A Note on the International Coordination of Antidoping Policies

Abstract

We apply ideas put forward in the literature on the international coordination of macroeconomic and trade policies to study the merits of an international coordination of antidoping policies. Without international coordination of antidoping policies, sports associations and National Antidoping Agencies that comply with an international regulatory framework like the World Antidoping Code are at a disadvantage. We sketch a simple game-theoretic model to illustrate this disadvantage, and the advantage of an international coordination of antidoping policies. Finally, we address the limitations of our model and how it could be extended in future research.

Keywords

doping policy coordination duopoly model

Introduction

Many researchers have used techniques of modern game theory to study the effectiveness of antidoping policies. Much of this research has focused on how to set incentives such that athletes comply with antidoping rules (Berentsen, 2002; Bird & Wagner, 1997; Breivik, 1992; Eber, 2008; Eber & Thépot, 1999; Haugen, 2004, to name just a few). Results of this research show that athletes’ incentives to use forbidden illicit substances depend on, for example, the efficiency of doping tests, the fair play norms against doping, and the details of the regulatory system. Less is known about how to set incentives to sports associations and antidoping agencies to enforce antidoping rules. Focusing on the incentives of sports associations and antidoping agencies is important because, as Eber (2002) shows, sports associations and antidoping agencies may have an incentive to tolerate a suboptimal level of doping in the absence of a binding commitment to a strict antidoping policy. In order to enforce a binding commitment to a strict antidoping policy, Eber (2002) suggests that antidoping policy should be delegated to an institution like the World Antidoping Agency (WADA) and that this institution should be independent of sports federations and the International Olympic Committee (IOC).

WADA is a supranational organization that was founded after the World Antidoping Conference that took place in 1999 in Lausanne. It has been registered since 2002 in Montreal. WADA tries to coordinate and harmonize on behalf of the IOC the fight of National Antidoping Agencies (NADAs) against doping.¹ It has developed to this end a program (World Antidoping Code [WADC]) to improve compliance on the side of the international sports federations to the goal of a worldwide doping-free sport (WADA, 2009). The WADC was adopted by the World Conference against Doping in Sport on March 5, 2003, in Copenhagen, and subsequently implemented for the first time on the occasion of the 2004 Olympic Games in Athens.²

In our research, we take Eber’s (2002) research one step further to argue that an international antidoping agency like WADA may also help to resolve the problem of international coordination of antidoping policies. We argue that in the absence of a binding commitment and without the possibility to enforce compliance to the rules, a lack of international coordination of antidoping policies may inflate doping in international competitions. Our research builds on earlier research in the international macroeconomics (Hamada, 1976) and international trade (Spencer & Brandner, 1983) literature, where the effects of international coordination of national (economic) policies have been extensively studied by means of game-theoretic models.

In order to develop our argument, we study the Cournot–Nash equilibrium of a simple two-country duopoly model. While the number of international “players” is greater than two in the real world, we deem a simple two-country model a reasonable approximation of the international competition for Olympic success. Figure 1 shows, based on medal counts for the sample period 1998-2012, that only few countries dominate the Olympic medal tables. The “Top 2” countries in the case of the summer (winter) Olympics are the United States and Russia (Germany and the United States). Beyond the group of leading medal-winning countries, the number of medal counts rapidly decreases. Medal counts, thus, are highly concentrated among a few countries.

In our model, sports associations and NADAs minimize a loss function defined over sporting performance of athletes and the prevalence of doping among athletes. We measure sporting performance of athletes relative to the performance of foreign athletes. Moreover, we assume that sporting performance of athletes can be promoted by using doping in a systematic way. We argue that in the absence of international coordination and a binding commitment, sports associations and NADAs have an incentive to inflate sporting performance of athletes by tolerating doping. If foreign antidoping agencies, in response to such a policy, seize retaliation measures by loosening their fight against doping, the Cournot–Nash equilibrium of our model shifts to a configuration in which overall sporting performance artificially increases in both countries because sports associations and antidoping agencies tolerate doping among athletes. We organize the remainder of this research as follows. In the second section, we develop a simple model to illustrate our argument. In the third section, we discuss our results and offer some concluding remarks.

