Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

Contents lists available at ScienceDirect

Transportation Research Interdisciplinary Perspectives

journal homepage: https://www.journals.elsevier.com/transportation-research-
interdisciplinary-perspectives

COVID-19 and airline employment: Insights from historical uncertainty

shocks to the industry
Joseph B. Sobieralski
School of Aviation and Transportation Technology, Purdue University, West Lafayette, IN, USA

A R T I C L E I N F O A B S T R A C T

Article history: The purpose of this paper is to analyze the effects of uncertainty shocks on airline employment in the light of the cur-
Received 6 April 2020 rent global pandemic. The airline industry has faced many threats throughout history, but none quite as rapid and
Accepted 11 April 2020 severe as the one posed by the spread of COVID-19. One constant during uncertainty shocks and industry downturns
Available online 1 May 2020
is that airline labor bears the brunt of the decline. As the industry reduces capacity amid the increase in travel restric-
tions, the post-stimulus impacts to airline labor are not known. Using time series analysis, the dynamics of historical
Keywords:
Airline labor
uncertainty shocks to the industry are examined. During periods of uncertainty shocks, the estimated job loss is nearly
COVID-19 7% of the airline workforce with an upper bound of over 13%. Major airline employment is most impacted, while low-
Recessions cost and regional airline employment is least impacted. The hardest hit employees are ones related to passenger han-
Uncertainty dling and flight operations, while management employees fair slightly better during these uncertain periods. Further,
Unemployment recovery following uncertainty shocks is estimated to take between 4 and 6 years. Overall, the labor impacts to the
airline industry from uncertainty events are substantial and provide insight into the expected industry job loss from
COVID-19.
© 2020 The Author. The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction The interaction between pandemics and economic dynamics has been
studied in theoretical modelling by Eichenbaum et al. (2020). Their
The global pandemic from the spread of COVID-19 has created uncer- model suggests that containment policies that promote reductions in con-
tainty for many as quarantines, hospitalizations, and tragically, related- sumption and work can reduce the number of deaths attributed to a pan-
deaths continue to growth. Travel restrictions have reduced the mobility demic. These policies unfortunately negatively impact the economy and
of individuals across the globe and the associated negative impacts have worsen the effects of the subsequent recession. Another study, Barro et al.
propagated to numerous industries. The transportation industry has been (2020), uses the Spanish flu and WWI deaths as proxies to quantify the eco-
especially hard hit as air travel continues to shrink due to flight cancella- nomic impacts a nation could expect during a pandemic. Their analysis es-
tions and capacity reductions. The airline industry has experienced a timates that a nation on average will experience a 6% decline in
decrease in capacity of roughly 60–80% at major carriers (Josephs, consumption and an 8% decline in GDP. As the spread of COVID-19 con-
2020). As demand wains, further reductions by airlines are on the horizon. tinues, the expected decline in consumption is being realized as jobless
The indefinite timeline for removing social distancing and travel restric- claims reach nearly 10 million in March 2020 (Wolfers, 2020). Barro
tions has increased uncertainty for the industry as a whole. In the U.S., et al. (2020) also find increased stock market and interest rate volatility
the government has passed a large stimulus bill to assist the crippled indus- during these periods. Their work implies that situations similar to the cur-
try (Gilbertson, 2020). Despite this stimulus to aid the wounded industry, rent COVID-19 panic can result in financial conditions much like the
the labor impacts of COVID-19 in the industry following expiration of Great Recession. Atkeson (2020) notes that models of the spread of
stimulus-imposed requirements remain to be seen. This paper provides COVID-19 suggest that social distancing and the associated economic
novel insight into the effects of uncertainty shocks to airline labor by exam- tradeoffs could last for nearly 18 months. Additionally, McKibbin and
ining the historical relationship between the industry and uncertainty Fernando (2020) use data from China to model the potential impacts
shocks such as pandemics. from the COVID-19 spread. They find rough estimates of a 1%–5% shock
to consumption, an equity risk premium shock of 1.07%–1.33%, and a
labor supply shock of up to 1.4%.
The true economic and societal impacts from the COVID-19 pandemic
E-mail address: sobieral@purdue.edu.. are still materializing; however, the uncertainty surrounding the pandemic

http://dx.doi.org/10.1016/j.trip.2020.100123
2590-1982/© 2020 The Author. The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/).
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

