Asset specificity in the promotion of elite sports: efficient institutions of governance for the production of long-term future sporting success.

Flatau, Jens ; Emrich, Eike

Introduction

From an economic perspective, the promotion of elite sports can be interpreted in the context of the production of sporting success. Here, we apply an economic approach that takes into account transaction costs of the production of goods, the New Institutional Economics (NIE), to the analysis of the production of elite athletic success. NIE analyzes institutions of governance suitable to optimize the efficiency of transactions (Williamson, 1991, p. 269).

This article applies a central concept from transaction cost economics, the dependency of the optimal governance structure (decentralized or centralized) on certain characteristics of the production process for some good, called asset specificity (Williamson, 1989, p. 142), to a specific stage in the production of national Olympic sporting success, the development of young talented athletes. We compare the medal success of graduates of German elite sport schools (ESS) with the medal success of other athletes. We find that ESS graduates, promoted by a centralized organization, are comparatively successful but that their success depends on the asset specificity of the respective sports. This result has important implications for national policies aimed at increasing Olympic success.

The Production of Elite Sporting Success

Virtually all nations participate in international sporting competitions, especially the Summer and Winter Olympic Games and FIFA World Cup; national sporting success for representation and prestige purposes is desirable in such events. Nations promote domestic elite sports in order to enhance their chances for national success in international competitions, again for the purpose of national representation (e.g., Germany, Deutscher Bundestag, 2010, p. 15). This success has both external (towards the rest of the world) and internal (towards their own population) effects. In the period between the end of the Second World War and the breakdown of the Eastern Bloc, elite athletic success served to legitimate public expenditures on sport. In democratic societies, however, like any nonessential public good, such spending should be supported by a majority of the population or generate sufficient benefits to justify the expenditure. Humphreys, Johnson, Mason, and Whitehead (2011) and Breuer and Hallmann (2011) show that this is true at least for Canada and Germany, in that the national promotion of elite sport meets individual needs, especially concerning the Olympic Games.

Like other products, the inputs to the production of elite sporting performance are labor, land, capital, and know-how. Human capital in the form of talent is also an input. But without training, which takes place at specialized facilities under the guidance of a trainer with a certain competence (human capital), the best talent will not perform as well as possible.

The production of sporting success can be divided into two different temporal stages. Training of elite senior athletes focuses on success in the present or near future, for example in the next relevant competition. For junior athletes, international competitions either do not exist or, if they do, they are paid little attention by the wider public. Thus, only a minimum of recognition of the athletes, and in this way national prestige, is produced. Junior athletes are developed for the future, when they may participate in major international sporting events. Hence, this development can, in economic terms, be regarded as an investment in long-term future sporting success.

Effectiveness and Efficiency in the Production of Elite Sport in Germany

The organization of the promotion of national sport sporting success was very different in East and West Germany. Because of the dissolution of the GDR (German Democratic Republic) and reunification in 1989/1990, Germany contains a unique governance structure of sporting organizations. During reunification, the governance structure of the entire sport system, as well as of each organization responsible for individual sports, had to be merged.

During the Cold War, efficiency was neglected in favor of effectiveness, particularly in the GDR, where the promotion of elite sports was characterized by an extremely high use of resources (Teichler, 1999) in relation to GDP in order to demonstrate the superiority of the ideological system in place. After reunification, former centrally planned economic production was integrated into the Western market economic system and many government-sponsored undertakings were closed down; the promotion of elite sport represents an important exception. A number of hierarchical aspects of the East German elite sport development system were retained in the contemporary athletic development system in reunified Germany.