A Simple Model

The world consists of two countries, home and foreign. Every country is populated by an NADA and athletes. The antidoping agency of the home country minimizes the following loss function³:

L = - λ (m - m^{*}) + (1 / 2) {(d - d_{T})}^{2},

where λ > 0 is a weighting factor, m denotes (the log of) athletes’ sporting performance, d denotes the (log of the) prevalence of doping among athletes, d_T > 0 denotes the doping “target” of the antidoping agency, and an asterisk denotes a foreign variable.⁴ The loss function specifies that the antidoping agency suffers from a loss if the sporting performance, m, of athletes falls short of the sporting performance of foreign athletes, m*. In some sense, our loss function, thus, can be interpreted to capture the “keeping up with the Joneses” effect well known from the macroeconomics literature (see e.g., Pierdzioch 2003). This effect specifies that consumers do not only derive utility from their own consumption, but that, for a given level of own consumption, households feel worse off if other households increase their consumption. In addition, the loss function specifies that the antidoping agency suffers a loss whenever the prevalence of doping among athletes deviates from the doping target. The quadratic functional form captures the idea that a prevalence of doping that is lower than the doping target may result in funding cuts and a lower public antidoping budget. A prevalence of doping that is higher than the agencies’ doping target, in turn, may trigger questions concerning the competence of the antidoping agency. The positive doping target reflects that the capabilities of an NADA to fight doping depend on financial constraints set by its stakeholders (government and sports associations). Even if an NADA wanted to detect as many dopers as possible, available financial resources would be a binding constraint. A positive doping target, thus, can be interpreted as reflecting a pragmatic antidoping policy, given financial constraints. The loss function of the foreign antidoping agency is identical to the loss function given in Equation 1, except that the foreign doping target may differ from the home doping target.

Sporting performance depends positively on the prevalence of doping and negatively on the sporting performance of foreign athletes. The “production function” of sporting performance is given by:

m = c d - b m^{*},

where c > 0 denotes the sensitivity of sporting performance with respect to the prevalence of doping, and 0 < b < 1 denotes the degree of cross-national “substitutability” of sporting performance. In other words, if a foreign athlete wins an Olympic gold medal in a certain discipline, a home athlete cannot win a gold medal in the same discipline. The home athlete, however, can still win a silver or bronze medal, implying that the sporting performance of foreign athletes is not a perfect substitute for the sporting performance of home athletes. We, thus, assume that the parameter b is positive but strictly less than unity. We also note that m can be interpreted as a broad measure of sporting performance. This measure of sporting performance does not necessarily refer only to, Olympic medals.⁵ The foreign production function is identical to the home production function, except that the roles of m and m* are interchanged, and d is replaced by d*.

Minimizing Equation 1 subject to Equation 2, and doing the same for their foreign counterparts, yields the Cournot–Nash equilibrium of the model under discretionary national antidoping policies.⁶ Discretionary national antidoping policy implies that, without a binding commitment, both home and foreign can set their respective doping policies without consulting the other country. The solution of the model for the home country under discretionary national antidoping policy is given by⁷:

d = d_{T} + λ c,

m = c (d_{T} - b d_{T}^{*} + (1 - b) λ c) / (1 - b^{2}) .

The solution for the foreign country is identical to the solution for the home country except that the roles of d_T and $d_{T}^{*}$ are interchanged. Resembling the results reported by Eber (2002) for a closed-economy model, Equation 3 shows that a discretionary national antidoping policy implies that an NADA has an incentive to tolerate doping. The doping bias is given by λc. Even more, the home antidoping agency has an incentive to actively promote the prevalence of doping by setting a higher doping target, d_T , such that the sporting performance of home athletes increases at the expense of the sporting performance of foreign athletes. Formally, this argument can be illustrated by increasing d_T and holding $d_{T}^{*}$ fixed. Equation 4 shows that we have:

\partial m / \partial d_{T} = c / (1 - b^{2}),

which is strictly positive. The corresponding cross-national foreign derivative is given by:

\partial m^{*} / \partial d_{T} = - c b / (1 - b^{2}),

which is strictly negative. The negative cross-national effect of a looser national antidoping policy sets an incentive for the foreign country to seize a retaliation measure by increasing its doping target. If both the home and the foreign antidoping agency increase their doping targets by the same amount, we get:

\partial m / \partial d_{T} = \partial m^{*} / \partial d_{T}^{*} = c (1 - b) / (1 - b^{2}) = c / (1 + b),

which simply is a combination of Equations 5 and 6. This expression is strictly positive, that is, the sporting performance of both countries increases: Athletes run faster, jump longer, and they put the shot farther, in a potentially larger number of international competitions. The increase in sporting performance, however, is artificial insofar as it only traces back to a higher prevalence of doping tolerated by antidoping agencies. Relative sporting performance is unchanged.

