Doping in Sports: A Compliance Conundrum

Abstract

This chapter reviews the history of doping regulations, contemporary anti-doping policies and the effectiveness thereof, as well as the public’s perception of the current state of doping in sports. We discuss how detection, testing and punishment influence compliance and, ultimately, the prevalence of doping. We offer a general framework to understand why anti-doping objectives are difficult to achieve. Finally, we assess some of the proposed solutions to improve current anti-doping policies.

Full text

65
Doping in Sports: A Compliance Conundrum
Jeffrey Cisyk and Pascal Courty
Abstract: This chapter reviews the history of doping regulations, contemporary
anti-doping policies and the effectiveness thereof, as well as the public’s percep-
tion of the current state of doping in sports. We discuss how detection, testing and
punishment influence compliance and, ultimately, the prevalence of doping. We
offer a general framework to understand why anti-doping objectives are difficult
to achieve. Finally, we assess some of the proposed solutions to improve current
anti-doping policies.
65.1 introduction
What could be more disturbing than the revelation that one’s beloved sports hero has used
performance-enhancing drugs (PED) to cheat their way to victory? The illicit use of PEDs,
the practice colloquially known as doping, compromises the credibility of fair competition
and our faith in the spirit of sports. Our animosity towards doping co-exists with the silent
truth that most of us know that doping is rampant in many sports. Although doping is the
object of intense controversies within sporting bodies, the media and academia, it seems that
the end of doping in sports is not within sight.
This chapter reviews the literature on doping and explains why compliance with doping
regulations is controversial. Scholars from medicine, sports management, ethics, law and
economics have different – and sometimes opposing – views on how to approach doping
and how to improve anti-doping policies. While the consensus is still in favour of a zero-
tolerance approach, this chapter will discuss why an outright elimination of doping is likely
impossible.
Anti-doping raises a myriad of issues that cannot be adequately covered in a short chapter.
While this chapter reviews the literature, we had to be selective, and we refer the reader to
specialized sources for topics that cannot be given fair treatment. The main contribution of
this chapter is to propose a simple framework to explain why compliance with doping
regulations raises unique issues. At the heart of the problem is the very high aspiration of
the anti-doping agenda (zero-tolerance) paired with great limitations in detection technolo-
gies (testing). We then summarize our analysis with a statement of the sports compliance
conundrum.
We dedicate the end of this chapter to reviewing current proposals to improve doping
regulations. While some of these proposals (e.g., collective responsibility, deferred com-
pensation) have great potential, we conclude that what is mostneedednowarereliable,
comparable measures of prevalence and systematic evaluations of anti-doping
interventions.
949

65.2 current state of anti-doping regulations
Although doping is not new to sports, formal anti-doping regulations are. The past thirty years
have seen wide adoption of policies across countries, sport federations and sporting events that
have evolved towards a detection-based deterrence approach established upon a list of banned
substances. These deterrence regulations, known at secondary measures, complement pri-
mary interventions that attempt to educate athletes and the sports community about the
consequences of doping and to establish drug-free norms. The current system is mostly
repressive and punitive, where punishment takes, in many instances, the form of
a participation ban.
However, there is still much variation across counties, federations and sporting events in
the definition of doping and in the implementation of doping countermeasures. In 1999, the
World Anti-Doping Agency (WADA) was created in part to standardize these rules and
regulations for international sport. WADA’s vision is “a world where all athletes can compete
in a doping-free sporting environment,” and the main justifications for anti-doping are to: (1a)
protect the athlete to participate in doping-free sport and thus promote (1b) health and (1c)
fairness (level playing field); (2) coordinate international anti-doping programs with regard to
(2a) detection, deterrence, and punishment, and (2b) prevention (Møller 2016).
Despite its best intentions, there is considerable controversy and disagreement about
fundamental questions regarding these aforementioned goals of WADA’s code. While few
would challenge the objectives stated in (1a–1c), many debate whether (2a) is the appropri-
ate approach to achieve these objectives and whether these objectives are attainable at all.
The literature has questioned the type of institutions needed to enforce anti-doping
measures, what compliance means, whether to emphasize a preventative or a punitive
approach, and so on. Another debate questions the current zero-tolerance and repressive
approach to doping (Kayser, Mauron and Miah 2007; Mazanov and Connor 2010)dueto
high consumer demand for the opportunity to witness the exceptional events that PEDs are
purported to produce.
65.3 the building blocks of anti-doping regulations
There are fundamental challenges in answering basic empirical questions about the preva-
lence of drug use in most sports, let alone questions about compliance or about the
effectiveness of anti-doping regulations. We review what we know about doping in sports
starting from what we know best (detection and deterrence) and moving to issues for which
we have less precise knowledge (prevalence and compliance).
65.3.1Detection
The current standard in detecting PED use involves the testing of urine and/or blood samples
against a list of banned substances. To be prohibited and featured on WADA’s list of banned
substances, known as a ‘negative list’, a substance must satisfy two of three requirements: (1)
enhance performance; (2) have potential health risk; (3) violate the spirit of the sport (WADA
2019). A key criticism of the current enforcement approach is that it is hard to determine what
should be on a negative list. Caffeine, for example, has been on some negative lists while other
performance-enhancement techniques such as altitude training or hypobaric chambers are
difficult to evaluate on the grounds of criteria (2) because they do not have well-understood
950 Jeffrey Cisyk and Pascal Courty