is evident and increasing as nations attempt to contain and mitigate the 2. Empirical strategy
spread of the virus. Uncertainty is noted as a key driver in numerous reces-
sions and uncertainty related economic shocks can propagate over time Data are used from U.S. airline and government datasets from 1991 to
(Caggiano et al., 2014). These uncertainty shocks can be caused by many 2019. Air travel data are from the Bureau of Transportation Statistics
factors to include financial crises, terror-related events, disease outbreaks, T-100 database. This data includes revenue passenger miles (RPMs), avail-
and natural disasters. The fluctuations in uncertainty create a distorted able seat miles (ASMs), and departures for each airline. Airline employment
view for individuals, businesses, and policymakers alike. Empirical eco- data are provided by airline Form 41 data as well as the data provider
nomic research has addressed the topic of uncertainty shocks with respect Cirium. Airlines are sorted into three categories: major, low-cost, and
to macroeconomic conditions. An increase in uncertainty tends to reduce regional airlines. Within each airline, employee data was sorted into six dif-
real activity in an economy (Basu and Bundick, 2017). The contraction in ferent categories to provide a more detailed understanding of employees
real macroeconomic activity is caused by precautionary savings by individ- impacted. Many skilled airline employees at airlines (e.g. pilots, mechanics)
uals and businesses as well as the subsequent job losses. The Great Recession are represented by labor organizations with lower levels of representation
was driven by financial uncertainty in the market even as policymakers re- of low skilled airline workers (von Nordenflycht and Gittell, 2013).
duced policy rates (Ludvigson et al., 2015). This uncertainty overshadowed Table 1 provides a description of the employee categories and percentage
policymaker action given the constraints imposed by a federal funds rate makeup of the airline workforce. Economic variables such as gross domes-
near 0% (Caggiano et al., 2017). The current state of the economy is also tic product (GDP) and crude oil prices are from the St. Louis Federal Reserve
noted as a determinant of the responsiveness of actors to uncertainty Economic Data. Stock market index data are from the CRSP database. Fig. 1
(Alessandri and Mumtaz, 2019). Gross domestic product growth and global- provides a macro viewpoint of the relationship between total air transpor-
ization impact an economy's responsiveness to these uncertainty shocks tation related employment, the measure of airline supply (ASMs), and a
(Bloom, 2014). Globalization has created an interwoven system of economies volatility index (measured by the VIX volatility index discussed later).
connected by the robust air transportation network (Mahutga et al., 2010). Until 2001, air transportation related employment followed the trend of
The airline industry has been threatened by global crises throughout growth in air transport; however, after 9/11 employment has declined
history. From the oil embargo, airline deregulation, terrorist attacks, the while air transport has on average increased. Further, we notice that vola-
industry has weathered many storms (Davies, 2016). The survival of an air- tility spikes, as measured by the VIX, have a slight aftereffect on ASMs,
line during and after these downturns is often related to the airline's busi- but employment appears to follow a somewhat similar trend to average
ness model. Franke and John (2011) note that the attacks on 9/11 caused volatility over time.
a significant reduction in air travel and airlines struggled to remain opera- The primary model used in this study is the vector autoregression (VAR)
tional. As the spread of SARS followed these events, demand only returned model. Introduced by Sims (1980), the VAR models the joint dynamics of
to pre-2001 levels in 2004 (Franke and John, 2011). During this period, low economic variables in time-series analysis. VAR models are an extension
cost carriers gained market share as major air carriers struggled to recover of univariate autoregressive models where all variables in the multi-
as pricing competition increased (Tan, 2016). Rising fuel prices also hin- equation system are treated as endogenous (Greene, 2011). The VAR
dered the recovery of the industry following 2001. Major airlines adjusted model without constraints on the coefficients is considered a seemingly
capacity to compensate for this increased cost leaving lower cost airlines unrelated regression model and regression on each equation provides the
opportunities to enter new markets (Sibdari et al., 2018). The Great Reces- maximum likelihood estimates for the coefficients of the lagged variables.
sion has further changed the network and routing structure of airlines as The VAR model allows for feedback between variables of interest in a
they attempt to mitigate the impacts of the decline in demand by cutting multi-equation system and affords the opportunity to analyze the real
capacity (Pearce, 2012). Less noted in the literature are the industry work- effects of uncertainty shocks in a dynamic system. Several measures of un-
force changes due to these shocks. Following 9/11, airlines moved to re- certainty shocks will be used in this analysis. The first measure of uncer-
structure their businesses to reduce both fixed and variable costs (von tainty is the stock market volatility measured by VIX. The VIX is the most
Nordenflycht and Gittell, 2013). These efforts were marked by employee commonly used measure of uncertainty shocks in the macroeconomic liter-
layoffs and renegotiation of labor contracts. Even as airline demand was re- ature (e.g. Bloom et al., 2007; Caggiano et al., 2018). It provides a forward-
covering, the industry was still plagued by continual job loss (Goodman, looking measure of uncertainty over the next period (Bloom, 2014). The
2008). Bankruptcies and mergers during this period furthered the work- second measure is the index of consumer expectation (ICE) from the Uni-
force reductions and labor contract renegotiations. The surviving em- versity of Michigan (Colombo, 2013; Leduc and Liu, 2016). This index is
ployees were faced with wage and benefits cuts that varied depending derived from surveys of households asking about expectations related to
upon occupation and ranged from 9% to 50% (von Nordenflycht and future economic and business conditions. The ICE allows for a forward-
Gittell, 2013). Air travel demand levels returned to pre-recession levels looking indicator of movement in consumer behavior which is akin to a
within 18 months, signaling that even deep crises eventually correct. The shock in aggregate demand. Finally, we also build an uncertainty indicator
current COVID-19 pandemic has disrupted the airline industry with the un- similar to the one constructed in Bloom (2009). Using information of uncer-
certainty of reduced capacity, weak demand, and an indefinite return to tainty events related to volatility in the stock market (Bloom, 2009) and
pre-2020 levels. The impact of the pandemic on the airlines is a continual volatility in the airline industry (Franke and John, 2011), the built indicator
source of debate for policymakers given the large economic impact and im-
portance of the industry; however, the discussion concerning the post-
stimulus airline labor impacts is typically an afterthought. This paper will Table 1
add to the policy discussions by providing extensive insight into the Airline employee categories.
expected impact on airline labor from the current crisis. Additionally, no
Category Example job title % of total airline employment
known study uses historical industry data and econometric modelling to for carrier type
estimate the impact of uncertainty shocks on major, low-cost, and regional
Major Low-cost Regional
airline labor. This study represents the first attempt to model the impact of
uncertainty shocks to airline labor. Cargo handling Cargo or freight agent 9% 1% 1%
Aircraft handling Line Service technician 7% 14% 13%
The remaining sections of this paper are organized as follows: the next
Maintenance Aircraft mechanic 11% 5% 13%
section provides a description of the data and the framework for modelling Passenger handling Passenger service agent 45% 48% 37%
the uncertainty shocks to the industry and airline labor. Section 3 provides Aircrew Pilot 14% 18% 28%
the results of the analysis and places them in the light of the current pan- Management Manager 7% 6% 4%
Other Security 7% 8% 4%
demic. Finally, Section 4 provides insight to policymakers and labor leaders
as well as paths for future work. As of 2019.

