Schreyer, Dominik; Torgler, Benno
Working Paper
Football spectator no-show behavior in Switzerland:
Empirical evidence from season ticket holder behavior
CREMA Working Paper, No. 2021-06
Provided in Cooperation with:
CREMA - Center for Research in Economics, Management and the Arts, Zürich
Suggested Citation: Schreyer, Dominik; Torgler, Benno (2021) : Football spectator no-show behavior
in Switzerland: Empirical evidence from season ticket holder behavior, CREMA Working Paper, No.
2021-06, Center for Research in Economics, Management and the Arts (CREMA), Zürich
This Version is available at:
https://hdl.handle.net/10419/234621
Standard-Nutzungsbedingungen:
Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen
Zwecken und zum Privatgebrauch gespeichert und kopiert werden.
Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle
Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich
machen, vertreiben oder anderweitig nutzen.
Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen
(insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten,
gelten abweichend von diesen Nutzungsbedingungen die in der dort
genannten Lizenz gewährten Nutzungsrechte.
Terms of use:
Documents in EconStor may be saved and copied for your personal
and scholarly purposes.
You are not to copy documents for public or commercial purposes, to
exhibit the documents publicly, to make them publicly available on the
internet, or to distribute or otherwise use the documents in public.
If the documents have been made available under an Open Content
Licence (especially Creative Commons Licences), you may exercise
further usage rights as specified in the indicated licence.
Center for Research in Economics, Management and the Arts
Raumplanung:
Rückzonungen sollen
Einzonungen ermöglichen
René L. Frey
Artikel erschienen in Basellandschaftliche Zeitung, 28. November 2012, S. 30,
aufgrund des Referats «Mehrwertabschöpfung: Eine politisch-ökonomische Analyse»,
gehalten am 1. November 2012 in Zürich im Rahmen des «Forums Raumwissenschaften»,
Universität Zürich und CUREM
Beiträge zur aktuellen Wirtschaftspolitik No. 2012-04
CREMA Gellertstrasse 18 CH-4052 Basel www.crema-research.ch
Football spectator no-show behavior in
Switzerland: Empirical evidence from season
ticket holder behavior
Working Paper No. 2021-06
CREMA Südstrasse 11 CH - 8008 Zürich www.crema-research.ch
1
Football spectator no-show behavior in Switzerland:
Empirical evidence from season ticket holder behavior
Dominik Schreyer & Benno Torgler*
This version: February 10, 2021
Current version: February 10, 2021
Abstract: For football executives, understanding the determinants of spectator no-show
behavior better is of utmost importance. Recent research efforts, however, have
primarily focused on exploring the potential effects of determinants that the club
management can hardly influence (e.g., potential scheduling effects, the visiting
team's quality, and the weather). In contrast, our understanding of factors relating
to both accommodation (e.g., the ticket price), socio-demographics (e.g., age),
and also emerging no-show habits, in predicting no-show behavior is still
limited. Here, departing from more traditional survey-approaches, we address
this shortcoming by exploring disaggregated behavioral season ticket holder data
provided by an established Swiss Super League club. Analyzing a rich data set
containing roughly 2.09 million attendance decisions made by ticket holders in
Switzerland between 2013 and 2016, we observe that both a season ticket
holder’s accommodation and his (or her) socio-demographic information can
help predict subsequent no-show behavior. In particular, we notice an important
role of a season ticket holders’ age, his (or her) domicile, and emerging no-show
habits, as well as the season ticket price. Although our results suggest that the
management of clubs with a strong demand for tickets might be well-advised to
begin experimenting with strategies to exploit emerging no-show habits among
their season ticket holders, most executives, i.e., those operating at clubs that
sell-out their stadium only occasionally, might want to prioritize efforts to
increase the inherent ticket value (e.g., by reducing the ticket supply).
Keywords: attendance, decision-making, demand, football/soccer, no-shows, season
tickets,
spectator sports, stadiums
JEL Codes D12, L83, R22, Z20
Running head: Football spectator no-show behavior in Switzerland
Schreyer: WHU – Otto Beisheim School of Management, Erkrather Str. 224a, 40233, Düsseldorf, Germany (e-
mail: [email protected]); Torgler: Queensland University of Technology (QUT), QUT Business School,
School of Economics and Finance, Gardens Point 2 George St, Brisbane QLD 4000, and CREMA – Center for
Research in Economics, Management and the Arts, Switzerland (e-mail: [email protected]).
2
Football spectator no-show behavior in Switzerland:
Empirical evidence from season ticket holder behavior
Introduction
Today, for professional football executives, understanding the determinants of spectator no-
show behavior better is of utmost importance. Although one might argue that a ticket holder’s
decision to forgo physical attendance on matchday, thus being a no-show, is mostly irrelevant
to a football club once the ticketing department has sold the ticket to him (or her), and generated
the necessary income, this subsequent decision, however, poses several severe challenges for
the club management. First, as such spectator no-show behavior is more likely among a club's
many season ticket holders (STHs; e.g., Schreyer, 2019), those football executives running a
club facing a strong ticket demand lose substantial match day income if those season tickets,
usually offered at a significant discount, remain unused but could have been otherwise sold to
the public on the short term for the regular price or, perhaps, even more. Similarly, the very
same executives, second, increasingly need to manage complaints from impatient future STHs
on the waiting list that often remain empty-handed for long periods, despite frequently
observing empty seats. Third, as spectators must be considered an integral part of the product
offered to third parties (e.g., Correia & Esteves, 2007; Kuenzel & Yassim, 2007; Morrow,
1999), a large number of no-shows is, further, in diametric opposition to the interest of the club
management’s most valuable external stakeholders (e.g., broadcasters, corporate sponsors, and
also those customers in the hospitality section), all of whom benefit from the atmosphere
created in a packed stadium (e.g., McDonald, 2010). In other stakeholder groups, most notably
among TV audiences, sold-out yet unoccupied stadiums, fourth, might decrease future stadium
visit intentions (e.g., Oh et al., 2017), thus endangering the success of the management’s initial
ticketing strategy. Somewhat related, among STHs – the ticketing department’s most important
stakeholder group – frequent no-show behavior might, fifth, ultimately result in future season
3
ticket churn (e.g., McDonald et al., 2014). Sixth, in the stands, every no-show results in reduced
income generated from selling food and beverages, team merchandise, and often also parking
tickets. Despite this shortfall in matchday revenue, seventh, club management often has to plan
according to information from ticketing sales and might, thus, misspend not on only personnel
but also goods. Further, as football clubs are increasingly interested in diversifying their income
sources (cf., Schmidt & Holzmayer, 2018), some clubs, eight, might also benefit from a
decrease in auxiliary revenues generated through, for example, hotel stays, museum visits, and
stadium tour bookings if spectator no-show behavior increases. On the field, ninth, thousands
of empty seats might diminish an otherwise often significant home advantage (e.g., Bryson et
al., 2021; Krumer & Lechner, 2018; Reade et al., 2020a).
More recently, two mostly independent, complementary streams of empirical research
on the phenomenon of spectator no-show behavior have emerged in the literature. While their
authors have employed different methodological approaches to various sporting markets, most
of these studies, however, offer only a few answers to the critical question of how to manage
spectator no-show behavior.
Those authors in the first stream began early – perhaps often unintentionally – to
document the existence of no-shows in different sporting environments, but were mostly
interested in exploring related concepts, such as season ticket churn (e.g., McDonald, 2010),
relationship quality (e.g., Lee et al., 2019), and satisfaction (e.g., McDonald et al., 2017).