long-term effects (Vlad et al. 2018). A recent study supported by WADA evenrecommends that
spinach be added to the negative list (Isenmann et al. 2019).
Anti-doping testing is not cheap: Maenning (2014) reports that WADA ran 270,000 tests in
2012. The laboratory cost of these tests was $228 million and adding the costs of administering
each test brings the overall cost to about $500 million. Mountjoy et al. (2017) calculate that in
2015,28 summer Olympic International Federations ran 33,000 tests at a cost of $27.7million.
These costs do not include costs associated with challenges to test outcomes and litigation
around enforcement. Table 65.1reports testing policies in the four major North American
leagues. A large number of athletes are tested a couple of times each year.
Current regulations have given birth to a doping industry that tries to stay one step ahead of
an anti-doping industry. New drugs and delivery methods are constantly arising. In an attempt
to keep up, agencies update negative lists and design new testing protocols (Mazanov and
Connor 2010). Therefore, a ‘window of opportunity’ emerges wherein a new drug may be used
until it has been proven performance enhancing and/or risky and a testing protocol has been
adopted. Firms have an incentive to innovate with ways to avoid existing tests and to develop
‘designer drugs’ that are not yet on negative lists. While many have questioned the accredit-
ation of testing labs and the secrecy of testing protocols, there is a trade-off between test
transparency and revealing weaknesses that could then pave the way for masking strategies
and selective use of PEDs. A more fundamental problem is that athletes are not punished for
the substances that are not on negative lists. Due to the cat-and-mouse nature of anti-doping,
there is much debate over whether a negative list can ever be exhaustive (Mu
¨ller 2010).
Adverse test rates are surprisingly low. In 2012, WADA reported a rate of adverse analytical
finding of 1.76 per cent for a worldwide total of 267,000 samples analysed. Low adverse test
rates hold for most sports and sporting events. Section 65.3.3on prevalence will contrast the
testing evidence with the widespread belief that a significant fraction of athletes use PEDs.
An important advancement in detecting technology is the biological passport wherein
athletes are tested against their own baseline levels of various drug-markers. The athlete’s
passport records longitudinal results of, for example, blood or urine samples. Having
a complete history of an athlete’s test profile greatly increases the power of tests by using
temporal deviation from the athlete’s own normality: instead of using direct detection from
table 65.1 PED testing in the four major North American sports leagues
Baseball Hockey Football Basketball
MLB NHL NFL NBA
Number of tests 7,400 2,199 7,036 2,135
league-wide, per season
Size of testing pool 750 713 1,696 450
approximate
Number of tests ≥3≥3≥1≤6
per athlete per season
Sources: MLB and MLBPA Joint Drug Prevention and Treatment Program, expiring 1December 2021; NHLPA and NHL
Collective Bargaining Agreement, 16 September 2012–15 September 2022; NFL Policy on Performance-Enhancing
Substances, 2018; NBA and NBPA Collective Bargaining Agreement, 19 January 2017.
65 Doping in Sports: A Compliance Conundrum 951