2
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

(a)

(b)
Fig. 1. Data from 1990 to 2020 displaying (a) total U.S. airline ASMs and VIX volatility index and (b) all air transportation employees and VIX volatility index.

will take the value 1 when an uncertainty event occurs and 0 otherwise. The VAR model used to examine the uncertainty shocks and their rela-
Table 2 displays the list of events used in this study to develop this uncer- tion to economic variables is given by:
tainty indicator.
X
p
yt ¼ V þ Γ i Δyt−i þ βX t þ ut ð1Þ
Table 2 i¼1
Uncertainty Events.
Event Year(s) Event type where yt is the vector of endogenous variables ordered uncertainty shocks
Gulf Crisis/Gulf War I 1990/1991 War indicator, log of GDP, log of airline employment, V is the vector of con-
Asian Crisis 1998 Economic stants, Γ is a matrix of coefficients, β is a matrix of coefficients, Xt is a vector
9/11 Attacks 2001 Terrorism of exogenous variables, and ut is the vector of white noise. The order of the
SARs 2003 Epidemic variables in y are based on economic theoretic assumptions that shocks
Great Recession 2008/2009 Economic
impact the stock market instantaneously, followed by prices, then real

3
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

quantities such as output and employment. The ordering of these variables Table 4
will be varied during checks for robustness. The time series properties of the Cointegration tests.
data are important to address before proceeding with any estimation. Hypothesis r=0 r≤2 r≤3 r≤4
To estimate the parameters of the VAR, the variables used in the estima- ASM specification Trace Statistic 63.76⁎ 36.52⁎ 18.38⁎ 6.68⁎
tion are required to be covariance stationary. The error-correction repre- Log Likelihood 2991.38 2999.82 3001.32 3002.57
sentation of a VAR, a vector error correction model (VECM), may be EMP specification Trace Statistic 55.35⁎ 35.26⁎ 23.47⁎ 3.78⁎
more appropriate if the variables in yt are not covariance stationary. The Log Likelihood 2972.96 2975.96 2976.31 2977.42
augmented Dickey-Fuller (ADF) test will be conducted to investigate the ⁎ Indicates rejection at the 5% level. Variables include GDP, OIL, VOL with lag
degree of integration of the variables (Elliott, 1998). Additionally, a test order of four.
for the presence of a unit root in the variables developed in Phillips
(1987) and Phillips and Perron (1988) will be used to account for the poten-
tial heteroskedasticity in the residuals. The results of these tests on the five these uncertainty shocks, economic studies suggest that employment will
endogenous variables (including alternate measures of uncertainty) are decline as production falls (Altug et al., 2012; Bhadra, 2009; Dixit, 1997).
displayed in Table 3. The null hypothesis of the ADF test is that the process These results provide additional motivation to investigate the change in
is a random walk or that a unit root does exist. We fail to reject the null employment from the related shocks. We now examine the change in air-
hypothesis for GDP, EMP, and OIL variables in levels, but reject the null line employment from an uncertainty shock.
hypothesis for ASM and all uncertainty variables in levels and the first The IRFs for the three categories of airlines are displayed in Fig. 3. An
differences. The ADF test suggests that the ASM and all uncertainty vari- uncertainty shock, as measured by the uncertainty indicator, has a negative
ables in levels and the first differences of GDP, EMP, and OIL are I(0). The impact on major airline employment and the decline is sustained through
results of the Phillips-Perron test also indicate that the variables GDP, the remaining periods. This result is explained by the large financial expo-
EMP, and OIL are I(1), while ASM and all uncertainty variables are I(0). sure of major airlines and the larger payroll liabilities of these airlines.
Therefore, these results combined suggest that all of the series are I Major airlines are more greatly impacted by these shocks due to their inflex-
(1) with the exception of ASM and all uncertainty variables. ible business model versus low-cost carriers (Neal and Kassens-Noor, 2011)
Given these series are I(1) in levels, a common trend may exist. If the This inflexibility negatively impacts employment at these airlines as they
variables exhibit a common long-run trend, then the variables are consid- shed employees along with capacity during uncertainty shocks. This result
ered cointegrated. When one or more cointegrating relationship exists in would lead us to expect that major airline employees will receive the brunt
the series, then estimations using linear VAR models would provide spuri- of industry job loss following an uncertainty shock such as the recent pan-
ous long-run results and a VECM would be more appropriate. Therefore, demic. Low-cost carriers experience a sharp decline of approximately
testing for cointegration is conducted using the Johansen-Juselius test 0.035% but recover quickly most likely as a result of their business model
(Johansen, 1988; Johansen and Juselius, 1990). The Johansen maximum and lower employee wages. The low-cost model affords these carriers the
likelihood approach is used to test for cointegration, and the appropriate ability to offer lower fares which would be more attractive during periods
variable lag length is determined by using an Akaike Information Criterion of uncertainty (Cho and Min, 2018). This result aligns with previous find-
(AIC), likelihood ratio test, and Schwarz Criterion (SC). The appropriate lag ings that low-cost carriers rebounded quickly to 9/11 following an initial
length is determined to be four and using this lag length will reduce the pos- decline (Franke, 2004). Therefore, we would expect that employment at
sibility of over parameterization. Table 4 summarizes the results of the these airlines would only suffer in the short run following the uncertainty
cointegration test. The results suggest that we fail to reject the null hypoth- shock such as this pandemic. Regional airlines display an increase in
esis of no integration for each of the specifications. Since these results sug- employment following an uncertainty shock with greater fluctuations in
gest no cointegration exists, then we employ the VAR model with lag order total employment. This result could be caused by major airlines using
of four. regional carriers to operate on routes with smaller, more efficient aircraft
and lower relative ASMs. The contracting of regionals by major airlines
3. Results has been well documented in the literature to aid in shedding costs for
major airlines (Tan, 2018). Therefore, employment at regional airlines is
The impulse response function for total industry ASMs is provided in expected to increase as travel restrictions are lifted, the uncertainty shock
Fig. 2. An initial decline of nearly 0.04% in all ASMs in the U.S. occurs fol- is dated, and major airlines attempt to further reduce costs by contracting
lowing an uncertainty shock. This decline in ASMs is sustained for nearly with regional carriers.
five periods which corresponds to previous literature describing industry
shocks (Franke and John, 2011; Jean and Lohmann, 2016). Pre-shock
ASM levels eventually return; however, this return to pre-shock levels is
sluggish as seen in the IRF and noted in the aforementioned studies.
Given the contraction in air transportation production (ASMs) following