In contrast to the mechanism studied by Eber (2002), the mechanism that gives rise to an incentive to tolerate doping arises in our model because an NADA has an incentive to increase the “market share” of home athletes in international sporting competitions at the expense of the market share of foreign athletes. In other words, the mechanism that we study rests on the lack of commitment to fight doping in an international setting, whereas the mechanism that Eber (2002) studies rests on a lack of commitment to fight doping that arises because an antidoping agency is not independent of sports associations. Our mechanism can be effective even if NADAs were completely independent of sports associations. At the same time, if NADAs are not independent of sports associations, this lack of independence is likely to strengthen the mechanism at work in our model. The mechanism that Eber (2002) studies and the mechanism that is the focus of our analysis, thus, may reinforce each other.

The international coordination of antidoping policies can be modeled by assuming a monopolistic international antidoping agency. This international antidoping agency has the exclusive right to enforce antidoping policies. We assume that the international antidoping agency minimizes the sum of the home and the foreign loss function:

L^{c} = (1 / 2) {(d - d_{T})}^{2} + (1 / 2) {(d^{*} - d_{T}^{*})}^{2},

where the loss function, L^c , denotes the loss in the case of coordinated policy. Upon minimizing Equation 8 subject to Equation 2 and its foreign counterpart, we get $d^{c} = d_{T}$ and $d^{c *} = d_{T}^{*}$ , that is, coordination of antidoping policies eliminates the doping bias. The solution for home athletes’ performance is:

m^{c} = c (d_{T} - b d_{T}^{*}) / (1 - b^{2}) .

The foreign athletes’ performance follows from Equation 9 when d_T and $d_{T}^{*}$ are interchanged.⁸ As expected, sporting performance (and, thus, the prevalence of doping) under international coordination of antidoping policies is smaller than under discretionary national antidoping policies (m^c < m) because an international antidoping agency accounts for the negative spillover effect of discretionary national antidoping policies on the sporting performance of the respective other country.

Discussion

Our simple two-country model illustrates how international coordination of antidoping policies (or lack thereof) can influence sporting performance and the prevalence of doping. In our very stylized model, a lack of international coordination of antidoping policies, without a binding commitment and without punishing even an observable lack of commitment to WADC, results in a doping bias. International coordination of antidoping policies internalizes the negative spillover effect of discretionary national antidoping policies on the sporting performance of the respective other country and eliminates the doping bias. Before this result can be generalized to derive general policy conclusions, however, more research needs to be done.⁹ One could easily imagine extended versions of our model in which the optimality of international coordination of antidoping policies versus discretionary national antidoping policies is more complex than in our simple model.

Our simple model captures the observation that a critical problem standing in the way of achieving WADA’s goal of harmonizing and coordinating the worldwide fight against doping is the lack of compliance on the part of international sports associations. This lack of compliance is not surprising given that, while trying to cooperate in the fight against doping, international sports associations compete against each other both internationally and in the markets for resources and attention. The effectiveness of the WADC, hence, depends largely on whether and if so, to which extent, international sports associations are willing to bring their international and national antidoping regulations into line with the WADC and to act under the rules of this international regulatory system.

To reinforce compliance, stricter implementation requirements were imposed on international sports associations inasmuch as they were asked to include some provisions verbatim in their subordinate regulatory systems, and deviating regulations with respect to minimum and maximum standards are no longer permissible.¹⁰ An institutionalized control and enforcement procedure that allows the international and national implementation and enforcement of the WADC regulations to be monitored, and sanctions in the case of non- or misapplication to be imposed, however, does not yet exist.

An immediate consequence is that international sports associations are still forced to cooperate and to fight doping on a trust basis, albeit with the risk that other sports associations in the same sport fail to cooperate and, in fact, do not effectively fight doping. Under such a condition, where compliance to the rules of the WADC is more or less voluntary, cooperating sports associations which comply to the rules of the WADC are at a disadvantage. In other words, a sports association that enforces doping controls in competition and training reduces the international winning chances of its athletes as it cannot be sure that other sports associations impose equally effective controls.