a population norm, the passport allows intra-individual temporal variations in variables that
are sensitive to drug use. Although building statistical models using a wide range of variables
(longitudinal test results, behavioural variables, performance outcomes, etc.) is relatively
new, it is a promising agenda for broader and more accurate detection technologies.
65.3.2Deterrence
Deterrence incentives can take many forms. To start, athletes take into account health risk,
and the perception and understanding of these risks increase with exposure to information
and education campaigns (Mazanov, Huybers and Connor 2011). In addition, sports have
adopted or have considered adopting a number of punishment mechanisms based on test
outcomes:
1. competition ban
2. stripping medals and honours, and repayment of prize money
3. disclosure of test outcomes (negative impact on sponsorship)
4. fines and financial penalties.
Deterrence incentives vary greatly across sports and levels of competition. To illustrate, Table
65.2reports the punishment in several of the major sports leagues. The first offence can have
significant impact on most athletes’ incomes. The third offence completely eliminates the
value of the athlete’s human capital in five out of seven leagues.
65.3.3Prevalence
Low adverse test rates do not necessarily mean that most athletes comply with anti-doping
rules. Because one would expect that athletes who dope do so strategically to minimize the
chance of persecution, the incidence of positive tests should be lower than prevalence.
table 65.2 Competition bans for positive PED test as a percentage of
a sport’s season
Sport League 1st Offence 2nd Offence 3rd Offence
Football CFL 11%50%100%
NFL 25%63%200%
Hockey NHL 24%73% Lifetime
Basketball NBA 30%67% Lifetime
Baseball MLB 49%100% Lifetime
College sports NCAA 100% Lifetime
Soccer FIFA 400%800% Lifetime
Sources: CFL Drug Policy at a Glance – www.cfl.ca/2015/05/05/cfl-drug-policy-glance/;
NFL Policy on Performance-Enhancing Substances, 2018; NHLPA and NHL
Collective Bargaining Agreement, 16 September 2012–15 September 2022; NBA and
NBPA Collective Bargaining Agreement, 19 January 2017; MLB and MLBPA Joint
Drug Prevention and Treatment Program, expires 1December 2021; NCAA Frequently
Asked Questions about Drug Testing – www.ncaa.org/sport-science-institute/topics/
frequently-asked-questions-about-drug-testing; FIFA Anti-Doping Regulations, 2018
Edition.
952 Jeffrey Cisyk and Pascal Courty

In fact, there is little doubt that the prevalence of doping is much higher than positive test
results would indicate. Even WADA acknowledges that “testing has not proven to be
particularly effective in detecting doping.” However, estimating the prevalence of an illegal
practice is challenging. The simplest approach uses anonymous questionnaires of self-
reported drug use. This relies, however, on a profound assumption that respondents are
comfortable revealing compromising information. Some recent studies use the randomized
response technique, which is a slightly more complicated survey approach that guarantees
respondents’ confidentiality (Pitsch and Emrich 2011; de Hon, Kuipers and van Bottenburg
2015; Ulrich et al. 2018). Other methods include inference from changes in athletic perform-
ance, models of biological parameters, response to remission incentives, and testimonies by
retired athletes who fall outside statutes of limitations.
Pitsch and Emrich (2011), de Hon, Kuipers and van Bottenburg (2015) and Pielke (2018)
review these various methods. Although there is much variation in estimates of doping
prevalence, most studies agree that prevalence is high and most likely in the two-digit
range. For example, de Hon, Kuipers and van Bottenburg (2015) conclude that ‘the preva-
lence of doping in elite sports is likely to be between 14 and 39%’. Maennig (2014) offers
a minimum bound, stating that ‘at least 10% of all elite athletes are doping’.
65.3.4Test Validity: Sensitivity and Specificity
The validity of a test is measured by its sensitivity and specificity. Sensitivity is the probability
of a positive test conditional on actually doping: the probability of a ‘true positive’. Sensitivity
is therefore the probability of catching a non-complier. Conversely, specificity is a measure of
true negative: the probability of a negative test conditional on not doping. We denote
sensitivity by sand specificity by f. Together, sensitivity and specificity make up the adverse
test rate, or the probability of a positive test.
1
Although sensitivity and specificity are essential to understanding the deterrence effect of
testing, WADA and test laboratories do not disclose this information (Mazanov and Connor
2010). Sensitivity has been estimated for specific drugs and testing protocols. Hermann and
Henneberg (2014) review the literature and argue that ‘the highest rates of success of doping
detection have been reported as being 60% and 80% success rate’, with rates being as low as
10 per cent for select drugs and testing protocols.
When sensitivity or specificity is not perfect ( s<1and/or f < 1), making inference about
doping from laboratory tests requires making a Bayesian inference, updating one’s prior
probability of prevalence. This inference may be based on a single or multiple test result(s).
Berry (2008) argues that, ‘due to inherent flaws in the testing practices of doping laboratories’,
it is not possible to conclude that an athlete is guilty of doping when they test positive. Pitsch
(2009) comes to the similar conclusion that imperfect testing methods based on ‘Bayesian
logic lead to important ethical questioning of anti-doping policies’.
65.3.5Compliance
The same way that low adverse test rates do not mean that most athletes comply with anti-
doping rules, a high prevalence rate does not necessarily mean that compliance is low. For
1
Let d¼1 if an athlete dopes and d¼0 otherwise, and let t¼1 if the athlete tests positive and t¼0 otherwise. We
have: s¼Prðt¼1jd¼1Þ,f¼Prðt¼0jd¼0Þ, and Prðt¼1Þ¼sPrðd¼1Þþð1fÞPrðd¼0Þ.
65 Doping in Sports: A Compliance Conundrum 953