Table 3
Unit-roots tests.
Variables Dickey-Fuller Phillips-Perron

Levels 1st Levels 1st Difference

Difference
μa
t b
μa tb

GDP −2.696 −18.306⁎ −10.410 −1.982 −359.543⁎ −18.336⁎

ASM −3.858⁎ −23.965⁎ −23.689⁎ −3.552⁎ −436.360⁎ −24.053⁎
EMP −1.382 −11.272⁎ −2.206 −1.049 −204.189⁎ −11.603⁎
OIL −2.059 −12.681⁎ −6.280 −1.792 −216.650⁎ −12.591⁎
VOL (VIX) −5.903⁎ −21.787⁎ −61.634⁎ −5.809⁎ −363.828⁎ −22.490⁎
VOL (ICE) −4.451⁎ −9.284⁎ −25.512⁎ −5.203⁎ −144.887⁎ −9.419⁎
VOL −4.275⁎ −19.066⁎ −21.433⁎ −4.571⁎ −477.913⁎ −19.629⁎ 4
(Index)

Phillips-Perron test statistic without a trend, μ, and with a trend, t.

a,b

⁎ Indicates rejection at the 1% level. Fig. 2. Change in ASMs from an uncertainty shock.

4
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

(a)

(a)

(b)

(b)

(c)
Fig. 4. IRFs for varying uncertainty shock measures for (a) major, (b) low-cost, and
(c) regional airlines.

major airlines regardless of the uncertainty measure used. Using the VIX,
major airlines experience a steeper initial decline in total employment fol-
lowing an uncertainty shock, low-cost airline employment appears more
volatile, and regional airline employment is almost unaffected. The ICE
metric produces a similar declination as observed with the uncertainty
index for major and low-cost airlines with a slightly deeper negative impact
in latter periods for major airlines. The constructed index provides an aver-
(c) aged impact on employment for major airlines and low-cost airlines relative
to VIX and ICE. The regional airline employment IRF using the uncertainty
Fig. 3. The response of employment to an uncertainty shock for (a) major, (b) low-
index aligns with the results from the VIX and ICE when examining the
cost, and (c) regional airlines.
interval created by two standard errors. Using these results, estimates of
the total impact of these shocks to employment are calculated. The baseline
The measure of uncertainty shocks using the uncertainty index is now estimated job loss is over 32,000 employees. Using the two standard errors,
tested against the other measures of uncertainty described in Section 2. the impact is estimated to range between 5500 and 64,000. That corre-
These checks of robustness allow for a more stringent examination of the sponds to an approximate decline of nearly 7% of the airline workforce
impact of uncertainty that aligns with common measures in the literature. with a range between 1%–13%.
Fig. 4 provides the combined IRFs for total employment at major, low- Examining the employment impact of uncertainty shocks for different
cost, and regional airlines. Uncertainty shocks have a negative impact on employee categories provides insight into the occupations facing the

5
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

greatest threat from uncertainty shocks. Figs. 5–7 provide the IRFs for dif- are observed following these uncertainty shocks. These declines are evident
ferent airline category employees at major, low-cost, and regional airlines. in the literature as furloughs of these employees at major airlines typically
Uncertainty shocks cause an initial decline in cargo handling employees at occur in the early stages of the shock. These occupations are unaffected in
major airlines likely due to the decline in capacity. Passenger handling oc- later periods most possibly due to skilled nature of the workforce and
cupations such as ticket or gate agent see the largest shocks and volatility requirement of a minimum number of pilots, flight attendants, and aircraft
from uncertainty shocks. Comprising >45% of the total major airline work- maintainers. Management positions show no statistically significant impact
force, these occupations appear to be the first line of cost shedding for air- from uncertainty shocks at major airlines possibly due to the low propor-
lines during uncertainty shocks. Initial declines in maintenance and aircrew tion of employees in this category.

(i) (ii)

(iii) (iv)

(v) (vi)
Fig. 5. IRFs for major airline individual employee categories (i) cargo handling, (ii) aircraft handling, (iii) maintenance, (iv) passenger handling, (v) aircrew, and (vi)
management.

6
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

(i) (ii)

(iii) (iv)

(v) (vi)

Fig. 6. IRFs for low-cost airline individual employee categories (i) cargo handling, (ii) aircraft handling, (iii) maintenance, (iv) passenger handling, (v) aircrew, and (vi)
management.

For low-cost airlines, an uncertainty shock has small negative im- following an uncertainty shock. This result is consistent with the in-
pacts on aircrew, but these are quickly recovered as operations for crease in operations experienced by the transition from major airline
low-cost airlines tend to increase post shock. The increase in operations flights to contracted regional carrier operations. As capacity increases
in later periods as low-cost carriers increase capacity following the for these carriers due to their low operating costs relative to major air-
shock increases employment in passenger handling, aircrew, and man- lines, then demand for aircrew and passenger handling employees
agement. These increases in employment post-shock appear to diminish increases. This spike in employment is tempered in later periods as
in later periods as major airlines increase their capacity and low-cost major airlines recover from the uncertainty shock and regional airline
airlines slow expansion (Atallah et al., 2018). Examining the impacts employment becomes less attractive. The other employee categories at
on regional airlines employees, cargo handling, passenger handling, regional airlines, to include management and maintenance are found
and aircrew experience an increase in employment in the early periods to not be statistically impacted by the uncertainty shocks.