Accordingly, although most of the research generated in this stream of the literature relies on
cross-sectional survey data collected on the individual level, and, thus, might help us grasp the
magnitude and, perhaps, also infer the cultural robustness of the increasingly important
phenomenon better, only a few studies (e.g., McDonald et al., 2017; Sampaio et al., 2015;
Solberg & Mehus, 2014) add to our understanding of its potential antecedents. Further, as Karg
and McDonald (2011), among others, observe, it ultimately remains questionable as to whether
ticket holders, when surveyed, give an accurate indication of their – socially undesirable – no-
4
show behavior. As such, the use of survey-data to understand no-show behavior better might
be considered inappropriate.
1
In contrast, the authors in the second stream only recently began explicitly modeling the
determinants of spectator no-show behavior by employing behavioral rather than survey data.
2
Although not without its merits, most of this previous research (e.g., Schreyer & Däuper, 2018)
suffers from not only a focus on exploring the effect of determinants that the club management
can hardly influence (e.g., potential scheduling effects, the visiting team's quality, and the
weather) but also from the use of aggregated data that, in turn, has yet prevented the field from
analyzing relevant parameters such as the ticket price in the necessary detail. Those few authors
that do exploit individual data, however, only present results for a rather short period of
observation (e.g., Schreyer et al., 2016), and, as becomes evident during their analysis, fail to
exclude those debtors holding multiple season tickets, which, in turn, might affect the
robustness of the reported results.
In this paper – perhaps best understood as documenting an enhanced replication study
– we add to the emerging empirical literature on spectator no-show behavior in two important
ways. First, drawing on the core elements from both literature streams discussed above, i.e., the
use of data collected on the individual level and a timely move towards exploring behavioral
data rather than mere behavioral intentions, we are first to examine whether earlier results on
both the existence of no-show behavior and its antecedents are potentially generalizable beyond
the German market. More precisely, by exploring a unique and original data set containing
roughly 2.09 million distinct attendance decisions from a Swiss professional football club, we
loosely replicate the early work of Schreyer et al. (2016) – to the best of our knowledge, the
only similar study in the field – in an alternative, but more importantly different, environment;
1
On a more general note, as Katz et al. (2019) summarize, such behavioral intentions “are often a poor predictor
of actual behavior” (5)
2
Intriguingly, this stream of the literature originated in the late 70s (cf., Siegfried & Hinshaw, 1977; 1979) but,
then abated, perhaps because the access to such behavioral data is typically (still) scarce.
5
i.e., the Swiss football market, and over a significantly extended period of observation.
Although there is no need for a wide range of similar case studies from different countries, a
case study originating from Switzerland is particularly interesting because the rather small
Swiss football market is not only surprisingly under-researched but also differs from the
German market in several ways,
3
not least in terms of the championship format (e.g., Pawlowski
and Nalbantis, 2015) and the significance of matchday income.
4
In addition, exploring
individual no-show data from STHs over multiple seasons for the first time also allows us to
control for any time‐varying trends. Second, as we are primarily interested in those
determinants over which football club executives might ultimately have at least some control,
we provide a significantly more nuanced evaluation of the previously discussed role of STHs’
age, the paid season ticket price, and, even more important, emerging no-show habits in
predicting spectator no-show behavior, as well as the potential downside of offering previously
unexplored family areas and also a potential substitution effect due to STHs' birthday on
matchday. Although most (not all) of these factors have been explored earlier in Schreyer et al.
(2016), these previous attempts have predominantly failed to account for the potential non-
linearity in the relationship between, for instance, the season ticket price and subsequent
spectator decision-making. Similarly, the role of habit in shaping no-show behavior has – so far
– only been considered to a limited extent.
Intriguingly, our empirical results suggest that previous results observed in the German
Bundesliga seem to be only partially robust across borders. That is, although we observe a
crucial role of, for example, both a STHs’ geographical distance to the stadium and his or her
3
That is, unlike the 18 German Bundesliga teams, the 10 Swiss Super League (RSL) clubs play each other four
times each season in a quadruple round-robin tournament, which might affect STH decision-making. In addition,
RSL stadiums, some of which were significantly expanded before the 2008 UEFA European Championship, are
typically less crowded, which might affect the perceived atmosphere considerably, while matches staged in these
stadiums usually feature fewer stars as approximated by transfer market values.
4
For major European football clubs, today, matchday income still corresponds to about 15 percent of total turnover
(Deloitte, 2020), despite an increasingly important role of media income. In some leagues, most notably, Scotland
(43 percent), Switzerland (31), the Netherlands (29), Ireland (28), Sweden (24), Israel (24), and Belgium (22)
however, gate revenues often remain an even more significant part of the revenue mix (UEFA, 2020).
6
no-show habit when predicting spectator no-show behavior that is in line with previous
research, we note a significantly more nuanced role of the season ticket price. More specifically,
we discover a robust, non-linear relationship between the price and STH no-show behavior,
indicating that a club’s ticket pricing strategy might have a lasting effect even beyond the initial
sales process.
What do we know about spectator no-show behavior?
Although there already exists a rich and continuously growing body of literature on football
stadium attendance demand (e.g., Buraimo et al., 2018; Pawlowski & Anders, 2012; Reade et
al., 2020b; Watanabe et al., 2019; Valenti et al., 2020), as yet, only a few authors have made an
attempt to model admission behavior in general, and spectator no-show behavior in particular.
Putting the dichotomy between sports management and sports economic research aside, these
few behavioral studies can be structured along multiple lines, including, for example, the market
(e.g., Brazil: Sampaio et al., 2015; Scotland: Allan & Roy, 2008; United States: Putsis & Sen,
2000) or the sport under investigation (e.g., American Football: Zuber & Gandar, 1988), and
the chosen methodological approach – in particular, the distinction between whether the data
captures aggregated (e.g., Schreyer, 2019) or disaggregated (e.g., Schreyer et al., 2016), i.e.,
individual, spectator behavior decision-making. Further, content-wise, the existing literature
has provided first insights on both the antecedents (e.g., Schreyer et al., 2019), and the potential
consequences, most notably season ticket churn (e.g., McDonald, 2010; McDonald & Stavros,
2007; McDonald et al., 2014) of spectator no-show behavior.
To gain a better understanding of the literature on the potential determinants of football
spectator no-show behavior, in Table 1, we provide an extended overview of the determinants
of such no-show behavior in the German Bundesliga, i.e., the professional football league that
has recently attracted the greatest research interest. As can be clearly seen from that table, most
of this previous research has already established a robust relationship between spectator no-
7
show behavior and factors relating to either match quality aspects, primarily those shaped by
the visiting team (e.g., the market value), and the opportunity costs arising from attending live
matches (e.g., from extreme temperatures, mid-week fixtures, and precipitation).
- - - Insert Table 1 about here - - -
However, this does not mean that our current understanding of those factors shaping
football spectator no-show behavior is complete. In fact, as Schreyer et al. (2019) observe when
exploiting panel data on aggregated no-show behavior among 25 Bundesliga and Bundesliga 2
clubs between August 2014 and May 2017, those factors shaping Bundesliga no-show behavior
might not necessarily also affect spectator decision-making in Bundesliga 2. That is, while the
authors confirm earlier findings by Schreyer and Däuper (2018) regarding the role of both
midweek matches and the temperature in shaping Bundesliga no-show behavior, both factors
had no effect in Bundesliga 2. Further, as most authors still employ aggregated data on
behavioral decision-making generated from multiple external stakeholders, including, for
example, corporate sponsors in the hospitality section, whose behavior is likely to differ from
that of the typical fan in the stands, our understanding of socio-demographic factors in shaping
spectator no-show behavior such as age, gender, and habit, as well as the ticket price and
emerging no-show habits is still rather limited.