example, it may be that 70 per cent of athletes comply with anti-doping rules when preva-
lence is 30 per cent. Compliance should be measured as the reduction in doping associated
with a specific regulation or enforcement mechanism. If all athletes dope in the absence of
testing, the compliance rate is equal to one minus the rate of prevalence under anti-doping, as
computed in the example in Section 65.3.4.
One may argue that compliance is high on the grounds that prevalence, without any
doping regulations, would be much higher than current levels. In fact, historical evidence
that predates the rigorous modern anti-doping detection (i.e., pre-1990s) suggests a high
demand for doping. The Goldman dilemma points to the same conclusion wherein about
half of surveyed elite athletes reported that they would take a drug that guarantees success
even if it would cause death within five years (Goldman, Bush and Klatz 1984). Although
Connor, Woolf and Mazanov (2013) have questioned the validity of the dilemma, other
studies suggest that the demand for safe and undetectable doping is high (Overbye, Knudsen
and Pfister 2013) and survey questionnaire responses point to the same conclusion.
Another approach to understanding compliance is to compute the probability that an
athlete – who strategically dopes to avoid detection – is actually caught given existing testing
rules. This probability accounts for a test’s sensitivity, the window of detection (the time span
when a drug can be detected) and the frequency of tests. Hermann and Henneberg (2014)
conclude that this probability is low for most sporting events and for most drugs used. This is
because, too often, test sensitivity is low, the window of detection is short, and tests are
infrequent. They conclude that it is easy to avoid testing positive for ‘narrow-window drugs’.
This explains the low rates of adverse findings and points towards the fundamental problem
that athletes can avoid detection.
65.4 sport’s compliance conundrum
This section explains why compliance with anti-doping regulations in sports raises issues that
are absent in standard applications of deterrence theory. The core of the discussion is
technical and the reader not interested in the mechanics of compliance may move directly
to the main results stated in our Proposition below.
Denote policy p32 fa;∅g, where p¼aif there is an anti-doping policy and p¼∅
otherwise. Let dp
i¼1 if athlete i32 Idopes under policy pand dp
i¼0 otherwise. There are
three types of athlete:
1. unconditional non-compliers : d∅
i¼da
i¼1
2. unconditional compliers : d∅
i¼da
i¼0
3. conditional compliers: d∅
i¼1 and da
i¼0.
Doping prevalence under proposed anti-doping regulations is composed of the unconditional
non-compliers, Prðda
iÞ. The deterrence effect of anti-doping regulations is measured by the
fraction of conditional compliers, Prðd∅
iÞ Prðda
iÞ. If all athletes dope in the absence of
anti-doping regulations (Prðd∅
iÞ¼1), we obtain that the deterrence effect of anti-doping
regulations is equal to one minus the prevalence rate under anti-doping, 1 Prðda
iÞ.
If the benefit of doping is band the cost of non-compliance is c, an athlete complies if
b≤cs where sis the test’s specificity as defined in Section 65.3.4. Deterrence theory says that
athletes comply with doping regulations so long as the punishment when being caught is
large enough, cb=s. To illustrate, take the evidence discussed in Table 65.2and set the
value of s=0.5and cto six months’ salary loss. We obtain that athletes who can gain at least
954 Jeffrey Cisyk and Pascal Courty