Several possible explanations for this hybrid system exist. The GDR was exceptionally successful at the Olympic Games. For this reason, West Germany copied structural elements from the centralized East German elite sport development system before reunification, for example, by establishing Olympiastutzpunkte (Olympic Training Centers [OTC]) analogous to the GDR's Sportclubs. Reunification represented a unique opportunity to integrate "original" elements in the German elite sport development system. The existence of many top-class sport development structures in the GDR lead to the psychology of sunk costs (Arkes & Blumer, 1985) where decision makers tend to keep expensive assets "for the sake of the past" (Kelly, 2004, p. 60) regardless of cost. Therefore, many East German sports facilities and laboratories for performance diagnostics representing large one-time investments were not shut down. Another reason was the lack of an authority to evaluate sport development programs based on efficiency. While the final decision on the continuation of East German commercial enterprises was made by a special state department called Treuhandanstalt (trustee agency) based on a strict criteria of efficiency, political-normative considerations played an important role when assessing organizations in the state sector like elite sport development programs.

The centralized sport development organizations analyzed here are Eliteschulen des Sports (elite sport schools [ESS]). Currently, 11,300 student-athletes are educated in 39 ESS, 9,816 of them in the area of the former East Germany, which is a disproportionately high number (Emrich, Frohlich, Klein, & Pitsch, 2009). Their goal is to combine school and sporting education in a way that is comparable with regular schools while still providing time for training and development. ESS are typically located close to sports facilities and often are boarding schools. Generally, every Olympic sport is promoted by the ESS system and single ESS focus on different Olympic sports.

Emrich and Pierdzioch (2011) demonstrated the existence of inefficiencies in contemporary German ESS in the former GDR relative to similar organizations in the west. Our thesis is that the spatial location of the sport production organization (west vs. east) is a proxy for the spill-over effect of different institutions of governance, meaning that a higher degree of vertical integration persists in the former GDR where almost all production processes were formerly characterized by central planning.

Context

NIE-related research in sport is limited to the team sports industry (Downward & Dawson, 2000). Dietl, Franck, Hasan, and Lang (2009) compared cooperative with contractual governance in European team sports leagues and explained the advantages of the former with NIE. Sanderson and Siegfried (2006) reviewed contracting in the baseball labor market and found that market liberalization financially favors players, not team owners. Marburger (2002) examined the effects of player property rights on trades and found that intraconference trade prices were lower in the case of free agency and that this effect was smaller in interconference trades. On the other hand, the possession of own property rights under free agency served as an incentive against shirking (Marburger, 2003). The effect of rule changes on team effort and fan interest in ice hockey was analyzed by Easton and Rockerbie (2005). They concluded that allowing only four skaters in overtime periods and a modified allocation of points in case of a tie lead increased effort and fan interest.

Relatively little research has focused on efficiency in the development of elite sport. Pierdzioch and Emrich (2013) recently showed that corruption is an important factor influencing the impact of population and economic prosperity on elite sport success. They found that, on a national level, less corruption increases the chances of national Olympic success. Corruption in open societies undermines property rights (Shughart & Tollison, 1993; Campbell et al., 2005), athletes' ability access to the media, and resources provided by sports associations, limiting the ability to transform sporting success into financial rewards.

Pierdzioch, Emrich, and Klein (2012) showed that population and economic resources alone cannot explain GDR Olympic medal counts since 1970. The GDR was a highly centralized hierarchical political system; there appears to be a link between hierarchical organization and the efficiency.

The drawback to this approach is that the explanatory variables are measured at a highly aggregated level, leaving room for alternative interpretations. Explanations for Olympic success should go deeper, ideally to the individual level. (1) Another problem is that the explanatory variables are often not under the control of decision makers in sport organizations, reducing their usefulness for understanding the determinants of national sporting success. Hence, the high explained variation in Olympic success reported in the literature cannot be transformed into practical guidance for decision makers.

In a recent survey, Emrich and Gullich (2005) examined the relationship between individual elite sporting success and individual features of athletes' training and identified several key factors related to talent development. They found that greater diversity of sports participation at a young age, a lower scope of training in the primary sport, and relatively low participation in formal development programs in childhood and youth are associated with relatively high sporting success in the long term; in other words both higher and lower individual investment can be more effective, depending on the specific form of investment.

Previous studies on the German ESS were mostly descriptive and focused on school-specific outcomes like reduction in travel-costs (Teubert, Borggrefe, Cachay, & Thiel, 2006), increases in school duration (Evers, 2003), and the importance of compensation in increased physical and psychological stress of ESS-students (Richartz & Brettschneider, 1996).