To the best of our knowledge, so far, only two studies have modeled football spectator
no-show behavior by analyzing disaggregated STH data. First, Schreyer et al. (2016), primarily
interested in the role of match outcome uncertainty in shaping STH decision-making, explore
a panel data set containing 236,164 individual admission decisions from 13,892 STHs during
the Bundesliga season 2012-13. Interestingly, the authors provide some first empirical evidence
on the potential differences in spectator no-show behavior based on factors relating to
individual socio-demographics, and also stadium accommodation. More precisely, Schreyer et
al. (2016) observe, for example, that no-show behavior increases if a STHs’ age, his/her pitch
and travel distance, and the number of season tickets in possession increases, but decreases if
8
the ticket price increases. Also, no-show behavior was more likely to occur if the STH was
accommodated in the standing terraces, had already missed the last home match, and had
already resigned. In contrast, the authors did not observe substantial gender differences in
spectator no-show behavior – a finding that is largely consistent with related survey results
(e.g., McDonald et al., 2017; Schreyer, 2019; Solberg & Mehus, 2014). Second, in a short, and
related study employing the same data set, Schreyer et al. (2018) primarily focus on the role of
geographical distance in shaping STH behavioral loyalty, adding some necessary nuance to our
current understanding of the travel distance. However, the authors of both studies not only
present results generated from only one season, i.e., 17 consecutive Bundesliga home matches,
but, as becomes evident during their analysis, also fail to exclude those STHs holding multiple
season tickets – a procedure that might ultimately affect the robustness of the reported effect
sizes. As such, our current understanding on both the role of socio-demographic antecedents
and factors relating to a STHs’ accommodation in shaping spectator no-show behavior is still
limited at best.
Data set, explanatory variables and empirical approach
To address previous shortcomings in the literature, we base our analysis on an original dataset
containing roughly 2.09 million ticket holder attendance decisions made between February
2013 and December 2016. This information, generously provided by RSL club FC Basel 1893
(FCB), was recorded with the help of the club’s stadium access system and shared at the end of
season 2017-18. During our period of observation, the club hosted a total of 72 RSL matches,
all of which we considered in our subsequent analysis.
5
Founded in 1893, FCB is certainly the most successful club in modern Swiss football.
In fact, the club secured 12 out of their currently 20 domestic championships since the season
5
We chose this period of observation for two reasons: First, in Basel, ticketing executives sell season tickets at the
begin of each year rather than at the begin of each season; Second, in January 2017, the management altered STH
allocation numbers.
9
2001-02 and, during our period of investigation, ranked first at the end of each season.
Therefore, it is perhaps not surprising that based on their estimated winning probability, FCB
was the bookmakers' (often heavy) favorite before each of the 72 RSL home matches in our
data set (M = 0.65, SD = 0.08; Min = 0.45; Max = 0.81). Having reached the lucrative group
stage of the UEFA Champions League (UCL) in the three seasons 2013-14, 2014-15, and 2016-
17, and of the UEFA Europa League (UEL) in the seasons 2012-13 and 2015-16, FCB has
successfully competed in European football over the entire sample period.
6
Since 2001, the club has played their home matches at St. Jakob-Park, also known as
Joggeli, which with 38,512/37,500 seats (domestic/international matches), is currently the
largest football stadium in Switzerland. However, as indicated in the club’s official annual
reports (e.g., FCB, 2017), stadium attendance demand rarely matched ticket supply during our
four-year-long period of observation. More precisely, the club distributed, on average, between
27,595 and 29,706 tickets per domestic league match, most of them to their roughly 24,000
STHs.
7
As is common practice in European professional football, these season tickets are sold
at a reasonable discount. For instance, in the recent year 2019, STHs attending all RSL matches
received a discount between 39 and 46 percent relative to purchasing the 18 individual match
tickets (c.f., FCB, 2020), depending on STH accommodation.
- - - Insert Figure 1 about here - - -
In Figure 1, we first present information on the club’s no-show rate (NSR). On average,
this rate – the relative share of distributed, though subsequently unused tickets – was effectively
about 26.77 percent during our period of observation, and has slightly increased over time.
More precisely, we observe that the NSR in RSL home matches is about 25.81, 23.41, 26.61,
and 31.27 percent in the four consecutive years 2013, 2014, 2015, and 2016, respectively.
6
More precisely, in 2014-15, the club reached the UCL round of Last 16 and was UEL a semi-finalist and a
quarter-finalist in the seasons 2012-13 and 2013-14, respectively.
7
According to information in the club’s official annual report (FCB, 2017), FCB distributed between 23,671 and
24,265 season tickets per year. Intriguingly, between the stadium opening in 2001 and the 2019, the club always
distributed more than 20,000 season tickets.
10
Somewhat similarly, in the German Bundesliga, the most-attended football league in the world,
the NSR also has recently significantly increased from roughly 9.25 percent in season 2014-15
to about 11.96 percent in season 2017-18 (cf., Schreyer, 2019). In this and other European
markets, anecdotal evidence has it that many football clubs tend to observe two-digit NSRs, in
particular among their STHs (e.g., FC Business, 2018). For example, Schreyer et al. (2016) –
in an early attempt to analyze spectator no-show behavior among 13,892 STHs of a German
club – observe a NSR of approximately 17 percent. In Brazil, Sampaio et al. (2015) document
a similar NSR among Porto Alegre FC STHs during the three seasons 2008 to 2010.
Intriguingly, as can be seen from Figure 1, we observe a significantly higher NSR among
STHs than among matchday ticket holders. As STHs are, therefore, responsible for the vast
majority of all no-show occurrences in our data set, we focus our subsequent analysis on better
understanding spectator no-show behavior by exploring detailed (panel) data from this
particular group.
- - - Insert Table 2 about here - - -
In Table 2, we present descriptive information on our explanatory variables, including
the controls capturing either quality aspects or factors relating to the opportunity costs arising
from attending live matches (cf., once more, Table 1). As we are primarily interested in those
factors that link to potential management implications, we focus on either a STHs’ socio-
demographics or those explanatory variables capturing STH accommodation. Nonetheless, to
provide a more complete picture of the determinants of football spectator no-show behavior in
Switzerland, in the Appendix, we also present extended specifications, i.e., including control
variable effect sizes (cf., Table A1).
The first set of explanatories includes variables capturing a STHs’ accommodation.
Whereas most previous research has centered on factors that relate to either match quality
aspects or the opportunity costs arising from attending a home match in the stadium (e.g.,
Schreyer & Däuper, 2018), our understanding of the role of these factors in shaping STH no-
11
show behavior is still underdeveloped. In fact, as sketched out above, there exist, to the best of
our knowledge, only two studies that have started to explore the effect of such factors using
STH data generated in German football (e.g., Schreyer et al., 2018). Here, we modify their
empirical approach by analyzing a total of five factors, one of which has previously been
neglected. First, we add the season ticket price, including its squared term, on a per match basis.
Although Schreyer et al. (2016; 2018) observe that STH no-show behavior decreases as the
season ticket price increases, a finding that has subsequently been replicated using survey data
(Schreyer, 2019), it seems reasonable that the respective relationship is non-linear with an
inverted u-shape, due to diminishing price sensitivity among those STHs with a rather high
income (e.g., Solberg & Mehus, 2014). Second, we add the impact of being located in the
standing terraces. In line with previous research (e.g., Schreyer et al., 2016), we expect a higher
NSR among STHs in these sections of the stadium, primarily because those STH in the stands
might face rather high opportunity costs due to, for example, occasional work commitments on
the weekend. Third, we add a somewhat related dummy that takes the value of 1 if a STH is
located in the family area of the stadium. Previously unexplored, we expect a higher NSR
among these STHs because such family visits are typically associated with an increase in
coordination costs. As Schreyer et al. (2016) demonstrate, such costs are correlated with an
increase in STH no-show behavior. Further, to control for differences in atmosphere between
stands and stadium sectors, we also include stand and sector fixed effects, respectively.
The second set of explanatory variables includes a number of variables capturing STH
characteristics. First, we are interested in revisiting the role of STH age in no-show behavior.