7
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

(i) (ii)

(iii) (iv)

(v) (vi)

Fig. 7. IRFs for regional airline individual employee categories (i) cargo handling, (ii) aircraft handling, (iii) maintenance, (iv) passenger handling, (v) aircrew, and (vi)
management.

4. Conclusion this decline in capacity can last for years. The uncertainty shock due to
the events of September 11, 2001 caused a halt in air traffic throughout
The pandemic arising from the spread of COVID-19 has caused airlines the U.S. that lasted several days. Although air travel has yet to be
to reduce capacity and lobby the government for emergency relief. As these completely curtailed, COVID-19 is causing airlines around the globe to
reductions in capacity continue to grow and the post-stimulus employment reduce capacity due to travel restrictions and/or out of diminishing de-
requirements expire, the uncertain future of the airline workforce is appar- mand. The results of this analysis suggest that employees at major airlines
ent. To provide insight into the movements of airline employment during will be the most impacted during uncertain times such as the one faced
uncertainty shocks to the industry, this paper examined the dynamic inter- today by the global pandemic. The major airlines' size and financial expo-
actions among economic and industry variables when exposed to these sure increase their vulnerability to uncertainty shocks throughout history.
types of shocks. As shown, ASMs decline during uncertainty shocks and This analysis has shown that employment reductions are the likely response