Only a few studies evaluated the elite sporting success of ESS-students. Beckmann, Elbe, Szymanski, and Ehrlenspiel (2006) compared the sporting success of products of two types of hierarchical organizations in sport schools by comparing boarding school students with day students and found no significant differences in sporting success. Emrich et al. (2009) found a substantial difference in the success achieved by boarding students (49% of eventual medal winners) and day students (19%). Former ESS students were also compared to non-ESS students in terms of their medal success in Olympic Games. ESS students were more successful in Olympic Winter Games than competitors whose talent had exclusively been developed in clubs, but this did not hold true for Olympic Summer Games.

Emrich and Pierdzioch (2011) compared East and West German elite students and determined that the former won slightly more medals than the latter; the production efficiency of the elite schools, measured by medal prevalence of athletes trained, however, was substantially higher in the old federal states owing to a much lower student prevalence in the Olympics.

Flatau and Emrich (2011) is the only study to compare ESS students and non-ESS students to examine the effectiveness of centralized governance structures relative to that of decentralized structures in the production of long-term future sporting success by the site and physical asset specificity of the production. Flatau and Emrich reported a significant interaction effect of (former) ESS-membership and both site and physical asset specificity. They also found that Germany's national sporting success is generally greater in sports with high asset specificity, meaning sports like rowing or bobsled.

The present study is the first that tests a model that includes several explanatory variables measured at the individual and organizational level. In addition, it is informed by economic theory and compares the effectiveness of centralized and decentralized governance structures depending on asset specificity.

Transaction Cost Economics and Production of Elite Sporting Success

Unlike neoclassical economics, the NIE takes into account that, due to information asymmetries, transactions in a perfectly competitive market do not necessarily lead to collectively efficient results (Voss & Abraham, 2000). The correction of these inefficiencies by the institutions that can prevent opportunistic behavior (for example, moral hazard) implies that transaction costs must be weighed against the losses incurred in uncorrected transactions.

Klein, Crawford, and Alchian (1978) and Williamson (1985) argue that there are other sources of inefficient outcomes in competitive markets related to the production process, called asset specificity. Here, transaction costs are responsible for market inefficiencies (Eggertson, 1990). One way to avoid these transaction costs is through vertical integration of production. Again, these efficiency advantages have to be weighed against the disadvantages arising from misallocation of resources. If the former predominate, this answers Coase's (1937) classical question for the existence of firms (i.e., centralized units) in generally decentralized market economies: They result from the market itself since efficiency is its exclusive selection criterion (see Williamson, 1971; Arrow, 1974).

Williamson (1989) identified five kinds of asset specificity: (2)

* Site specificity: dependence on resources only available at a certain location and transportable only at major cost;

* Physical asset specificity: specialized and/or expensive machines designed for a single purpose;

* Human asset specificity specialized human skills, arising only by a specialized and time-consuming education or by extensive experience;

* Dedicated assets: a discrete investment in a plant to produce a customer-specific good that cannot readily be put to work for other purposes;

* Brand name specificity investment in the reputation of certain products.

Specific assets are investments that "lose value if employed in alternative uses" (Williamson, 1983, p. 526). This creates a risk of sunk costs, depending on the degree of uncertainty of the transaction stability and the size of the investment. Thus, vertical integration is more favorable the higher the asset specificity (Whyte, 1994).

Applying Asset Specificity to Sports

Since the promotion of elite sports can be regarded as a production process, we assume that asset specificity affects efficiency. We first discuss which kinds of asset specificity exist in this production process and then identify analogies to Williamson's (1975) ideal type categories of governance structure.

Sporting success at the Olympic Games is the sum of success in a number of different sports across which asset specificity varies substantially. A high level of site specificity is important if elite training in a sport requires certain site prerequisites. For example, multi-use sports halls and football pitches can be constructed just about anywhere, implying that the site specificity of football and handball is low. Alpine skiing requires high, steep mountains, low winter temperatures, and snow. Some countries have an abundance of these conditions (for example, Austria), implying low site specificity, while others have few (for example, Germany), implying high site specificity; many countries have no ideal sites for Alpine skiing.