Interestingly, while earlier research (e.g., Schreyer et al., 2018) has established a robust linear
relationship between AGE and no-show behavior, indicating that the NSR is higher among
younger STH, survey data suggests a non-linear, inverted U-shaped relationship between age
and spectator no-show appearances (e.g., Solberg & Mehus, 2014). Second, and somewhat
related, we also test whether a STH’s birthday on matchday increases no-show behavior. Here,
12
perhaps best understood as (further) increasing the opportunity costs arising from attending a
match in the stadium, STHs may omit attendance on their birthday to instead celebrate with
their family and friends. Third, primarily to explore the robustness of previous results, we add
a STH’s gender, a factor that has been previously reported as insignificant (e.g., Schreyer et al.,
2016). Fourth, we add DISTANCE, a proxy that captures the beeline between St. Jakob-
Park/Basel and a STHs’ place of residence in kilometers, including its squared term. As
Schreyer et al. (2018) observe, the NSR seems to be particularly low among those STHs living
in close proximity to the stadium. Somewhat similarly, Solberg and Mehus (2014), analyzing
survey data from Norwegian STHs, observe that travel time seems to shape the decision
whether to attend a home match. Finally, fifth, complementing an approach from Schreyer et
al. (2016), we successively test several dummy variables with differing cut-off points capturing
a STHs’ emerging no-show habit. Here, the corresponding dummy variable takes the value of
1 if a STH has previously omitted the last home match or home matches, respectively. In
contrast, most likely due to the rather short observation period of only 17 home matches,
previous research has refrained from exploring emerging STH no-show habits in more detail.
Although STHs are often characterized as being behaviorally loyal (e.g., Benz et al., 2009),
missing a home match or two might increase the probability of not attending a subsequent
match.
Unlike previous research (e.g., Schreyer et al., 2016), we explore STH no-show
behavior using a variety of different specifications, primarily to increase the robustness of our
empirical results. More specifically, as can be seen from Table 3, for three different models, we
estimate both a random-effects probit model and a pooled probit model in which standard errors
are estimated with a cluster-robust covariance-estimator.
8
We use matchday as the cluster
variable to measure the effect of STH no-show behavior using micro unit STH data with
8
All reported effects are robust to employing additional logit specifications, which are available from the authors
upon request.
13
aggregate match level information. Clustering allows to take into account the multi-level
structure of data. Not clustering can lead to standard errors that are biased downward and
subject to the possibility that the random disturbances in the regression are correlated within
groupings (Moulton, 1990). The bias of the standard errors can result in spurious findings of
statistical significance for those aggregate variables. In addition, for the first two panel models,
we also estimate additional results excluding the higher-level aggregate level variables but
adding matchday dummies to control for unobserved match-level characteristics.
Here, we measure a STH’s decision whether to attend a particular home match using a
binary scale. Accordingly, NO-SHOW, our dependent variable, takes the value of 1 if STH i
omits a particular home match t and a value of zero otherwise. Further, mirroring the approach
chosen in Schreyer et al. (2018), we also present two alternative specifications applying Poisson
and fractional probit regression on two alternative dependent variables; i.e., a STHs’ absolute
and relative no-show behavior in the complete period of observation.
Breaking the habit? Results and arising management implications
To ensure the robustness of our empirical results, we initially report the estimates from a total
of three different models. First, in Model (01), we only include those factors that relate to either
a STH’s accommodation and his or her socio-demographic information, thereby excluding
higher-level secondary data. Second, in Model (02), we add a dummy for no-show habit that
take the value of 1 if a STH has for attendance for the last two home matches. By definition,
there is no information on a STHs’ no-show habit for the first two home matches in our data
set; thus, we employ a slightly smaller panel data set containing information on about 611,380
individual attendance decisions made by 8,734 permanent STHs on 70 consecutive matchdays.
Finally, in Model (03), we add those controls relating to either the match quality or the
opportunity costs that may arise from attending a football match on site. Note that, for these
first three models, we present estimates generated through both panel probit regression and
14
pooled probit regression with standard errors clustered by subsequent matchdays. Further, we
also present additional estimations from a reduced specification (3c),
9
were we only include
those explanatory variables that were, and remain, significant across the previous models.
In addition, to ensure the comparability of our results to previous STH research –
primarily that of Schreyer et al. (2018) – we present two more models that include both the total
number of absolute no-show appearances during the period of investigation, i.e., 72 consecutive
home matches (04), and its relative share (05).
10
On average, the 8,734 STHs skipped about 21
home matches over the course of four years. Thus, STHs missed, on average, between five to
six home matches per year. Interestingly, only about one percent of those STHs were
behaviorally loyal and attended all 72 matches; excluding these behaviorally loyal STHs from
our sample does not affect the reported results.
- - - Insert Table 3 about here - - -
In Table 3 and Figure 2, we report our estimation results and present an overview of the
relative effect sizes, respectively. Although we can confirm some of the results obtained earlier
in those studies exploring individual spectator decision-making in the German Bundesliga (cf.,
once more, Table 1), we observe some notable differences that are worth discussing in more
detail. This is particularly true with respect to the potential role of a STHs’ age, the paid season
ticket price, and – even more important – emerging no-show habits in predicting spectator no-
show behavior.
- - - Insert Figure 2 about here - - -
Interestingly, and in contrast to earlier findings (e.g., Schreyer et al., 2016), we first
observe a non-linear, inverted U-shaped relationship between a STHs’ age and spectator no-
9
It is noteworthy that the reported effects from specification (03c) are robust to the exclusion of those STH with
either zero (Pseudo R² = 0.0863) or perfect attendance (0.0865). Further, the reported effects are robust to
employing alternative habit dummies capturing the omission of either only one (0.0853), three (0.0818), four
(0.0766), and five (0.0699) subsequent home match(es), despite the gradually decreasing number of observations.
10
Although we employ Poisson and fractional probit regression to model the cumulated absolute number (04) and
the relative share (05) of no-show appearances, respectively, all reported effects are also robust to employing
alternative ordinary least squares (OLS) specifications. These additional estimates are available from the authors
upon request.
15
show behavior with a turning point at about 40 years. Although earlier research based on survey
data (e.g., Solberg & Mehus, 2014) has long indicated that no-show behavior might be
particularly likely among those STHs in their midlife phase – perhaps because team
identification, i.e., the degree to which a fan identifies with a particular team, varies over time
(cf., Bergmann et al., 2016) – previous behavioral research has not explored such a potential
non-linearity.
Accordingly, those football executives interested in maximizing their stadium capacity
utilization need to solve an interesting paradox: Interested in selling more season than matchday
tickets, their ticketing executives are most likely to sell season tickets to individuals from the
club’s most natural customer segment, i.e., middle-aged fans who can afford the cost of
attending matches live in the stadium. Somewhat ironically, it is these STHs, in particular, that
are more likely to skip matches, at least occasionally. In fact, as can be seen in Figure 3, here,
we observe a similar tendency, as about 56 percent of all STHs in our data set were between
the age of 40 and 60 years. Therefore, as long as decreasing the number of available season
tickets is not a viable option to reduce spectator no-show behavior, ticketing executives might
want to explore alternative customer segments; i.e., local teenagers. Interestingly, anecdotal
evidence has it that football clubs such as Manchester United have already started moving in
this direction to address a perceived decrease in stadium atmosphere due to a collectively aging
audience (cf., Irish Times, 2018).
11
Alternatively, primarily reflecting the significantly higher
NSR among STHs in the family section of the stadium, increasingly targeting children and their
grandparents, rather than their parents (perhaps the more traditional market), might be an
interesting option to increase stadium attendance on matchday. Further, as we observe an
increase in no-show behavior among birthday boys/girls, football clubs might want to explore
ways to make a birthday visit to the stadium more appealing to their STHS – for example, by
11
In our initial data set, we observe a similar aging effect – a moderate increase from about 42.48 years, on average,
on the first matchday in 2013 to about 43.83 years, on average, on the first matchday in 2016.