8
J.B. Sobieralski Transportation Research Interdisciplinary Perspectives 5 (2020) 100123

these airlines have to uncertain times such as the one facing the industry Caggiano, G., Castelnuovo, E., Pellegrino, G., 2017. Estimating the real effects of uncertainty
shocks at the zero lower bound. Eur. Econ. Rev. 100, 257–272. https://doi.org/10.1016/
today. In total, approximately 7% of the airline workforce will be displaced j.euroecorev.2017.08.008.
by these shocks. Due to the global nature and severity of the shock of Caggiano, G., Castelnuovo, E., Figueres, J.M., 2018. Economic policy uncertainty spillovers in
COVID-19, the impact may be of the magnitude near the upper bound of booms and busts. Oxf. Bull. Econ. Stat. https://doi.org/10.1111/obes.12323.
Cho, W., Min, D.J., 2018. Longitudinal examination of passenger characteristics among airline
the estimates with a reduction of over 13% of the airline workforce. types in the US. J. Air Transp. Manag. 72, 11–19. https://doi.org/10.1016/j.
Another finding is that the low-cost and regional airlines' business jairtraman.2018.06.004.
models afford these airlines the ability to weather uncertain times without Colombo, V., 2013. Economic policy uncertainty in the US: does it matter for the euro area?
Econ. Lett. 121 (1), 39–42. https://doi.org/10.1016/j.econlet.2013.06.024.
the large employment reductions seen by the major carriers. One explana-
Davies, R.E.G., 2016. Airlines of the Jet Age: A History. Smithsonian Institution.
tion is that these airlines are built around low fares driven by low costs of Dixit, A., 1997. Investment and employment dynamics in the short run and the long run. Oxf.
operation. The already lean nature of these airlines may make them more Econ. Pap. 49 (1), 1–20. https://doi.org/10.1093/oxfordjournals.oep.a028592.
Eichenbaum, M., Rebelo, S., Trabandt, M., 2020. The Macroeconomics of Epidemics
able at allocating personnel in a cost-effective manner. Regional airlines re-
(No. w26882). National Bureau of Economic Research https://doi.org/10.3386/
ceive a boost during these periods as major airlines contract out routes to w26882.
these airlines. The lower costs related to personnel and aircraft make Elliott, G., 1998. On the robustness of cointegration methods when regressors almost have
them ripe candidates for substitution of major airline capacity. These unit roots. Econometrica 66 (1), 149–158. https://doi.org/10.2307/2998544.
Franke, M., 2004. Competition between network carriers and low-cost carriers—retreat battle
changes ultimately diminish as demand recovers post uncertainty shock or breakthrough to a new level of efficiency? J. Air Transp. Manag. 10 (1), 15–21.
and major carriers resume mainline operations. Another important finding https://doi.org/10.1016/j.jairtraman.2003.10.008.
of this study is that certain categories of employees will face deeper work- Franke, M., John, F., 2011. What comes next after recession? — Airline industry scenarios and
potential end games. J. Air Transp. Manag. 17 (1), 19–26. https://doi.org/10.1016/j.
force cuts than others. The occupations related to passenger handling at jairtraman.2010.10.005.
major airlines are noted as a category that is hardest hit and could benefit Gilbertson, D., 2020. US airlines must continue flying if they accept coronavirus relief. USA
most from a safety net such as career transition assistance. The lower skilled Today, 31 March. Available at:. https://www.usatoday.com/story/travel/airline-news/
2020/03/31/coronavirus-stimulus-package-us-airlines-must-continue-flying/
employees at airlines appear to receive the majority of the impact from 5093641002/, Accessed date: 3 April 2020.
workforce reductions. This finding aligns with prior economic research re- Goodman, C.J., 2008. Takeoff and descend of airline employment. Monthly Lab. Rev. 131, 3.