The production of sporting success also requires human capital (athletic talent and training diligence, as well as coaching ability) and training conditions: both must be present. Together with the expenses for the required sporting facilities, these factors determine production costs. Since human capital can be assumed to be equally distributed across a country, the mean distance between high human capital and an adequate training site is high in case of high site specificity. In a decentralized system, the talented athletes have to travel (on average) relatively long distances for each training session. (3) These transaction costs reduce the efficiency of a decentralized system and, at a certain point, exceed the costs for centralized organization like boarding schools.

The degree of physical asset specificity varies across sports. Elite training in some sports requires physical assets in the form of cost-intensive facilities and/or equipment while other sports do not. Long distance runners, for example, can train almost anywhere and require little equipment. Bob sleds and bob sled runs are very expensive. But not every sport that requires cost-intensive prerequisites has high physical asset specificity, since specificity is also characterized by the assets' purpose. Thus, the selectness of the development process must also be considered. Only when cost-intensity and selectness come together does the production of success in certain kinds of sports exhibit high physical asset specificity. This applies to the bob sled example since the runs and sleds are exclusively built for elite athletes' training in bob sled, toboggan, and skeleton. Recreational sliders cannot use them for leisure purposes. Similar examples are artificial white water channels and ski jumps. Swimming pools, which are expensive to build and maintain, are not exclusively used by elite athletes, so swimming has low asset specificity. The same applies to running tracks that are used by a substantial number of recreational runners. And there are also sports facilities with no alternative use. Asphalt roads, indispensable for road cyclists, are a public good constructed primarily for traffic, but velodromes are only used by elite cyclists and have high asset specificity.

Facilities for developing athletes in a sport with high physical asset specificity are relatively rare. This leads to an assumption analogous to sports with high site specificity: Since the average travel (transaction) costs for athletes and coaches (human capital inputs) are high in a decentralized elite sports training system, centralized training may have cost advantages.

Empirical Analysis

Research on the production of sporting success using aggregate level (social/national) data is advanced, but knowledge about explanatory factors at the level of institutions and individuals is still scarce. The size of a country's population and financial resources are crucial determinants of sporting success, but decision makers cannot be sure what to do with these resources (for example how to spot and develop talented young athletes, which can be influenced by policy makers). Various disciplines of sport science have contributed theories and empirical evidence at disaggregated levels, but so far the empirical results in this literature have been ambiguous and contradictory.

By developing and estimating an empirical model explaining individual sporting success that contains proxies for characteristics of institutional governance (decentralized vs. centralized) and on the respective kind of sports in which sporting success is to be produced (asset specificity), we develop new evidence that can be used by decision makers. Independent variables in the empirical model at the individual level:

* Age at start of training: The theory of deliberate practice (Ericsson, Krampe, & Tesch-Romer, 1993) predicts that an early beginning of training enhances the chances of later success since every training adds to an athlete's abilities.

* Number of additional sports played: Emrich and Gullich (2005) posit that every kind of sport participated in when young raises the chance of a functional matching between athlete and sport.

* Gender: Due to a relatively high degree of women's freedom in German society, we expect greater sporting success by females.

Independent variables at the institutional level:

* ESS attendance: These institutions aim to save time that can be used to extend training time and also provide young athletes with a high quality of training through good sports facilities and qualified trainers. Thus, a positive impact of ESS membership on later sporting success is expected.

* Site specificity: Germany has a comparatively wide geographical and climatic diversity (e.g., the Alps, North and Baltic Sea coasts, cold winters). Since the dependence on high mountains, low temperatures, and various bodies of water are the reasons for high asset specificity of sports, it is presumed to be a factor of (German) sporting success.

* Physical asset specificity: Due to a relatively high level of prosperity, Germany can afford to build expensive sports facilities exclusively for the purpose of producing (future) sporting success.