16
allowing those STHs to bring a friend to the often underutilized stadium, or by offering cost-
effective presents.
- - - Insert Figure 3 about here - - -
As with the role of a STH’s age, we also observe a non-linear, U-shaped relationship
between the season ticket price and no-show behavior with a turning point at about 23 CHF per
match; i.e., an equivalent of about 20 Euro.
12
This indicates that a simple strategy of increasing
the season ticket price to reduce no-show behavior as suggested by earlier research might not
necessarily work in all customer segments and across all European football clubs. More
precisely, as we observe exceptionally high NSRs among both those STHs in our sample
holding free season tickets and those with relatively expensive season tickets, increasing entry-
level prices might, in fact, decrease STH no-show behavior, whereas raising already high prices
might not. Intriguingly, this finding is supported by previous survey results, indicating that an
increase in income is likely to reduce STH appearances in Norway (Solberg & Mehus, 2014).
While our results suggest that those executives operating at a football club where the
demand for tickets does not exceed its (fixed) supply might want to reconsider their season
ticket pricing strategies, perhaps by offering significant discounts only to those STHs with
perfect (or near-perfect) attendance, they, in turn, also pave a way to monetize no-shows for
those executives operating at in-demand clubs. More precisely, curing the symptoms of
spectator no-show behavior rather than eliminating the initial cause (cf., Schreyer, 2019), these
executives might not only want to increase season ticket prices until the demand nears available
ticket supply – thus generating more initial income on matchdays – but might then also want to
overbook some of these tickets. Intriguingly, this hints at an important point: That is, depending
on a clubs’ ticketing strategy, a large number of no-shows can – but does not necessarily –
present a challenge. In contrast, it can also be an opportunity. In Germany, for example,
12
During the 4-year-long period of investigation, 1 CHF was, on average, worth approximately 0.87 Euro.
17
anecdotal evidence has it that football club Bayern Munich have already begun overbooking
parts of their stadium (cf., Smart Pricer, 2018), although the approach, for now, seems not to
include substantial season ticket price increases.
Although both a STHs’ age and the paid season ticket price are robust predictors of STH
no-show behavior, the associated effects are modest when compared to a STHs’ no-show habit.
Intriguingly, as can be seen in the three graphs depicted in the lower row of Figure 3, the
observed effect size, further, seems to increase with every additional no-show appearance. That
is, while we observe that the NSR among those STHs forgoing physical attendance once is
about 47 percent on the subsequent matchday, it increases to 60, 68, 79, 89, and 96 percent after
two, three, five, ten and twenty consecutive no-show appearances, respectively. Although this
tendency seems to point to a growing disinterest over time, it is, perhaps, better explained by
an increasing differentiation of STHs, all of which might have distinct motivations to purchase
a ticket in the first place (cf., Karg et al., 2019). In fact, survey-based research has long
established that a fair share of fans purchases their season ticket to either financially support or
to feel more involved with the club rather than to gain a reserved seat (e.g., McDonald &
Stavros, 2007). In other words, there might be no habit to break. As such, while anecdotal
evidence has it that the management of clubs with a strong demand for tickets, i.e., a ticket
demand continuously exceeding ticket supply, such as Borussia Dortmund, have recently begun
experimenting with strategies based on STH punishment to break negative STH no-show habits
(e.g., Ruhr Nachrichten, 2019), executives operating at football clubs that sell-out their stadium
only occasionally might want to prioritize their efforts in selling additional season tickets rather
than reduce subsequent no-show behavior.
13
On a more general note, however, this observation directly points to the important
strategic question of whether these clubs, per se, supply too many seats, resulting in an inferior
13
As one reviewer has rightfully argued, for these executives, the potential disadvantages arising from STH no-
show behavior must be balanced against the main advantage of selling season tickets in the first place; i.e.,
mitigating a club's financial risk through a guaranteed minimum funding early in the season.
18
stadium experience for those fans with (season) tickets. Here, a significant decrease in the
stadium capacity available to interested parties, for instance, by partly dismantling/remodeling
an oversized stadium, might help revalue the (season) ticket in the medium term by inducing a
scarcity effect.
The remaining findings are in line with previous research and our expectations. More
specifically, we observe a higher NSR among STHs located in both the standing terraces and
the family area of the stadium, and a lower NSR among those STHs that live in close proximity
to the stadium. To better understand the distance effect geographically, in Figure 4, we illustrate
this outcome by plotting the average number of no-show appearances per postcode on a map of
Switzerland. In contrast, we do not observe a significant gender effect.
- - - Insert Figure 3 about here - - -
Limitations and future research
Although our empirical results presented above shed some additional light on the determinants
of STH no-show behavior, we believe that there are still a number of open questions left to be
answered in future research. This is particularly true for questions addressing the management
of such no-show behavior. As Schreyer (2019) summarizes, previous research on reducing the
negative effects from no-show behavior in alternative environments seems to distinguish two
strategies to maximize capacity utilization. The first is to treat the immediate cause, which in
this case is to motivate STHs to either use or share their ticket(s). The second is to cure the
negative symptoms arising from this immediate cause: here, to explore alternative ways to
compensate for the resulting shortfall in additional matchday revenue.
Accordingly, analyzing the effectiveness of corresponding measures could make for an
exciting new path in sports management research; i.e., no-show management. In fact, as
discussed above, increasing anecdotal evidence has it that some football clubs have already
begun experimenting with mechanisms of rewards and punishment to reduce no-show behavior
19
among STHs. Alternative research approaches might explore the role of reminders (e.g.,
Schreyer et al., 2020), alternative payment and pricing mechanisms, and consumer apps that
allow for convenient season ticket sharing in reducing spectator no-show behavior.
In addition, we suggest future research to explore the robustness of our empirical results
not only in alternative football leagues such as the English Premier League but also across
different sports. In fact, as most previous research exploring behavioral intentions has focused
on analyzing stadium attendance demand for US sports leagues, it would be interesting to
explore whether these intentions to attend ultimately translate into subsequent behavior in
alternative environments such as in College Football. High NSRs in this context have not only
become a topic of media concern recently (Wall Street Journal, 2018) but, apparently, have
already begun motivating further research in the field (cf., Popp et al., 2019).
Finally, although we hope to add to the field’s current understanding of those factors
shaping STH decision-making, there are still numerous possible factors influencing spectator
decision-making that, unfortunately, are not part of our otherwise rich data set. For example, as
Deserpa (1994) argues, STHs might purchase their ticket(s) because of the included additional
rights. In fact, some clubs offer season tickets that also include access to matches in domestic
and/or international cup competitions. Accordingly, if STHs have a specific interest in these
sets of matches, it seems likely that such product characteristics, as well as potential variations
in the season ticket bundle, affect subsequent STH decision-making. Further, as both team
identification and variations in social status might play an essential role in explaining spectator
decision-making (cf., Schreyer, 2019), future research might be well-advised to revisit the
determinants of spectator no-show behavior by combining behavioral and survey data, thus
drawing a more complete picture of those factors that need to be addressed to minimize NSRs.
20
Conclusions
In a rapidly changing football environment (cf., Merkel et al., 2016), a better understanding of
the determinants of spectator no-show behavior is of utmost importance to football executives.
Exploring a rich data set containing roughly 2.09 million attendance decisions made by ticket
holders in Switzerland between 2013 and 2016, we observe that both a STH’s accommodation
and socio-demographic information can help predict subsequent no-show behavior. In
particular, we notice an important role of a STHs’ age, his (or her) domicile and emerging no-
show habits, as well as the season ticket price. In contrast, we observe no gender differences in
football spectator no-show behavior. While these results can be helpful in defining measures to
overcome the challenges arising from spectator no-show behavior, the ideal measure might be
context-sensitive.
21
References
Allan, G., & Roy, G. (2008). Does television crowd out spectators? New evidence from the
Scottish Premier League. Journal of Sports Economics, 9(6), 592-605.