Given the difficulties with the legal enforcement of global antidoping policies, a fruitful direction for future research is to think about decentralized solutions to the coordination problem that our simple model describes. In this respect, a cross-country approach to antidoping policy immediately comes to mind because it combines elements of a coordination solution and a decentralized solution. Such a cross-country approach would require that antidoping inspectors of the home country are responsible for carrying out antidoping tests in the foreign country, and vice versa. Such a cross-country approach, however, does not solve the international coordination problem that we have described in this research. If, for example, doping inspectors had a diplomatic status and Chinese controllers could control American athletes in the United States (within some time window that makes detecting illicit substances likely), the result would be a secondary competition resulting in an increase in the number of detected dopers of the competitor country. Foreign inspectors even could promote export of doping substances to the respective other country in an attempt to increase the number of detected dopers by making illicit substances easily available.

A more powerful approach to support international antidoping policy is to make the fight against doping more transparent. To this end, WADA could provide information on the number of doping controls that have been conducted in every nation such as to make transparent how often athletes were tested in competition and training. During Olympic Games, commentators, journalists, athletes, and spectators could then see on a board how often athletes had been controlled. Spectators could detect easily a lack of compliance to the standards of the WADC on the side of a sport association, implying that athletes and sports associations that do not participate in the fight against doping run the risk of damaging their reputation. Media, sponsors, and spectators may even boycott athletes that do not adequately support the fight against doping. Transparency, thus, could act as a substitute for a lack of commitment to the WADC.

The idea that transparency may help to mitigate a coordination and commitment problem, of course, is not new. For example, economists have analyzed whether transparency in central banks’ communication policies helps to reduce the kind of inflation bias described in the literature on discretionary monetary policy. Eber (2002) draws on the insights reported in this literature to undertake his analysis. Our argument for improved transparency in international antidoping policies shows that this literature may hold yet unexplored insights that could help to solve the commitment and coordination problem in international antidoping policy.

Footnotes

Acknowledgments

The authors are grateful to two anonymous reviewers for helpful comments.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the financial support from the Bundesinstitut für Sportwissenschaft (project Wettbewerbs—verzerrungen im Sport).

Notes

References

Berentsen

(2002). The economics of doping. European Journal of Political Economy, 18, 109–127.

Bird

E. J.

Wagner

G. G.

(1997). Sport as a common property resource. A solution to the dilemma of doping. Journal of Conflict Resolution, 41, 749–766.

Breivik

(1992). Doping games. A game theoretical exploration of doping. International Review for the Sociology of Sport, 27, 235–255.

Drazen

(2000). Political economy in macroeconomics. Princeton, NJ: Princeton University Press.

Eber

(2002). Credibility and independence of the world anti-doping agency: A Barro-Gordon-type approach to antidoping policy. Journal of Sports Economics, 3, 90–96.

Eber

(2008). The performance-enhancing drug game reconsidered: A fair play approach. Journal of Sports Economics, 9, 318–327.

Eber

Thépot

(1999). Doping in sport and competition design. Recherches Économiques de Louvain/Louvain Economic Review, 65, 435–446.

Gravelle

Rees

(2004). Microeconomics. Harlow, UK: Prentice Hall.

Hamada

(1976). A strategic analysis of monetary interdependence. Journal of Political Economy, 84, 677–700.

10.

Haugen

K. K.

(2004). The performance-enhancing drug game. Journal of Sports Economics, 5, 67–86.

11.

Jakob

Berninger

(2008). Die wichtigsten Änderungen des WADA-codes. SpuRt—Zeitschrift für Sport und Recht, 15, 61–62.

12.

Pierdzioch

(2003). Keeping up with the Joneses: Implications for the welfare effects of monetary policy in open economies. Jahrbuch für Wirtschaftswissenschaften/Review of Economics, 54, 167–177.

13.

Spencer

B. J.

Brandner

J. A.

(1983). International R&D rivalry and industrial strategy. Review of Economic Studies, 50, 707–722.

14.

WADA. (2009). World anti doping code 2009. World anti doping agency (Montreal). Retrieved September 7, 2012, from http://www.wada-ama.org/Documents/World_Anti-Doping_Program/WADP-The-Code/WADA_Anti-Doping_CODE_2009_EN.pdf