one year of additional salary from doping do not comply. This is plausible as compensation
schemes in elite sports are characterized by high pay-offs from small, incremental perform-
ance improvements (Rosen 1981).
The punishment must increase as sensitivity decreases or as the benefit from doping
increases. In the case of individual sport with a winner-takes-all prize scheme, for example,
the benefit from doping is equal to the incremental probability of winning multiplied by the
prize. Thus, until otherwise corrected, doping should prevail for effective drugs and within
sports with high financial rewards. Circumstantial evidence supports this prediction. For
example, Maennig (2014) argues that the commercialization of sports and the increase in
benefit bshould decrease compliance, consistent with the observation that doping is com-
mon in media sports with large financial rewards and in sports where doping has a significant
impact on winning (e.g., weightlifting).
We now turn to our main result. Recall WADA’s key objectives from Section 65.2:
Definition: Anti-doping aims for: (1) drug-free sport, defined as Prðda
iÞ¼0; and (2) fair
competition (‘level playing field’), defined as drug-free sport without false negative test
results, that is, Prðda
iÞ¼0 and f¼1.
The drug-free sport condition embodies the zero-tolerance approach to doping. The level
playing field condition adds the requirement that no athlete can be wrongly accused of
doping. We can state the sport’s compliance conundrum as follows:
Proposition: Anti-doping interventions based on testing and punishment fail to achieve
WADA’s key objectives for two reasons: (1) unlike standard deterrence applications, a drug-
free sport objective does not value incremental increases in compliance, only absolute
compliance (which is difficult to achieve with low test sensitivity and/or high benefits to
doping); and (2) a level playing field cannot hold with imperfect specificity (f<1).
Three outcomes that we label the ‘paradox of doping’ can happen simultaneously: high
deterrence (Prðd∅
ida
iÞis high); high prevalence (Prðda
iÞis high); and low adverse test rate
(Prðt¼1Þis low). According to the literature, prevalence is around 30 per cent, an upper
bound for the deterrence effect of anti-doping is 70 per cent, and the adverse test rate is
2per cent. A low adverse test rate occurs because doping tests have low sensitivity. To achieve
high deterrence, low sensitivity is balanced with powerful punishments (high costs to doping;
see Table 65.2) that debase the value of a non-complier’s human capital. The challenge is that
the benefits to doping are large in commercialized sports. When combined with a low test
sensitivity, full compliance is not achievable under realistic punishment schemes.
Although these three outcomes are expected in any use of deterrence elsewhere, they
generate a paradox in sports because the objective of anti-doping is absolute, rather than
incremental, compliance in order to eliminate the highly unpopular suspicion that
a competition was won by a non-complier. To contrast compliance in doping with other
domains, take crime deterrence as an illustration. An increase in the punishment may reduce
crime’s prevalence, and any incremental reduction in crime is valuable to society. Although
important, incremental reduction in doping, however, is not the main concern in sports. The
issue at stake is that doping gives some athletes an unfair advantage. While crimes can be seen
as independent events that can be treated separately, the fact that athletes compete against
one another makes doping a collective problem. The goal is not to reduce doping down to
a given pre-set target level of non-compliance but to eliminate it outright so that no athlete
can get an unfair advantage. Therefore, any non-zero level of prevalence compromises the
65 Doping in Sports: A Compliance Conundrum 955

spirit of competition. The high prevalence rate speculated to exist in many sports suggests that
current penalties for doping are not sufficiently high. Worse, penalties would arguably have to
be set at unrealistically high levels to achieve full compliance.
The other concern originates when imperfect specificity compromises the second object-
ive of anti-doping regulations. When specificity is imperfect, athletes may be wrongly
punished. This challenges a deterrence approach because it compromises fairness (Berry
2008; Delanghe et al. 2014). An anti-doping programme that deters doping and achieves low
prevalence may fail the level playing field objective if some athletes who do not dope are
wrongly punished due to imperfect testing.
As emphasized in the model, imperfect specificity goes against ‘fairness and equality’
because it introduces uncertainty to the competition. Considerations should be given to
a substance’s ability to enhance performance when specificity is low. The negative list should
therefore trade-off specificity and performance enhancement.
65.5 current proposals
Many proposals have been made to improve current anti-doping programmes: Section 65.3.1
has already discussed progress in detection technologies with the recent introduction of the
biological passport. We focus here on proposals that, to the best of our knowledge, have not
yet been implemented.
65.5.1Collective Responsibility
Teams, leagues and federations do not bear many of the costs of doping. Cisyk and Courty
(2017) find small declines in attendance at Major League Baseball games following the
announcement of a PED suspension and there is evidence of similar negative responses for
television audiences or endorsements in other sports (Cisyk 2020; van Reeth 2013). Although
such demand responses demonstrate that the public cares about PED use, the financial
consequences are limited. Taking into account the positive impact of doping on performance
and overall demand, it is not implausible that event organizers sometimes benefit from poorly
detected non-compliance, or, stated differently, lose from taking an aggressive stance on
doping (Mazanov and Connor 2010).
2
For example, random testing in the Ultimate Fighting
Championship may negatively affect revenue when last-minute competition bans force
cancellations to matches.
Maennig (2014) proposes holding all parties involved in a competition collectively respon-
sible for doping outcomes. For example, one could ban the broadcast of high-doping-
prevalence sports the same way that countries with state-sponsored doping programmes
have been banned from participating in the Olympic Games. Broadcast bans would add
financial consequences through loss of sponsorship and advertisement. One would expect
sports to develop norms that would internally discipline or exclude non-compliers. Such
norms are not unusual; for example, golfers have the obligation to report rule violations by
competitors. Even informal sports, such as pick-up basketball, require each player to police
others’ actions under the ‘call your own fouls’ rule.
3
2
An exception is the cancellation of the 2009 Tour of Germany (cycling) which was attributed to the pull-out of
public broadcasters after doping revelations.
3
‘Call your own fouls’ requires players to explicitly announce when their opponent has committed an infraction.
956 Jeffrey Cisyk and Pascal Courty