lated to workforce reductions due to uncertainty. The implications to Greene, W., 2011. Econometric Analysis. Seventh edition. Pearson.
Hiltzik, M., 2019. Column: In anti-union campaign, Delta becomes latest firm offering dumb
policymakers from these findings are that resources should be devoted to financial advice. Los Angeles Times, 13 May. Available at:. https://www.latimes.com/
the less skilled employees working at airlines. As many of these employees business/hiltzik/la-fi-hiltzik-delta-union-20190513-story.html, Accessed date: 30 March
are not represented by labor unions, the at-will nature of their employment 2020.
Jean, D.A., Lohmann, G., 2016. Revisiting the airline business model spectrum: the influence
may make them more susceptible to workforce reductions (Slotnick, 2020).
of post global financial crisis and airline mergers in the US (2011−2013). Res. Transp.
It would appear unlikely to organize all airline workers as airlines have in- Bus. Manag. 21, 76–83. https://doi.org/10.1016/j.rtbm.2016.06.002.
creasingly lowered the percentage of organized workers in their employee Johansen, S., 1988. Statistical analysis of cointegration vectors. J. Econ. Dyn. Control. 12 (2–
counts via anti-union campaigns and outsourcing (Hiltzik, 2019). 3), 231–254. https://doi.org/10.1016/0165-1889(88)90041-3.
Johansen, S., Juselius, K., 1990. Maximum likelihood estimation and inference on
As COVID-19 continues to reduce air travel and airline capacity, future cointegration — with applications to the demand for money. Oxf. Bull. Econ. Stat. 52
analyses could provide a more complete picture of the impacts by account- (2), 169–210. https://doi.org/10.1111/j.1468-0084.1990.mp52002003.x.
ing for the indirect effects to employment. Although beyond the focus of Josephs, L., 2020. American Airlines cutting international summer schedule by 60% as coro-
navirus drives down demand. CNBC News, 2 April. Available at:. https://www.cnbc.
this study, the implications to employment that is tied to the transport of com/2020/04/02/coronavirus-update-american-airlines-cuts-summer-international-
passengers and cargo via air travel are vitally important. Without the re- flights-by-60percent-as-demand-suffers.html, Accessed date: 3 April 2020.
lated tourism, business travel, flow of goods, etc., these indirect employees Leduc, S., Liu, Z., 2016. Uncertainty shocks are aggregate demand shocks. J. Monet. Econ. 82,
20–35. https://doi.org/10.1016/j.jmoneco.2016.07.002.
will likely experience workforce reductions as well. An accurate account of Ludvigson, S.C., Ma, S., Ng, S., 2015. Uncertainty and Business cycles: Exogenous Impulse or
both the direct and indirect effects would also highlight the truly vital na- Endogenous Response? (No. w21803). National Bureau of Economic Research.
ture of air transportation and the need to understand the dynamics of the Mahutga, M.C., Ma, X., Smith, D.A., Timberlake, M., 2010. Economic globalisation and the
structure of the world city system: the case of airline passenger data. Urban Stud., 47
industry to uncertainty shocks such as this one. (9), 1925-1947. Doi:https://doi.org/10.1177/0042098010372684.
McKibbin, W., Fernando, R., 2020. The global macroeconomic impacts of COVID-19: seven
References scenarios. The Brookings Institution Report. Available at:. https://www.brookings.edu/
research/the-global-macroeconomic-impacts-of-covid-19-seven-scenarios/.
Neal, Z.P., Kassens-Noor, E., 2011. The business passenger niche: comparing legacy carriers
Alessandri, P., Mumtaz, H., 2019. Financial regimes and uncertainty shocks. J. Monet. Econ. and southwest during a national recession. J. Air Transp. Manag. 17 (4), 231–232.
101, 31–46. https://doi.org/10.1016/j.jmoneco.2018.05.001. https://doi.org/10.1016/j.jairtraman.2010.09.001.
