, p = 0.004), ground staff (H = 12.95, p = 0.002), maintenance staff (H = 59.93, p = 0.000), and management staff (H = 33.81, p = 0.000) among the airline groups. However, no significant difference was found for pilots (H = 4.662, p = 0.097).

Pairwise comparison tests were conducted for four employee types with significant differences, as presented in

Table 5. Kruskal–Wallis ANOVA results. (**: p ≤ 0.01).

Employee Type	Kruskal–Wallis H Test Statistic	p-Value	Pairwise Comparison
Pilots	4.662	0.097	N/A
Flight attendants	11.04	0.004 **	Pairs	Test Statistic	Sig.
			FSCs–Others	−7.565	0.353
			FSCs–LCCs	−19.434	0.001
			LCCs–Others	11.869	0.125
Ground staff	12.95	0.002 **	Pairs	Test Statistic	Sig.
			FSCs–Others	26.738	0.011
			LCCs–Others	35.968	0.000
			FSCs–LCCs	−9.230	0.228
Maintenance staff	59.93	0.000 **	Pairs	Test Statistic	Sig.
			FSCs–Others	50.109	0.000
			LCCs–Others	68.162	0.000
			FSCs–LCCs	−18.053	0.007
Management staff	33.81	0.000 **	Pairs	Test Statistic	Sig.
			FSCs–Others	42.755	0.000
			LCCs–Others	57.382	0.000
			FSCs–LCCs	−14.628	0.053

and Figure 5. Notably, LCCs and Others exhibited higher mean efficiency scores than FSCs for flight attendants. Among ground staff, the Others sector had a notably lower mean efficiency than LCCs and FSCs. For maintenance staff and management staff, significant differences emerged between all three airline sectors. LCCs and FSCs demonstrated markedly higher mean efficiency than the Others sector for both employee types.

It is imperative to clarify that these comparisons pertain to relative efficiency scores rather than the absolute inputs and outputs of airlines. In essence, they indicate how efficiently airlines can utilize their workforce resources to attain maximum outcomes when compared to other airlines. For instance, regarding pilot efficiency comparison, the findings reveal that despite variations in pilot-related inputs among the three airline groups, there is no significant difference in their capacity to efficiently manage pilots to achieve the maximum outcomes.

4.3. Efficiency Assessment Results

In this section, detailed efficiency comparisons were conducted for each airline from 2007 to 2020. Table 6, Table 7, Table 8, Table 9 and Table 10 present the efficiency results for all five models over the fourteen years. The efficiency scores are color-coded for easy categorization of efficiency levels. Green represents an efficiency score of 100%, indicating efficient airlines. Yellow, orange, and red indicate inefficiency on different levels. Yellow indicates light inefficiency (efficiency score between 90 and 100%), orange indicates moderate inefficiency (efficiency score between 80 and 90%), and red indicates severe inefficiency (efficiency score less than 80%). Those inefficiency levels inform us of the extent of airlines’ weaknesses and improvement needed in each employee type to make the inefficient airlines become efficient.

Pilot efficiency comparison: As shown in Table 6, most airlines consistently achieved good efficiency scores, with no significant differences among the three sectors: FSCs, LCCs, and Others. Notably, American, Delta, U.S. Airways, Southwest, Virgin America, Hawaiian, and Allegiant demonstrated efficient pilot utilization from 2007–2019. Although United, Northwest, Frontier, Alaska, and Spirit maintained efficiency for the most part, they had occasional inefficiencies, mainly at the yellow or orange levels. Specifically, United and Frontier faced yellow-level inefficiencies from 2014–2019. In contrast, Continental, jetBlue, and AirTran consistently struggled with pilot inefficiency, with jetBlue’s recent performance standing out as poor, hitting red and orange levels. In 2020, while most airlines maintained efficiency, jetBlue stood out as inefficient, with a score of 84.6%.