Along the same lines, Maennig (2014) suggests that funding for inter-sport anti-doping
agencies should be proportional to the sport’s prevalence rather than to its share of the testing
pool. This would shift the burden on doping-prone sports to take a more active role against
doping.
65.5.2Leniency and Self-Reporting
An effective way to understand prevalence is to offer leniency for self-reporting and
whistle-blowing. Such exemptions from punishment incentivize athletes to reveal
shortcomings in existing testing protocols. The National Basketball Association
(NBA), for example, allows for a one-time voluntary self-admission of any prohibited
substance without penalty or public announcement. Following the NBA model,
sports’ governing bodies may handle self-reporting discreetly to reduce the risk of
public backlash and boycott.
Another form of exemption from doping punishments includes statutes of limitation.
Athletes tend to withhold information about doping until the end of their career. For
example, Figure 65.1shows that baseball players are more likely to admit to doping at or
near the end of their career. However, remission incentives and statutes of limitation for self-
reporting are double-edged because one would like to hold athletes accountable in order to
preserve the credibility of deterrence.
65.5.3Life-Cycle Considerations
When repression takes the form of a lifetime competition ban, athletes at the end of
their career face a lower cost to exposure because their human capital fully depreci-
ates at retirement. Moreover, athletes near the end of their career may benefit more
from doping because they have to keep pace with younger athletes. For these reasons,
the deterrence approach is less effective for athletes who are approaching retirement.
A solution to this problem would be to require athletes to contribute a portion of their
annual earnings to a fund that would be used as a bond in the event of doping violations. In
such a deferred compensation scheme, an athlete’s earnings would be held in escrow until
the culmination of a doping-free career in order to solve the end-of-career problem (Maennig
2014). As an aside, note that deferred compensation also solves the problem of fine enforce-
ment. Many sports punish doping violations with fines or by requiring the athletes to repay
prize awards. Collecting these fines, however, has proven difficult when, for example, an
athlete is near bankruptcy. Enforcement would be easier if each athlete had a personal escrow
fund.
65.5.4Governance
While anti-doping agencies were originally created to protect athletes, some have
since argued that these agencies perpetuate anti-doping rhetoric in hopes to gain
legitimacy, public support and, ultimately, additional funding (Mazanov and Conner
2010). For example, with its strict laboratory accreditation process, WADA has created
a monopoly over drug testing and has been accused of focusing on the quantity over
quality of tests while keeping a veil of opacity over the science of testing (Maennig
2014).
65 Doping in Sports: A Compliance Conundrum 957

The WADA centralized model for anti-doping was established upon the urge to stand-
ardize doping regulations that were essential to allow athletes from different countries and
sports to compete on a fair playing field. At the same time, a centralized approach to doping
risks stifling innovation and experimentation to new anti-doping ideas. A mixed approach
towards harmonization encourages federations, leagues and countries to adopt innovative
and un-tested anti-doping policies. This is happening to some extent with the for-profit
leagues that largely self-regulate (e.g., the leagues presented in Tables 65.1and 65.2),
although even in these instances there are pressures from politicians and sponsors to follow
international anti-doping norms.
65.5.5Policy Evaluation
We are not aware of standardized ex post evaluations that assess currently used anti-doping
policies. Such policy evaluations are standard in other domains (e.g., education, welfare
intervention, etc.) and established on well-accepted scientific methods (Athey and Imbens
2017). Estimation of the causal impact of an anti-doping intervention on compliance may
0
25
50
75
100
Percentage (%)
30 35 40 45
Age
figure 65.1 Age of MLB players who admitted to doping
Note: The figure displays the probability that one of the eighty-nine Major League Baseball (MLB)
players admitted to doping after being accused by an independent investigation led by the US
Congress in 2007. Of the implicated players, twenty-eight (31 per cent) admitted to PED use,
noticeably those in the latter stages of their career or already retired.
Source: Report to the Commissioner of Baseball of an Independent Investigation into the Illegal
Use of Steroids and Other Performance Enhancing Substances by Players in Major League
Baseball, or ‘Mitchell Report’ (Mitchell 2007).
958 Jeffrey Cisyk and Pascal Courty