Benz, M. A., Brandes, L., & Franck, E. (2009). Do soccer associations really spend on a good
thing? Empirical evidence on heterogeneity in the consumer response to match
uncertainty of outcome. Contemporary Economic Policy, 27(2), 216-235.
Bergmann, A., Schmidt, S. L., Schreyer, D., & Torgler, B. (2016). Age and organizational
identification: empirical findings from professional sports. Applied Economics Letters,
23(10), 718–722.
Buraimo, B., Tena, J. D., & de la Piedra, J. D. (2018). Attendance demand in a developing
football market: The case of the Peruvian first division. European Sport Management
Quarterly, 18(5), 671-686.
Bryson, A., Dolton, P., Reade, J. J., Schreyer, D., & Singleton, C. (2021). Causal effects of an
absent crowd on performances and refereeing decisions during Covid-19. Economics
Letters, 198, 1–5. DOI: 10.1016/j.econlet.2020.109664
Correia, A., & Esteves, S. (2007). An exploratory study of spectators' motivation in football.
International Journal of Sport Management and Marketing, 2(5-6), 572–590.
Deloitte (2020). Deloitte Football Money League 2020. Manchester, United Kingdom: Deloitte.
Deserpa, A. C. (1994). To err is rational: A theory of excess demand for tickets. Managerial
and Decision Economics, 15(5), 511–518.
FC Basel [FCB] (2017). Geschäftsberichte FC Basel 1893, FC Basel 1893 AG und
Stadiondienst AG [Annual report]. Retrieved from https://bit.ly/3n1YEDt
FC Basel [FCB] (2020). Geschäftsberichte FC Basel 1893, FC Basel 1893 AG und
Stadiondienst AG [Annual report]. Retrieved from bit.ly/2WVPqOj.
22
FC Business (2018). Spanish Champions FC Barcelona Sees Improved Attendances at Camp
Nou as They Approve New Incentives to Ticket Exchange Programme. Retrieved from
https://bit.ly/2R31Cbn
Frevel, N., & Schreyer, D. (2020). Behavioral responses to terrorist attacks: empirical evidence
from professional football. Applied Economics Letters, 27(3), 244–247.
Karg, A. J., & McDonald, H. (2011). Fantasy sport participation as a complement to traditional
sport consumption. Sport Management Review, 14(4), 327–346.
Karg, A., McDonald, H., & Leckie, C. (2019). Channel preferences among sport consumers:
Profiling media-dominant consumers. Journal of Sport Management, 33(4), 303–316.
Katz, M., Heere, B., & Melton, E. N. (2019). Predicting Fan Behavior Through Egocentric
Network Analysis: Examining Season-Ticket Holder Renewal. Journal of Sport
Management. DOI: 10.1123/jsm.2019-0018
Krumer, A., & Lechner, M. (2018). Midweek effect on soccer performance: Evidence from the
German Bundesliga. Economic Inquiry, 56(1), 193–207.
Kuenzel, S., & Yassim, M. (2007). The effect of joy on the behaviour of cricket spectators: The
mediating role of satisfaction. Managing Leisure, 12(1), 43-57.
Irish Times (2018, April 3). Manchester United introducing £15 tickets for young fans.
Retrieved from https://bit.ly/33sYqfo
Lee, M. A., Kunkel, T., Funk, D. C., Karg, A., & McDonald, H. (2019). Built to last:
relationship quality management for season ticket holders. European Sport
Management Quarterly. DOI: 10.1080/16184742.2019.1613438
McDonald, H. (2010). The factors influencing churn rates among season ticket holders: An
empirical analysis. Journal of Sport Management, 24(6), 676–701.
23
McDonald, H., & Stavros, C. (2007). A defection analysis of lapsed season ticket holders: A
consumer and organizational study. Sport Marketing Quarterly, 16(4), 218–229.
McDonald, H., Karg, A. J., & Leckie, C. (2014). Predicting which season ticket holders will
renew and which will not. European Sport Management Quarterly, 14(5), 503-520.
McDonald, H., Leckie, C., Karg, A., & Zubcevic-Basic, N. (2017). Female season ticket
holders: How their satisfaction is derived differently from males. European Sport
Management Quarterly, 18(2), 156-174.
Merkel, S., Schmidt, S. L., & Schreyer, D. (2016). The future of professional football: A
Delphi-based perspective of German experts on probable versus surprising scenarios.
Sport, Business and Management: An International Journal, 6(3), 295–319.
Morrow, S. (1999). The new business of football. London, United Kingdom: Macmillan.
Moulton, B. R. (1990). An illustration of a pitfall in estimating the effects of aggregate variables
on micro units. The Review of Economics and Statistics, 72(2), 334–338.
Oh, T., Sung, H., & Kwon, K. D. (2017). Effect of the stadium occupancy rate on perceived
game quality and visit intention. International Journal of Sports Marketing and
Sponsorship, 18(2), 166–179.
Pawlowski, T., & Anders, C. (2012). Stadium attendance in German professional football–The
(un) importance of uncertainty of outcome reconsidered. Applied Economics Letters,
19(16), 1553–1556.
Pawlowski, T., & Nalbantis, G. (2015). Competition format, championship uncertainty and
stadium attendance in European football–a small league perspective. Applied
Economics, 47(38), 4128–4139.
24
Popp, N., Simmons, J., Watanabe, N., & Shapiro, S. (2019). Why Didn't They Show Up? An
Examination of Factors Influencing No Show Rates at College Football Games. 2019
Sport Marketing Association (SMA) Conference (SMA XVII). November 7, 2019.
Putsis Jr, W. P., & Sen, S. K. (2000). Should NFL blackouts be banned?. Applied Economics,
32(12), 1495-1507.
Reade, J. J., Schreyer, D., & Singleton, C. (2020a). Echoes: what happens when football is
played behind closed doors? Available at SSRN 3630130. DOI: 10.2139/ssrn.3630130
Reade, J. J., Schreyer, D., & Singleton, C. (2020b). Stadium attendance demand during the
COVID-19 crisis: Early empirical evidence from Belarus. Applied Economics Letters.
DOI: 10.1080/13504851.2020.1830933.
Ruhr Nachrichten (2019, May 29). BVB kündigt 500 Dauerkarten-Inhabern [BVB cancel 500
season tickets]. Retrieved from https://bit.ly/2I0yyOw
Sampaio, C. H., Sordi, J. D., & Perin, M. G. (2015). How price bundling affects football ticket
purchases and consumption behaviour. International Journal of Sports Marketing and
Sponsorship, 16(3), 35-51.
Schmidt, S. L., & Holzmayer, F. (2018). A Framework for Diversification Decisions in
Professional Football. Routledge Handbook of Football Business and Management.
Schreyer, D. (2019). Football spectator no-show behaviour in the German Bundesliga. Applied
Economics, 51(45), 4882–4901.
Schreyer, D., & Däuper, D. (2018). Determinants of spectator no-show behaviour: first
empirical evidence from the German Bundesliga. Applied Economics Letters, 25(21),
1475–1480.
Schreyer, D., Schmidt, S. L., & Torgler, B. (2020). Using Reminders with Different Reward
Opportunities to Reduce No-Show Behavior: Empirical Evidence from a Large-Scale
25
Field Experiment in Professional Sports. Available at SSRN 3703040. DOI:
10.2139/ssrn.3703040
Schreyer, D., Schmidt, S. L., & Torgler, B. (2019). Football spectator no-show behavior.
Journal of Sports Economics, 20(4), 580–602.
Schreyer, D., Schmidt, S. L., & Torgler, B. (2018). Predicting season ticket holder loyalty using
geographical information. Applied Economics Letters, 25(4), 272–277.
Schreyer, D., Schmidt, S. L., & Torgler, B. (2016). Against all odds? Exploring the role of game
outcome uncertainty in season ticket holders’ stadium attendance demand. Journal of
Economic Psychology, 56, 192-217.