Table 6. Pilot Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	83.0%	86.0%	88.0%	92.3%	74.5%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	87.4%	100.0%	100.0%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	90.0%	100.0%	100.0%	96.2%	93.9%	95.1%	100.0%
U.S. Airways	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	92.6%	92.3%	86.1%	89.0%	86.8%	80.8%	76.6%	80.6%	76.6%	85.3%	79.6%	84.6%
AirTran	89.9%	91.6%	87.5%	97.4%	80.2%
Frontier	96.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	92.3%	99.3%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Alaska	88.6%	100.0%	93.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Hawaiian	100.0%		100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Spirit	93.3%	82.4%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Table 6. Pilot Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	83.0%	86.0%	88.0%	92.3%	74.5%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	87.4%	100.0%	100.0%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	90.0%	100.0%	100.0%	96.2%	93.9%	95.1%	100.0%
U.S. Airways	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	92.6%	92.3%	86.1%	89.0%	86.8%	80.8%	76.6%	80.6%	76.6%	85.3%	79.6%	84.6%
AirTran	89.9%	91.6%	87.5%	97.4%	80.2%
Frontier	96.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	92.3%	99.3%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Alaska	88.6%	100.0%	93.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Hawaiian	100.0%		100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Spirit	93.3%	82.4%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Flight attendant efficiency comparison: Table 7 reveals flight attendant efficiency results for airlines. American, Southwest, Frontier, and Allegiant consistently achieved industry benchmark efficiency levels. Delta, another FSC, performed well, with just one instance of yellow-level inefficiency in 2015, after which it maintained efficiency. However, the remaining airlines faced challenges. United struggled in the past five years, with inefficiency levels reaching orange since 2017. Non-existent FSCs Continental, Northwest, and U.S. Airways were highly inefficient before mergers, as was LCC AirTran, mainly at the yellow level. Alaska faced frequent inefficiency at the yellow level, and Virgin America was efficient until 2015 when it became inefficient at the orange level. jetBlue, Hawaiian, and Spirit had two instances of inefficiency, one at the red level and another at the yellow level. Notably, jetBlue and Hawaiian improved and achieved efficiency in 2019. In 2020, despite the pandemic, all the airlines were flight-attendant-efficient except for jetBlue (90.4%).

Table 7. Flight Attendant Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	81.1%	80.6%	73.3%	68.9%	60.1%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	94.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	89.6%	85.6%	79.4%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	95.6%	100.0%	84.7%	87.7%	88.7%	100.0%
U.S. Airways	100.0%	100.0%	98.8%	77.7%	76.9%	80.2%	97.3%	89.6%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	77.7%	100.0%	96.3%	100.0%	100.0%	90.4%
AirTran	100.0%	95.8%	100.0%	97.9%	95.8%
Frontier	98.5%	88.8%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	89.8%	100.0%	100.0%
Alaska	100.0%	100.0%	97.2%	94.6%	95.0%	95.9%	100.0%	100.0%	93.2%	100.0%	100.0%	100.0%	100.0%	100.0%
Hawaiian	100.0%		100.0%	100.0%	100.0%	90.9%	100.0%	100.0%	54.5%	100.0%	100.0%	100.0%	100.0%	100.0%
Spirit	100.0%	100.0%	77.7%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	96.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Table 7. Flight Attendant Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	81.1%	80.6%	73.3%	68.9%	60.1%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	94.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	89.6%	85.6%	79.4%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	95.6%	100.0%	84.7%	87.7%	88.7%	100.0%
U.S. Airways	100.0%	100.0%	98.8%	77.7%	76.9%	80.2%	97.3%	89.6%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	77.7%	100.0%	96.3%	100.0%	100.0%	90.4%
AirTran	100.0%	95.8%	100.0%	97.9%	95.8%
Frontier	98.5%	88.8%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	89.8%	100.0%	100.0%
Alaska	100.0%	100.0%	97.2%	94.6%	95.0%	95.9%	100.0%	100.0%	93.2%	100.0%	100.0%	100.0%	100.0%	100.0%
Hawaiian	100.0%		100.0%	100.0%	100.0%	90.9%	100.0%	100.0%	54.5%	100.0%	100.0%	100.0%	100.0%	100.0%
Spirit	100.0%	100.0%	77.7%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	96.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Ground staff efficiency comparison: Table 8 displays ground staff efficiency results for airlines over fourteen years. Most airlines had poor performance, with many inefficiency scores at the orange and red levels. Southwest was the sole consistently efficient airline from 2007–2019. Among non-existing airlines, Continental, Northwest, and U.S. Airways performed poorly before mergers. AirTran and Virgin America were efficient initially, then became inefficient. Among FSCs, Delta struggled with inefficiency, especially in the orange and red levels, but improved since 2018. American transitioned from inefficiency to efficiency after 2015, while United, once efficient, became consistently inefficient after 2015. Among LCCs and Others, jetBlue and Hawaiian consistently performed poorly with red-level inefficiencies. Frontier was initially very inefficient before becoming efficient. On the other hand, Allegiant was efficient until 2016, then slipped into red-level inefficiency for three years. Spirit’s and Alaska’s performance varied over the years. In 2020, United, jetBlue, Hawaiian, and Alaska were found inefficient.