now be possible given the current progress towards constructing reliable longitudinal meas-
ures of prevalence.
Surprisingly, much of the anti-doping resources is dedicated towards managing and
implementing detection; anti-doping agencies have so far not bothered estimating prevalence
using the methods described in Section 65.3.3, or, for that matter, the impact of their own
interventions using modern methods of policy evaluation. Specifically, debates around the
effectiveness of the current detection-based deterrence agenda rely on speculative theories
rather than rigorous causal evidence (Kayser, Mauron and Miah 2007; Mazanov and Connor
2010).
A different and indirect way to evaluate the effectiveness of anti-doping programmes would
be to use the simple framework presented in Section 65.4. We are not aware of any study that
investigates whether fines and punishments respond as predicted by deterrence theory. The
point would be to compare anti-doping programmes (using, for example, the information in
Tables 65.1and 65.2in the case of North-American leagues) across sports, countries, feder-
ations and events and to investigate whether testing and punishment policies are designed to
optimally deter doping. For example, punishment should increase with higher stakes and
lower test sensitivity. The point would be to investigate whether anti-doping programmes are
designed in a way that is consistent with the goal of minimizing doping.
65.6 summary and conclusions
We have reviewed current anti-doping policies and their effectiveness, the discontent about
the state of doping in sports, and offered a general framework to understand why anti-doping
objectives are difficult to achieve. Anti-doping in sports does not aim for an incremental
reduction in doping but for a complete elimination as predicated under the ‘doping-free
sport’ objective of most anti-doping policies falling under the umbrella of WADA. The
current state of discontent with WADA’s anti-doping approach is that a detection-based
deterrence approach cannot eliminate doping given current detection technologies and
high stakes.
However, significant improvement in testing technologies, such as the biological passport,
could shift the balance in favour of the WADA agenda. Additional proposed methods
including collective punishment, deferred compensation and laboratory competition could
further WADA’s goals. That being said, the public’s knowledge regarding compliance with
doping regulations will be incomplete until (i) modern methods to accurately measure
prevalence are adopted and (ii) systematic evaluations of anti-doping interventions are
undertaken.
references
Athey, Susan, and Guido W. Imbens. 2017. ‘The State of Applied Econometrics: Causality and Policy
Evaluation’. Journal of Economic Perspectives 31, no. 2:3–32. DOI: 10.1257/jep.31.2.3.
Berry, Donald A. 2008. ‘The Science of Doping’. Nature 454, no. 7205:692. DOI: 10.1038/454692a.
Cisyk, Jeffrey. 2020. ‘Impacts of Performance-Enhancing Drug Suspensions on the Demand for
Major League Baseball’. Journal of Sports Economics 21,no.4:391–419. DOI: 10.1177/
1527002520906529.
Cisyk, Jeffrey, and Pascal Courty. 2017. ‘Do Fans Care about Compliance to Doping Regulations in
Sports? The Impact of PED Suspension in Baseball’. Journal of Sports Economics 18, No. 4:323–50.
DOI: 10.1177/1527002515587441.
65 Doping in Sports: A Compliance Conundrum 959