Siegfried, J. J., & Hinshaw, C. E. (1979). Effect of lifting television blackouts on professional
football no-shows. Journal of Economics and Business, 32(1), 1–13.
Siegfried, J. J., & Hinshaw, C. E. (1977). Professional Football and the Anti‐Blackout Law.
Journal of Communication, 27(3), 169–174.
Smart Pricer (2018, November 7). FC Bayern München zählt auf Smart Pricer für die No-Show
Reduzierung [FC Bayern München counts on Smart Pricer to reduce No-Show].
Retrieved from https://bit.ly/2Uyv7ne
Solberg, H. A., & Mehus, I. (2014). The challenge of attracting football fans to stadia?.
International Journal of Sport Finance, 9(1), 3–19.
Union des Associations Européennes de Football [UEFA] (2020). The European Club
Footballing Landscape: Club Licensing Benchmarking Report (Financial Year 2018).
Nyon, Switzerland: UEFA.
Valenti, M., Scelles, N., & Morrow, S. (2020). The determinants of stadium attendance in elite
women’s football: Evidence from the UEFA Women's Champions League. Sport
Management Review, 23(3), 509–520.
26
Wall Street Journal (2018, August 30). College Football’s Growing Problem: Empty Seats.
Retrieved from https://on.wsj.com/37ZeoQo.
Watanabe, N. M., Yan, G., Soebbing, B. P., & Fu, W. (2019). Air Pollution and Attendance in
the Chinese Super League: Environmental Economics and the Demand for Sport.
Journal of Sport Management, 33(4).
Zuber, R. A., & Gandar, J. M. (1988). Lifting the television blackout on no-shows at football
games. Atlantic Economic Journal, 16(2), 63–73.
27
Figures and Tables
Figure 1 Football spectator no-show behavior per ticket category
Abbreviations and notes: All figures are rounded.
28
Figure 2
Determinants of season ticket holder no-show behavior in Switzerland
Abbreviations and notes: Marginal effects of selected explanatory variables (cf., specification 03c in Table 3)
Figure 3 Distribution of STH age (N = 8,734)
Abbreviations and notes: STH age in years on matchday 72.
29
Figure 4 Average number of STH no-show appearances per postcode area
Abbreviations and notes: Figure based on 8,734 STHs (M = 20.75, SD = 14.79).
30
Table 1 Determinants of football spectator no-show behavior in the Bundesliga
Schreyer
et al.
1
Schreyer
et al.
2
Schreyer
& Däuper
Schreyer
et al.
Frevel &
Schreyer
Schreyer
No-show-rate increases…
(
2016
)
(
2018
)
(
2018
)
(
2019
)
(
2020
)
(
2019
)
17
Matches
17
Matches
704
Matches
710
Matches
285
Matches
1,149
Matches
Economic aspects
Market size ↓ …rather not (no significant effect).
Unemployment ↓ …rather not (no significant effect).
Quality aspects
Home win probability
↑↓ ***
…if the winning probability of the home team
increases, then decreases.
Competitive balance
↑ *** ↑
↑ *** …if the absolute difference in the winning
p
robability of home and away team increases.
Competitive intensity +/- …rather not (no significant effect).
Market value +/- *** ↓ ***
Market value, home +/- …rather not (no significant effect).
Market value, away ↓ *** …if market value (away) decreases.
Promotion, home ↓ *** ↓ *** …if home team was not promoted.
Promotion, away ↓ *** …if away team was not promoted.
Pioneer, home + …rather not (no significant effect).
Tradition, away ↓ *** ↓ *** ↓ *** …if away team tradition decreases.
Geographical derby ↓ *** ↓ *** ↓ *** …if match is not a derby.
Pitch quality ↓ …rather not (no significant effect).
Opportunity costs
Midweek match ↑ *** ↑ *** ↑ *** …if match is play midweek.
Match day ↑↓ *** ↑↓ *** ↑↓ *** …until midseason, then decreases.
First half ↑ …rather not (no significant effect).
31
Holidays ↓ …rather not (no significant effect).
Substitute, free-TV ↓ ↓ +/- …rather not (no significant effect).
Temperature ↓↑ *** ↓↑ *** ↓↑ *** ...for extreme temperatures.
Precipitation ↑ *** ↑ ↑ *** …if rain sets in.
Air pressure ↑ …rather not (no significant effect).
Interval ↓ …rather not (no significant effect).
Terrorist attacks ↑ *** ...temporarily after an attack.
Season fixed effects Yes Yes Yes …over time.
Other
Stadium capacity ↑ *** +/- *** ↑ ***
Distributed tickets ↑ *** …if number of tickets increases.
Sold tickets ↑ *** …if number of sold tickets increases.
Sold tickets, season ↑ *** …if number of season tickets increases.
Sold tickets, match ↓ *** …if number of match tickets increases.
Free tickets ↑ …rather not (no significant effect).
Accommodation
Ticket quantity ↑ *** ↑ *** …if number of tickets increases.
Ticket, standing area ↑ *** ↑ *** …if spectator stands.
Ticket, distance to pitch ↑ *** ↑ *** …if distance to pitch increases.
Ticket, cost ↓ *** ↓ *** …if ticket price decreases.
Socio-demographics
Age ↓ *** ↓ *** …if spectator age decreases.
Gender, male ↓ ↓ † …rather not (no significant effect).
Inhabitant ↓ *** …if spectator lives not in host city.
Geographical distance
↑↓ ***
…if the distance in kilometers between home
and stadium increases, then decreases.
Habit, missed match ↑ *** …if spectator missed last match.
Churn ↑ *** …if spectator has already resigned.
Abbreviations and notes:
1
Final specification(s);
2
Excl. an alternative dichotomous dependent variable capturing perfect STH attendance (LOYALTY); significant effect (***).
32
Table 2 Descriptive statistics of explanatory variables, including controls
Explanatory variables Description Source M SD
Accommodation
Price Price of season ticket
(
in CHF
)
Club 23.51 8.14
Standin
g
terraces
1
Season ticket is located in the standin
g
area Club 0.18 0.38
Famil
y
are
a
1
Season ticket is located in the famil
y
are
a
Club 0.05 0.21
Stand
(
fixed effects
)
Stand in which the STH is accommodate
d
Club
Secto
r
(
fixed effects
)
Sector in which the STH is accommodate
d
Club
Sociodemographic
A
g
e STH a
g
e
(
in Years
)
Club 46.27 13.28
Birthda
y
1
STH’s birthda
y
is on matchda
y
Club 0.00 0.05
Female
1
STH is female Club 0.17 0.38
Distance Distance between
p
lace of residents and the stadium
(
in km
)
Club 17.62 132.96
Habit
1, 2
STH has missed last two home matches Club 0.13 0.34
Controls (Quality
3
)
APD
4
Absolute difference in home and awa
y
win
p
robabilit
y
Football-data.co.u
k
0.50 0.13
Market value
(
home
)
Summed market value of the home teams’ matchda
y
s
q
uad
(
in m
€)
Transfermarkt.com 42.78 7.09
Market value
(
awa
y)
Summed market value of the home teams’ matchda
y
s
q
uad
(
in m
€)
Transfermarkt.com 12.95 5.83
Promoted
(
awa
y)
1
Awa
y
team has been
p
romote
d
Kicke
r
.de 0.11 0.31
Tradition
(
awa
y)
Awa
y
team’s
y
ears in the RSL
(
in
y
ears
)
Swiss Su
p
er Lea
g
ue 52.22 24.31
Beeline/Derb
y
Distance between the stadium of home/awa
y
team
(
in km
)
Luftlinie.or
g
104.69 41.30
Controls (Opportunity effects
3
)
First half
1
Match is scheduled in the first half of the seaso
n
Kicker.de 0.50 0.50
Holida
y
s
1
Match is scheduled durin
g
the holida
y
s Schulferien.or
g
0.35 0.48
Midwee
k
1
Match is scheduled durin
g
the wee
k
Kicker.de
Matchda
y
Matchda
y
Kicker.de
Tem
p
erature Avera
g
e tem
p
erature on matchda
y
(
in °C
)
Meteoblue.com
5
14.80 6.93
Preci
p
itatio
n
1
Preci
p
itatio
n
Meteoblue.com
5
0.26 0.44
Interval/
p
ause Absolute number of da
y
s
p
ast since the last home match Kicker.de 20.42 16.81
Abbreviations and notes: All figures are rounded;
1
Dummy variable (Yes = 1; otherwise = 0);
2
Unavailable for home match #01 and #02;
3
In the Appendix, we present extended
specifications, i.e., including control variable effect sizes (cf., Table A1);
4
In line with previous research, we calculate APD using adjusted probabilities after excluding the
bookmakers’ margin (cf., Benz et al., 2009);
5
all values as of matchday, 12:00.