Table 8. Ground Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	92.6%	93.6%	81.1%	88.6%	87.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	72.7%	66.9%	69.2%	66.4%	61.3%
Delta	93.8%	84.3%	85.8%	100.0%	100.0%	92.9%	95.1%	99.6%	94.0%	94.7%	96.3%	100.0%	100.0%	100.0%
Northwest	77.1%	68.0%	69.6%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	95.6%	100.0%	86.3%	87.3%	87.9%	75.9%
U.S. Airways	100.0%	90.7%	88.7%	87.8%	85.0%	76.6%	73.8%	73.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	89.2%	81.3%	83.1%	83.4%	81.0%	80.0%	77.7%	79.0%	79.4%	75.9%	77.3%	80.4%
AirTran	100.0%	100.0%	100.0%	98.8%	97.6%
Frontier	63.2%	62.1%	69.6%	68.1%	75.5%	59.7%	60.3%	57.8%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	89.8%	92.1%	84.2%
Alaska	91.0%	83.8%	91.7%	94.4%	100.0%	100.0%	99.8%	100.0%	93.2%	93.2%	90.7%	100.0%	100.0%	92.9%
Hawaiian	85.2%		80.3%	68.0%	59.3%	55.2%	51.1%	55.2%	54.5%	53.6%	53.3%	44.4%	42.1%	38.5%
Spirit	100.0%	100.0%	80.3%	76.0%	85.1%	83.9%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	70.1%	66.8%	54.8%	100.0%	100.0%

Table 8. Ground Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	92.6%	93.6%	81.1%	88.6%	87.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	72.7%	66.9%	69.2%	66.4%	61.3%
Delta	93.8%	84.3%	85.8%	100.0%	100.0%	92.9%	95.1%	99.6%	94.0%	94.7%	96.3%	100.0%	100.0%	100.0%
Northwest	77.1%	68.0%	69.6%
United	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	95.6%	100.0%	86.3%	87.3%	87.9%	75.9%
U.S. Airways	100.0%	90.7%	88.7%	87.8%	85.0%	76.6%	73.8%	73.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	89.2%	81.3%	83.1%	83.4%	81.0%	80.0%	77.7%	79.0%	79.4%	75.9%	77.3%	80.4%
AirTran	100.0%	100.0%	100.0%	98.8%	97.6%
Frontier	63.2%	62.1%	69.6%	68.1%	75.5%	59.7%	60.3%	57.8%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	89.8%	92.1%	84.2%
Alaska	91.0%	83.8%	91.7%	94.4%	100.0%	100.0%	99.8%	100.0%	93.2%	93.2%	90.7%	100.0%	100.0%	92.9%
Hawaiian	85.2%		80.3%	68.0%	59.3%	55.2%	51.1%	55.2%	54.5%	53.6%	53.3%	44.4%	42.1%	38.5%
Spirit	100.0%	100.0%	80.3%	76.0%	85.1%	83.9%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	70.1%	66.8%	54.8%	100.0%	100.0%

Maintenance staff efficiency comparison: Table 9 presents maintenance staff efficiency results, revealing American and Delta as highly efficient FSCs and jetBlue and Spirit as efficient LCCs. Northwest and Virgin America, among non-existing airlines, achieved efficiency. Continental, AirTran, and U.S. Airways consistently operated inefficiently, mostly at the red level. United initially struggled with maintenance staff efficiency but improved in the second half. Southwest consistently performed well among LCCs. Frontier and Allegiant’s efficiency records were inconsistent, while Alaska and Hawaiian consistently scored low with red-level efficiency. In 2020, United, Alaska, and Hawaiian were inefficient in using maintenance staff. These findings align with the test results in Section 4.2, showing differences across airline groups, with LCCs performing better than FSCs.