Altug, S., Tan, B., Gencer, G., 2012. Cyclical dynamics of industrial production and employ- von Nordenflycht, A., Gittell, J.H., 2013. Context, process, and outcomes of collective
ment: Markov chain-based estimates and tests. J. Econ. Dyn. Control. 36 (10), bargaining in the US airline industry. In: Stanger, Howard R., Frost, Ann C., Clark, Paul
1534–1550. https://doi.org/10.1016/j.jedc.2012.03.016. F. (Eds.), Collective Bargaining Under Duress: Case Studies of Major US Industries,
Atallah, S., Hotle, S.L., Mumbower, S., 2018. The evolution of low-cost carrier operational pp. 9–43.
strategies pre-and post-recession. J. Air Transp. Manag. 73, 87–94. https://doi.org/ Pearce, B., 2012. The state of air transport markets and the airline industry after the great reces-
10.1016/j.jairtraman.2018.08.011. sion. J. Air Transp. Manag. 21, 3–9. https://doi.org/10.1016/j.jairtraman.2011.12.011.
Atkeson, A., 2020. What Will Be the Economic Impact of COVID-19 in the US? Rough Esti- Phillips, P.C., 1987. Time series regression with a unit root. Econometrica, 277–301 https://
mates of Disease Scenarios (No. w26867). National Bureau of Economic Research doi.org/10.2307/1913237.
https://doi.org/10.3386/w26867. Phillips, P.C., Perron, P., 1988. Testing for a unit root in time series regression. Biometrika 75
Barro, R., Ursua, J., Weng, J., 2020. The Coronavirus and the Great Influenza Pandemic: Les- (2), 335–346. https://doi.org/10.1093/biomet/75.2.335.
sons from the “Spanish Flu” for the Coronavirus’s Potential Effects on Mortality and Eco- Sibdari, S., Mohammadian, I., Pyke, D.F., 2018. On the impact of jet fuel cost on airlines' ca-
nomic Activity (No. w26866). National Bureau of Economic Research https://doi.org/ pacity choice: evidence from the US domestic markets. Transp. Res. E, 111, 1–17. doi:
10.3386/w26866. https://doi.org/10.1016/j.tre.2017.12.009.
Basu, S., Bundick, B., 2017. Uncertainty shocks in a model of effective demand. Econometrica Sims, C.A., 1980. Macroeconomics and reality. Econometrica 48 (1), 1–48. https://doi.org/
85 (3), 937–958. https://doi.org/10.3982/ECTA13960. 10.2307/1912017.
Bhadra, D., 2009. Race to the bottom or swimming upstream: performance analysis of US air- Slotnick, D., 2020. US airport and airline contractors are being laid off nationwide and risk
lines. J. Air Transp. Manag. 15 (5), 227–235. https://doi.org/10.1016/j. being forgotten in a bailout plan meant to protect workers. Bus. Insid. 24 March. Avail-
jairtraman.2008.09.014. able at:. https://www.businessinsider.com/airline-contract-workers-laid-off-bailout-
Bloom, N., 2009. The impact of uncertainty shocks. Econometrica 77 (3), 623–685. https:// covid19-2020-3, Accessed date: 3 April 2020.
doi.org/10.3982/ECTA6248. Tan, K.M., 2016. Incumbent response to entry by low-cost carriers in the US airline industry.
Bloom, N., 2014. Fluctuations in uncertainty. J. Econ. Perspect. 28 (2), 153–176. https://doi. South. Econ. J. 82 (3), 874–892. https://doi.org/10.1002/soej.12117.
org/10.1257/jep.28.2.153. Tan, K.M., 2018. Outsourcing and price competition: an empirical analysis of the partnerships
Bloom, N., Bond, S., Van Reenen, J., 2007. Uncertainty and investment dynamics. Rev. Econ. between legacy carriers and regional airlines. Rev. Ind. Organ. 53 (2), 275–294. https://
Stud. 74 (2), 391–415. https://doi.org/10.1111/j.1467-937X.2007.00426.x. doi.org/10.1007/s11151-017-9610-z.
Caggiano, G., Castelnuovo, E., Groshenny, N., 2014. Uncertainty shocks and unemployment Wolfers, J., 2020. The unemployment rate is probably around 13 percent. New York Times 3
dynamics in US recessions. J. Monet. Econ. 67, 78–92. https://doi.org/10.1016/j. April. Available at:. https://www.nytimes.com/2020/04/03/upshot/coronavirus-job-
jmoneco.2014.07.006. less-rate-great-depression.html, Accessed date: 3 April 2020.