Connor, James, Jules Woolf, and Jason Mazanov. 2013. ‘Would They Dope? Revisiting the Goldman
Dilemma’. British Journal of Sports Medicine 47, No. 11:697–700. DOI: 10.1136/bjsports-2012-
091826.
de Hon, Olivier, Harm Kuipers, and Maarten van Bottenburg. 2015. ‘Prevalence of Doping Use in Elite
Sports: A Review of Numbers and Methods’. Sports Medicine 45, No. 1:57–69. DOI: 10.1007/
s40279-014-0247-x.
Delanghe, Joris R., Thomas M. Maenhout, M. M. Speeckaert, and Marijn L. De Buyzere. 2014.
‘Detecting Doping Use: More Than an Analytical Problem’. Acta Clinica Belgica 69, No. 1:
25–9. DOI: 10.1179/0001551213Z.0000000009.
Goldman, Bob, Patricia J. Bush, and Ronald Klatz. 1984.Death in the Locker Room: Steroids and Sports.
London: Century.
Hermann, Aaron, and Maciej Henneberg. 2014. ‘Anti-doping Systems in Sports Are Doomed to Fail:
A Probability and Cost Analysis’. Journal of Sports Medicine & Doping 4, No. 5:148. DOI: 10.4172/
2161-0673.1000148.
Isenmann, Eduard, Gabriella Ambrosio, Jan Felix Joseph, Monica Mazzarino, Xavier de la Torre,
Philipp Zimmer, Rymantas Kazlauskas et al. 2019. ‘Ecdysteroids as Non-conventional Anabolic
Agent: Performance Enhancement by Ecdysterone Supplementation in Humans’. Archives of
Toxicology 93, No. 7:1–10. DOI: 10.1007/s00204-019-02490-x.g.
Kayser, Bengt, Alexandre Mauron, and Andy Miah. 2007. ‘Current Anti-doping Policy: A Critical
Appraisal’. BMC Medical Ethics 8, No. 1:2. DOI: 10.1186/1472-6939-8-2.
Maennig, Wolfgang. 2014. ‘Inefficiency of the Anti-doping System: Cost Reduction Proposals’.
Substance Use & Misuse 49, No. 9:1201–5. DOI: 10.3109/10826084.2014.912065.
Mazanov, Jason, and James Connor. 2010. ‘Rethinking the Management of Drugs in Sport’.
International Journal of Sport Policy 2, No. 1:49–63. DOI: 10.1080/19406941003634032.
Mazanov, Jason, Twan Huybers, and James Connor. 2011. ‘Qualitative Evidence of a Primary
Intervention Point for Elite Athlete Doping’. Journal of Science and Medicine in Sport 14, No. 2:
106–10. DOI: 10.1016/j.jsams.2010.06.003.
Mitchell, George John. 2007. ‘Report to the Commissioner of Baseball of an Independent Investigation
into the Illegal Use of Steroids and Other Performance Enhancing Substances by Players in Major
League Baseball’.
Møller, Verner. 2016. ‘The Road to Hell Is Paved with Good Intentions – A Critical Evaluation of
WADA’s Anti-doping Campaign’. Performance Enhancement & Health 4, no. 3–4:111–15. DOI:
10.1016/j.peh.2016.05.001.
Mountjoy, Margo, Stuart Miller, Matteo Vallini, Jeremy Foster, and James Carr. 2017. ‘International
Sports Federation’s Fight to Protect the Clean Athlete: Are We Doing Enough in the Fight against
Doping?’ British Journal of Sports Medicine 51:1241–1242. DOI: 10.1136/bjsports-2017-097870.
Mu
¨ller, Rudhard Klaus. 2010. ‘History of Doping and Doping Control’. Doping in Sports: Biochemical
Principles, Effects and Analysis 195:1–23. DOI: 10.1007/978-3-540-79088-4_1.
Overbye, Marie, Mette Lykke Knudsen, and Gertrud Pfister. 2013. ‘To Dope or Not to Dope: Elite
Athletes’ Perceptions of Doping Deterrents and Incentives.’ Performance Enhancement & Health 2,
no. 3:119–34. DOI: 10.1016/j.peh.2013.07.001.
Pielke, Roger. 2018. ‘Assessing Doping Prevalence Is Possible: So What Are We Waiting For?’ Sports
Medicine 48, No. 1:207–9. DOI: 10.1007/s40279-017-0792-1.
Pitsch, Werner. 2009. ‘“The Science of Doping” Revisited: Fallacies of the Current Anti-doping
Regime’. European Journal of Sport Science 9, No. 2:87–95. DOI: 10.1080/17461390802702309.
Pitsch, Werner, and Eike Emrich. 2011. ‘The Frequency of Doping in Elite Sport: Results of
a Replication Study’. International Review for the Sociology of Sport 47, No. 5:559–580. DOI:
10.1177/1012690211413969.
Rosen, Sherwin. 1981. ‘The Economics of Superstars’. American Economic Review 71, No. 5:845–58.
www.jstor.org/stable/1803469.
Ulrich, Rolf, Harrison G. Pope, Le
´a Cle
´ret, Andrea Petro
´czi, Tama
´s Nepusz, Jay Schaffer,
Gen Kanayama, R. Dawn Comstock, and Perikles Simon. 2018. ‘Doping in Two Elite Athletics
Competitions Assessed by Randomized-Response Surveys’. Sports Medicine 48, No. 1:211–19. DOI:
10.1007/s40279-017-0765-4.
960 Jeffrey Cisyk and Pascal Courty

Van Reeth, Daam. 2013. ‘TV Demand for the Tour de France: The Importance of Stage Characteristics
versus Outcome Uncertainty, Patriotism, and Doping’. International Journal of Sport Finance 8,
No. 1:39–60.
Vlad, Robert Alexandru, Gabriel Hancu, Gabriel Cosmin Popescu, and Ioana Andreea Lungu. 2018.
‘Doping in Sports, a Never-Ending Story?’ Advanced Pharmaceutical Bulletin 8, No. 4:529. DOI:
10.15171/apb.2018.062.
World Anti-Doping Agency. 2019. ‘Athlete Biological Passport Operating Guidelines’. www.wada-ama
.org/en/resources/athlete-biological-passport/athlete-biological-passport-abp-operating-guidelines.
65 Doping in Sports: A Compliance Conundrum 961