Table 9. Maintenance Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	65.9%	63.4%	62.5%	56.7%	53.2%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	100.0%	100.0%	100.0%
United	97.4%	98.1%	89.5%	79.0%	71.9%	99.9%	95.1%	89.9%	100.0%	100.0%	100.0%	100.0%	100.0%	88.3%
U.S. Airways	91.7%	82.5%	75.4%	70.8%	75.2%	76.7%	80.6%	75.2%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
AirTran	45.1%	46.9%	100.0%	100.0%	100.0%
Frontier	28.0%	30.1%	77.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	90.4%	100.0%	100.0%
Alaska	38.5%	38.6%	57.3%	60.4%	62.8%	64.7%	73.8%	78.3%	80.8%	78.3%	79.5%	88.0%	88.8%	85.9%
Hawaiian	100.0%		71.4%	63.7%	65.0%	57.3%	55.5%	59.0%	57.8%	51.5%	47.5%	51.8%	100.0%	37.8%
Spirit	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	60.3%	63.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Table 9. Maintenance Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	65.9%	63.4%	62.5%	56.7%	53.2%
Delta	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Northwest	100.0%	100.0%	100.0%
United	97.4%	98.1%	89.5%	79.0%	71.9%	99.9%	95.1%	89.9%	100.0%	100.0%	100.0%	100.0%	100.0%	88.3%
U.S. Airways	91.7%	82.5%	75.4%	70.8%	75.2%	76.7%	80.6%	75.2%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
AirTran	45.1%	46.9%	100.0%	100.0%	100.0%
Frontier	28.0%	30.1%	77.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	90.4%	100.0%	100.0%
Alaska	38.5%	38.6%	57.3%	60.4%	62.8%	64.7%	73.8%	78.3%	80.8%	78.3%	79.5%	88.0%	88.8%	85.9%
Hawaiian	100.0%		71.4%	63.7%	65.0%	57.3%	55.5%	59.0%	57.8%	51.5%	47.5%	51.8%	100.0%	37.8%
Spirit	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	60.3%	63.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Management staff efficiency comparison: As presented in Table 10, Southwest and Allegiant had been consistently management-staff-efficient. Among non-existing airlines, Continental and Virgin America performed well, while Northwest, U.S. Airways, and AirTran struggled with consistently low efficiency scores. Major carriers like American, United, and Delta faced efficiency challenges initially but improved in recent years, even during the pandemic. Among LCCs, Frontier and Spirit initially struggled but improved over time. In contrast, jetBlue, Alaska, and Hawaiian consistently operated inefficiently, with all three reaching the red level in 2020. Overall, the Others sector performed worse than LCCs and FSCs, aligning with the Section 4.2 test results.

Table 10. Management Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	94.9%	89.1%	87.1%	73.5%	71.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	100.0%	100.0%	100.0%	100.0%	100.0%
Delta	73.5%	76.5%	81.2%	100.0%	100.0%	100.0%	92.3%	95.4%	92.6%	94.4%	95.8%	100.0%	100.0%	100.0%
Northwest	57.8%	51.6%	62.8%
United	100.0%	100.0%	100.0%	98.1%	84.2%	92.6%	92.5%	95.4%	98.5%	100.0%	100.0%	100.0%	100.0%	99.9%
U.S. Airways	65.6%	61.6%	65.5%	60.1%	73.8%	54.9%	60.2%	100.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	88.0%	80.3%	65.4%	68.5%	68.5%	72.9%	73.7%	75.7%	83.7%	84.5%	88.8%	87.7%	91.8%	79.8%
AirTran	73.4%	89.0%	87.4%	73.2%	100.0%
Frontier	90.9%	91.0%	76.3%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	86.7%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Alaska	62.4%	59.5%	60.2%	72.7%	70.6%	75.2%	78.6%	69.8%	66.7%	69.6%	65.4%	80.0%	82.6%	77.6%
Hawaiian	85.5%		100.0%	83.4%	91.2%	87.2%	79.7%	79.8%	77.3%	69.2%	82.9%	73.7%	66.8%	42.5%
Spirit	88.2%	63.7%	81.5%	63.1%	64.3%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

Table 10. Management Staff Efficiency Results.