33
Table 3 Determinants of season ticket holder no-show behavior in Switzerland
Match-Level
5
Full period of investigation
(01a) (01b) (02a) (02b) (03a) (03b) (03c) (04) (05)
Probit
regression
5
Probit
regression
Probit
regression
5
Probit
regression
Probit
regression
Probit
regression
Probit
regression
Poisson
regression
Fractional
probit
re
g
ressio
n
Accommodation
Price -.0182** -.0336*** -.0154** -.0261*** -.0208*** -.0278*** -.0278*** -.0386*** -.0333 ***
.0056 .0019 .0055 .0021 .0056 .0019 .0019 .0073 .0068
Price
2
.0004*** .0007*** .0003** .0006*** .0005*** .0006*** .0006*** .0009*** .0008 ***
.0001 .0000 .0001 .0000 .0001 .0000 .0000 .0002 .0001
Standin
g
terraces
1
.2601*** .1295*** .2424*** .0924*** .1633*** .0920*** .0918*** .1475*** .1294 ***
.0339 .0112 .0321 .0116 .0331 .0121 .0120 .0346 .0305
Famil
y
are
a
1
.3013*** .1964*** .2859*** .1670*** .2965*** .1725*** .1723*** .2518*** .2171 ***
.0468 .0160 .0447 .0176 .0460 .0168 .0168 .0519 .0463
Stand
(
fixed effects
)
Yes Yes Yes Yes Yes Yes Yes Yes Yes
Secto
r
(
fixed effects
)
Yes Yes Yes Yes Yes Yes Yes Yes Yes
Socio-demographics
A
g
e .0393*** .0230*** .0355*** .0186*** .0236*** .0184*** .0184*** .0299*** .0232 ***
.0028 .0009 .0027 .0010 .0028 .0010 .0010 .0040 .0032
A
g
e
2
-.0003*** -.0003*** -.0003*** -.0002*** -.0003*** -.0002*** -.0002*** -.0003*** -.0003 ***
.0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000 .0000
Birthda
y
1
.2424*** .1905*** .2504*** .2054*** .2460*** .1966*** .1971***
.0336 .0405 .0342 .0403 .0347 .0384 .0386
Female
1
.0009 .0027 -.0047 -.0060 -.0146 -.0068 -.0068 .0025 .0021
.0197 .0059 .0185 .0057 .0191 .0060 .0059 .0200 .0172
Distance .0030*** .0024*** .0029*** .0019*** .0028*** .0020*** .0020*** .0020*** .0024 **
.0003 .0001 .0002 .0001 .0002 .0001 .0001 .0006 .0007
Distance
2, 3
-3.46e*** -2.87e*** -3.29e*** -2.24e*** -3.24e*** -2.31e*** -2.31e*** -2.45e** -2.89e **
2.99e 1.64e 2.82e 1.55e 2.91e 1.55e 1.55e 7.20e 9.31e
34
Habit .3189*** .9176*** .2897*** .9236*** .9234***
.0057 .0248 .0058 .0161 .0158
Controls
4
No NoNoNoYes
,
Yes
,
Yes
,
No No
Full set Full set Reduce
d
Estimatio
n
Panel Poole
d
Panel Poole
d
Panel Poole
d
Poole
d
Cluster
(
SE
)
Matchda
y
Matchda
y
Matchda
y
Matchda
y
Observations 628
,
848 628
,
848 611
,
380 611
,
380 611
,
380 611
,
380 611
,
380 8
,
734 8
,
734
Grou
p
s 8
,
734 8
,
734 8
,
734
Matches 72 72 70 70 70 70 70 72 72
Wald chi2 842.91 3
,
321.31 4
,
169.02 5
,
776.60 26
,
596.38 14
,
991.28 12
,
321.62 511.68 498.42
Prob > chi2 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
Pseudo R² 0.0119 0.0615 0.0880 0.0875 0.0475 0.0120
Abbreviations and notes: All figures are rounded; Standard errors (SE) in bold; *, ** and *** indicate statistical significance at the 5% (p < .05), 1% (p < .01) and 0.1% (p < .001)
level, respective;
1
Dummy variable;
2
Squared term;
3
e-07 and e-08 for coefficients and robust standard errors, respectively;
4
For the full list of controls, including effect sizes,
please see Table 3b;
5
All effects are robust when adding matchday dummies to control for unobserved match-level characteristics.
Appendix
Table A1 Determinants of season ticket holder no-show behavior in Switzerland
(03a) (03b) (03c)
Probit
3
Probit
3
Probit
3
Quality aspects
APD .7619*** .5899* .5848*
.0228 .2559 .2501
Market value
(
home
)
.0016*** .0026
.0003 .0033
Market value
(
awa
y)
-.0135*** -.0133** -.0174***
.0006 .0044 .0048
Promoted
(
awa
y)
1
-.0142* -.0115
.0064 .0678
Tradition
(
awa
y)
-.0022*** -.0017
.0001 .0011
Beeline/Derb
y
.0010*** .0010
†
.0011*
.0001 .0006 .0006
Opportunity costs
First half
1
-.0454*** -.0274
.0094 .0870
Holida
y
s
1
-.0187*** -.0230
.0052 .0505
Midwee
k
1
.0023 .0255
.0059 .0640
Matchda
y
-.0144*** -.0120* -.0104**
.0005 .0056 .0033
Tem
p
erature -.0226*** -.0172*** -.0180***
.0004 .0042 .0042
Preci
p
itatio
n
1
.0062 .0155
.0046 .0412
Interval/
p
ause .0018*** .0021
†
.0023*
.0001 .0012 .0010
Season FEs YES YES YES
Ex
p
lanator
y
variables
2
Yes
,
Yes
,
Yes
,
Full set Full set Full set
Estimatio
n
Panel Poole
d
Poole
d
Cluster
(
SE
)
Matchda
y
Matchda
y
Observations 611
,
380 611
,
380 611
,
380
Grou
p
s 8
,
734
Matches 70 70 70
Wald chi2 26
,
596.38 14
,
991.28 12
,
321.62
Prob > chi2 0.0000 0.0000 0.0000
Pseudo R² 0.0880 0.0875
Abbreviations and notes: All figures are rounded. Standard errors (SE) in bold; †, *, ** and *** indicate statistical
significance at the 10% (p < .10), 5% (p < .05), 1% (p < .01) and 0.1% (p < .001) level, respective;
1
Dummy
variable;
2
For the full list of explanatory variables, including effect sizes, please see Table 3a;
3
Reported effects
are robust to employing additional logit specifications, and the exclusion of those STH with either zero (Pseudo
R² = 0.0863) or perfect attendance (0.0865), as well as employing alternative habit dummies capturing the omission
of either only one (0.0853), three (0.0818), four (0.0766), and five (0.0699) subsequent home match(es), despite
the gradually decreasing number of observations.