Airlines	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020
American	100.0%	100.0%	100.0%	94.9%	89.1%	87.1%	73.5%	71.6%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Continental	100.0%	100.0%	100.0%	100.0%	100.0%
Delta	73.5%	76.5%	81.2%	100.0%	100.0%	100.0%	92.3%	95.4%	92.6%	94.4%	95.8%	100.0%	100.0%	100.0%
Northwest	57.8%	51.6%	62.8%
United	100.0%	100.0%	100.0%	98.1%	84.2%	92.6%	92.5%	95.4%	98.5%	100.0%	100.0%	100.0%	100.0%	99.9%
U.S. Airways	65.6%	61.6%	65.5%	60.1%	73.8%	54.9%	60.2%	100.0%
Southwest	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
jetBlue	88.0%	80.3%	65.4%	68.5%	68.5%	72.9%	73.7%	75.7%	83.7%	84.5%	88.8%	87.7%	91.8%	79.8%
AirTran	73.4%	89.0%	87.4%	73.2%	100.0%
Frontier	90.9%	91.0%	76.3%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Virgin America			100.0%	86.7%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Alaska	62.4%	59.5%	60.2%	72.7%	70.6%	75.2%	78.6%	69.8%	66.7%	69.6%	65.4%	80.0%	82.6%	77.6%
Hawaiian	85.5%		100.0%	83.4%	91.2%	87.2%	79.7%	79.8%	77.3%	69.2%	82.9%	73.7%	66.8%	42.5%
Spirit	88.2%	63.7%	81.5%	63.1%	64.3%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%
Allegiant	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%	100.0%

4.4. Recommendations for Improvement

Table 11. Output Improvement Recommendations for Inefficient Airlines (Based on 2020 Efficiency Scores).

Employee Type	Airline	Output Variables	Actual	Target	Potential Improvement
Pilots efficiency	jet Blue—84.6%	Available seat miles	25,518	30,165.2	18.21%
		Revenue passenger miles	14,245	19,437.5	36.45%
		Total passenger revenue	1819	2150.3	18.21%
Flight attendant efficiency	jetBlue—90.4%	Available seat miles	25,518	28,241.0	10.67%
		Revenue passenger miles	14,245	15,765.1	10.67%
		Total passenger revenue	1819	2013.1	10.67%
Ground staff efficiency	United—75.9%	Available seat miles	60,868	80,244.6	31.83%
		Revenue passenger miles	38,273	52,737.1	37.79%
		Total passenger revenue	4804	6592.3	37.23%
	jetBlue—80.4%	Available seat miles	25,518	43,188.5	69.25%
		Revenue passenger miles	14,245	25,850.1	81.47%
		Total passenger revenue	1819	2263.6	24.44%
	Hawaiian—38.5%	Available seat miles	5946	30,367.9	410.73%
		Revenue passenger miles	3398	19,892.8	485.43%
		Total passenger revenue	467	1212.6	159.66%
	Alaska—92.9%	Available seat miles	29,282	42,002.5	43.44%
		Revenue passenger miles	16,110	25,299.0	57.04%
		Total passenger revenue	2013	2166.4	7.62%
Maintenance staff efficiency	United—88.34%	Available seat miles	60,868	80,845.5	32.82%
		Revenue passenger miles	38,273	43,323.2	13.20%
		Total passenger revenue	4804	6091.2	26.80%
	Alaska—85.9%	Available seat miles	29,282	38,807.6	32.53%
		Revenue passenger miles	16,110	21,982.1	36.45%
		Total passenger revenue	2013	2343.5	16.42%
	Hawaiian—37.77%	Available seat miles	5946	20,989.8	253.01%
		Revenue passenger miles	3398	12,613.3	271.20%
		Total passenger revenue	467	1236.4	164.75%
Management staff efficiency	jetBlue—79.85%	Available seat miles	25,518	34,889.7	36.73%
		Revenue passenger miles	14,245	18,959.1	33.09%
		Total passenger revenue	1819	2278.1	25.24%
	Alaska—77.63%	Available seat miles	29,282	39,292.4	34.19%
		Revenue passenger miles	16,110	21,205.0	31.63%
		Total passenger revenue	2013	2593.2	28.82%
	Hawaiian—42.53%	Available seat miles	5946	18,403.5	209.51%
		Revenue passenger miles	3398	10,549.2	210.45%
		Total passenger revenue	467	1098.2	135.16%

outlines improvement recommendations for jetBlue, primarily focusing on output enhancements. The “Actual” column displays the airline’s 2020 values, while the “Target” column presents the desired figures. The “Potential Improvement” column indicates the percentage change required. The recommendations for improvement are based on the comparison between actual values and target values. Since our DEA models are output-oriented, this study focuses mainly on output improvements.

The results show that to become efficient in using pilots, jetBlue needs to improve some outputs or inputs. More specifically, for pilot efficiency, jetBlue must aim to increase ASMs by 18.2%, RPMs by 36.5%, and total passenger revenue by 18.2% to achieve efficiency. Regarding flight attendance efficiency, jetBlue did not perform poorly during the pandemic; it mainly needs to enhance its ASM, RPM, and total passenger revenue by 10.7% to become efficient.

Additionally, four airlines need to make improvements to raise their ground staff efficiency scores. United, jetBlue, and Alaska seem to have similar recommendations. Specifically, they all need to increase ASM by 31–69%, RPM by 38–81%, and total passenger revenue by 7–37%. Hawaiian faces the most challenges since its outputs appear too low compared to the target values. The airline needs to improve ASM by 410.7%, RPM by 485.4%, and passenger revenue by 159.7%.

In 2020, given the impact of the pandemic, United, Alaska, and Hawaiian were inefficient in using maintenance staff. United and Alaska have similar efficiency scores of 88.34% and 85.9%, respectively. In contrast, Hawaiian has a much lower efficiency score of 37.8%. Table 11 shows the recommended improvements for these three airlines. Specifically, United and Alaska share similar goals of increasing ASM by about 32%, RPM by 13% and 36%, and passenger revenue by 26% and 16%, respectively. Hawaiian faces a much more challenging goal, needing to increase ASM by 253%, RPM by 271%, and passenger revenue by 165%.

Finally, for management staff efficiency, the Others group performed worse than LCCs and FSCs. Specifically, jetBlue and Alaska have similar goals, needing to increase ASM by 34–36%, RPM by 31–33%, and passenger revenue by 25–28%. Likewise, Hawaiian faces a more daunting task, needing to improve both ASM and RPM by over 210% and passenger revenue by 135%.

5. Discussions

This study has a notable finding that most airlines demonstrated relatively good efficiency when it came to pilots and flight attendants, reflecting the attention and investment given to these types of employees due to their direct passenger interactions. In contrast, those airlines showed considerably worse efficiency outcomes for ground staff, maintenance staff, and management staff, indicating that these employee groups, which work behind the scenes, may not receive the same level of attention. This imbalance poses a challenge to economic sustainability, as underperforming support functions can drive up operating costs, diminish operational effectiveness, and potentially compromise safety—all factors that are vital for long-term financial resilience. Therefore, it is imperative for airlines to develop targeted strategies that enhance the efficiency of these less visible yet essential employee groups, ensuring that every component of the workforce contributes to a more sustainable and economically viable operation.

The Kruskal–Wallis one-way ANOVA was conducted to compare the efficiency differences among three airline groups: FSCs, LCCs, and Others. The test results indicated significant differences in mean efficiency scores between the three airline groups, except for pilot efficiency. Overall, the Others group received lower mean efficiency scores than the LCCs and FSCs groups. Additionally, FSCs appeared to have slightly lower efficiency compared to LCCs, although this difference was only significant for maintenance staff efficiency.

In-depth comparisons of efficiency scores among airlines over time revealed interesting findings. Among the three existing FSCs, United faced more challenges in labor efficiency than the other two. For pilots and flight attendants, American and Delta maintained consistent efficiency over the years, whereas United received lower scores intermittently. This pattern was also observed for maintenance staff, where United was the only major carrier encountering difficulties. However, this airline made improvements and became more efficient in utilizing maintenance staff from 2015 to 2019. As for ground and management staff, all three major carriers displayed poor efficiency scores over time. Notably, American and Delta managed to implement necessary improvements to their operational strategies, leading to a recovery in their efficiency for ground and management staff. The results indicated that these two airlines consistently used all employee types efficiently, and even when they received low efficiency scores, they made the right decisions to regain efficiency. In 2020, when the industry was impacted by the pandemic, American and Delta maintained efficiency in labor usage, while United struggled with ground and maintenance staff efficiency, signaling some challenges ahead for the airline. The recommended improvement results indicated that to regain efficiency, United would need a progressive strategy to increase outputs (ASM, RPM, and passenger revenue) by at least 30%. United may need to develop a strategic plan to re-enter competitive markets.

Continental, U.S. Airways, and Northwest were FSCs that no longer exist due to their mergers with other carriers. Their labor efficiency records over the years were very poor, with low efficiency scores consistently at the orange or red level for most employee types. There were only three instances where these airlines were efficient: U.S. Airways was efficient with pilots, Continental with management staff, and Northwest with maintenance staff. In all other cases, those airlines were inefficient, which could partially explain their need to merge with other major carriers.

Among the group of LCCs, Southwest has performed exceptionally well overall. The airline has been consistently efficient with all employee types over the years, never receiving any efficiency score lower than 1. It remains the benchmark in this competitive market, which explains its ongoing success. Other airlines in this group have shown somewhat inconsistent operations. Except for jetBlue and AirTran, the remaining airlines operated well with pilots and, to some extent, flight attendants. These results indicate that those airlines have effective strategies for these employee types, focusing on flight experience and service for passengers. However, their efficiency records were low and inconsistent for other employee groups (ground, maintenance, and management staff). jetBlue and Spirit were only efficient with maintenance staff, while Allegiant was only efficient with management staff. Interestingly, before its merger, Virgin America had good efficiency records with maintenance and management staff. None of these airlines achieved good and consistent efficiency records for ground staff. Nevertheless, some managed to improve their operational efficiency, such as Allegiant, Frontier, and Spirit, which experienced low efficiency records for several years before regaining labor efficiency. This trend suggests that these airlines recognized their challenges and promptly implemented necessary improvements. AirTran was a defunct LCC, with consistently low records across all employee types, which could explain its merger with Southwest. Lastly, jetBlue is an existing LCC that requires better strategies for labor efficiency; aside from maintenance staff, it had low efficiency scores across all four other employee types, with particularly low scores in 2020 amidst the pandemic.

In the Others group, Alaska and Hawaiian exhibited similar poor efficiency outcomes. Except for pilot efficiency, both airlines maintained mediocre efficiency records for flight attendants and very poor efficiency records for ground, maintenance, and management staff over the years, including in 2020. Their efficiency records consistently reflected red indicators. The poorer performance of this group, compared to the LCCs and FSCs, aligns with the comparison test results. These airlines certainly need to reconsider their operational strategies.

Further analysis results indicated that jetBlue, Hawaiian, and Alaska must significantly enhance their operations. jetBlue and Alaska need to boost outputs (ASM, RPM, and passenger revenue) by 35% to 80% to regain efficiency. However, Hawaiian faces the greatest challenge, requiring improvements of almost 490% in ASM and RPM and nearly 170% in passenger revenue for efficiency. Note that these are smaller carriers with limited fleets and market shares, making these objectives very challenging. The post-pandemic air travel market shows a significant increase in demand, but many airlines struggle to meet this demand due to labor shortages. Additionally, the concerning number of flight delays and cancellations complicates efficient airline operations.

Possible reasons for airline labor inefficiency, particularly among less visible employee groups, include insufficient training and skill development programs, limited investment in technology and automation, and challenges in labor relations. Additionally, inefficient operational strategies, suboptimal resource allocation, and poor management practices contribute to the underutilization of these employees. External factors—such as volatile market conditions, intense competition, and complex regulatory requirements—further exacerbate these challenges. Overall, these factors lead to resource misallocation and disruptions in airline operations, undermining both service quality and economic sustainability. Since labor efficiency is a critical component of reducing operating costs and ensuring long-term financial viability, addressing these inefficiencies typically requires a combination of data analysis, process improvement initiatives, strategic investments in technology and training, and robust management practices to optimize workforce performance and support sustainable economic growth.

6. Conclusions

Labor efficiency is crucial for both operational success and economic sustainability in the airline industry, given the central role of employees in driving performance and managing costs. While previous airline efficiency models incorporated labor as an input, it was just one of many inputs. Consequently, evaluating airline efficiency based on labor utilization has remained challenging, as the impact of other inputs could distort the results. This study bridges this gap by conducting a comprehensive assessment of airline efficiency exclusively from a labor workforce perspective. The DEA models incorporate diverse aspects of the labor workforce, including employment size, wages, salaries, benefits, and workload. This study produces relative efficiency scores for all U.S. airlines longitudinally from 2007 to 2020, not only shedding light on areas of labor inefficiency but also providing a foundation for strategies that enhance economic sustainability by aligning workforce management with long-term financial and operational goals.

A noteworthy revelation from this study is the significant impact on ground, maintenance, and management staff. Most airlines seemed to prioritize pilots and flight attendants, resulting in high efficiency scores for these categories, given their direct interaction with passengers during flights. Conversely, many airlines grappled with maintaining efficiency for the less visible employee groups, including ground, maintenance, and management staff. To improve efficiency in these areas, airlines must significantly improve their outputs based on the targets identified by the DEA results. The potential risk is that failure to meet these output targets might lead to reduced employment and employee compensation and benefits (see Appendix A). Ground staff are crucial for passenger, cargo, and aircraft handling, while maintenance staff oversee aircraft maintenance. Cutting these labor resources could compromise safety and operational effectiveness. Moreover, airlines have already laid off many employees due to the COVID-19 pandemic, making it challenging to meet the increased post-pandemic demand. Airlines need tailored strategies for these challenges, as a one-size-fits-all approach may not suffice. Smaller carriers like jetBlue, Alaska, and Hawaiian face even greater hurdles, requiring substantial increases in ASM, RPM, and passenger revenue. Achieving these goals, given their limited market share and smaller fleets, necessitates organizational restructuring and possibly further industry mergers.

Airlines can establish multiple strategies to improve outputs that contribute directly to economic sustainability. Such strategies include boosting ASM, RPM, and passenger revenue through initiatives that strengthen operational efficiency, expand market presence, and enhance customer satisfaction. Key approaches may involve fleet optimization—such as acquiring newer, fuel-efficient aircraft while retiring older, less efficient models—and expanding routes by identifying underserved or high-demand markets. Marketing and pricing strategies, including dynamic pricing and loyalty programs, can stimulate demand and boost passenger revenue, while improved customer experiences drive repeat business.

Furthermore, partnerships such as code-sharing agreements and alliances can broaden network reach, and advanced revenue management systems can optimize pricing and seat allocation. Collaborations with travel agencies, online platforms, and aggregators further expand ticket sales and customer reach. In addition, sustainability initiatives that emphasize environmental stewardship, along with flexibility and adaptability in response to market dynamics, contribute positively to both reputation and revenue—integral components of long-term economic sustainability.

Airlines can also develop strategies to enhance labor efficiency for ground, maintenance, and management staff. They can optimize operational processes through workflow streamlining and reduction of manual tasks. Additionally, investing in advanced training and skill development programs ensures that employees are well-equipped to perform efficiently. Airlines should also embrace modern technology and automation tools, such as maintenance management software and automated check-in systems, to streamline tasks and reduce manual labor. Efficient resource allocation and motivating employees can contribute to labor efficiency. Finally, encouraging collaboration, cross-training, and integrating environmental sustainability initiatives all contribute to long-term economic sustainability by reducing costs and promoting operational resilience.

It is important to note that this paper primarily focuses on the labor workforce to examine airline labor efficiency. Therefore, other input variables, such as capital, fuel costs, facilities, materials, and operating costs, have not been included. This deliberate omission allows us to exclusively analyze how labor inputs influence airline efficiency. A comprehensive model incorporating various inputs would potentially distort or alter the effects of employee-related inputs. Additionally, this study used data only until the end of 2020, which reflects the peak impact of the pandemic and associated travel restrictions. Consequently, our evaluation of airlines’ labor efficiency is limited to this timeframe, and post-pandemic conditions cannot be assessed due to data unavailability beyond 2020. Hence, the recommendations provided for airlines to enhance their labor efficiency pertain to the mid-pandemic scenario characterized by travel restrictions. Nonetheless, the analysis provides valuable insights into the weaknesses and necessary improvements that inefficient airlines may face during similar disruptive events in the future, highlighting the importance of sustainable labor practices in maintaining economic viability.

Future research endeavors could expand the analysis by incorporating new data from 2021 and 2024, enabling the calculation of airline labor efficiency scores during and after the pandemic. Such an extended dataset would facilitate comparisons with previous years and provide a clearer understanding of the impact of labor-related changes on overall efficiency and economic sustainability. Furthermore, it would be valuable to assess how strategic labor improvements influence the competitive dynamics within the airline industry. Researchers might also explore multi-stage models that investigate how the labor workforce affects various resources and functions within airlines, ultimately impacting overall outputs. Such models could offer a more comprehensive understanding of the intricate relationships at play, deepening insights into how labor efficiency contributes to sustained economic performance in a challenging and dynamic market environment.