Sports clubs in highly site-specific sports are generally less evenly distributed across Germany. This mitigates production advantages from superior information (talent spotting) and resource allocation (talent development). Thus, we assume that in sports with high site specificity, the effectiveness advantage of hierarchical organizations over decentralized production in sports clubs is larger (its disadvantage is smaller respectively).

Physical asset specificity also reduces the contribution made by sports clubs in the production of national sporting success. This is because of the low financial resources available to most sports clubs as nonprofit organizations prevents them from building training facilities. Thus, we assume that in sports with high physical asset specificity, the advantage of hierarchical organizations over its decentralized production in sports clubs in the production of national sporting success is larger.

Based on Williamson's (1991) study of comparative economic organization, we compare the effectiveness of decentralized and centralized organizations, represented by sports clubs and ESS, in the production of sporting success. Data on costs do not exist, but a comparison of organization types under asset specificities can indirectly identify effectiveness and efficiency.

We analyze the data used by Flatau and Emrich (2011) and expand and modify their analysis. The data come from a 2006 survey of German participants in the previous Olympic Games (Summer 2004 or Winter 2006). Of the 611 athletes who were contacted, 199 from 40 sports responded. Surveys from 27 athletes from team sports were excluded from the analysis. Twenty-two athletes did not provide information on their sport. The remaining sample of 148 athletes on which the following analyses are based is representative of the overall body of Olympic participants with regard to relevant characteristics such as gender, age, and region (east/west).

Unlike Flatau and Emrich (2011), we quantify ESS membership as an interval-scaled variable reflecting duration of attendance in years with a value of zero for non-ESS students. Physical asset and site specificity variables are dichotomous: high vs. low. For classification into the high and low categories, 20 sports scientists and economists from various German universities were interviewed for expert evaluations. The final classification was made on the basis of majority decisions. Table 1 shows the classification of asset specificity.

Sporting success becomes most apparent through medal awards. The medal table ranks participating nations in the order of their medal success providing a simple overview of the nations' sport-related standing in the world. That is why we measure national sporting success at the individual level dichotomously (medal/no medal). As this study examines determinants of future sporting success, we only examine young athletes. Thus, we calculated junior training age as the time from the beginning of systematic training (at least twice a week on average) until the age of 18. We estimate the parameters of a probit model where the dependent variable is equal to one if an individual athlete won an Olympic medal and zero otherwise.

Results

The parameter estimates, estimated standard errors, Z-statistics for the null hypothesis that the parameter estimate is equal to zero, and associated P-values for the probit model are shown on Table 2.

Among six determinants of future sporting success, ESS attendance proves to empirically have a significant positive effect on later Olympic sporting success. The effects of systematic training at young age, gender, and the two forms of asset specificity do not explain observed variation in medal success. The model's overall explanatory power is moderate ([R.sup.2] Nagdkerke = 0.16).

To test the effectiveness of ESS attendance under different levels of asset specificity, we estimated two alternative models. First, we tested the interaction effects of the respective variables using a probit model. These results are displayed in Table 3.

The results support our hypothesis about asset specificity and effectiveness of hierarchically organized production of long-term future sporting success. Under low site specificity and physical asset specificity, respectively, the positive effect of the duration of ESS attendance is not statistically significant. When site specificity and physical asset specificity is high, ESS attendance increases the likelihood of later Olympic success significantly. More time spent in ESS attendance when young is only associated with later Olympic success when the sport exhibits either high site specificity or high physical asset specificity; ESS attendance is not associated with improved elite athletic performance in all sports.

Discussion and Conclusions

We developed an empirical model containing explanatory variables measured at the micro- and meso-levels--focusing on the young athletic talent development--rooted in contemporary theoretical approaches in social science. The results from this model are easier to convert to applicable knowledge for the promotion of elite sport. Although almost all explanatory variables contribute to the development of sporting success, the effects are more or less small and thus, in all cases but one, not statistically significant. The only significant independent variable is ESS attendance, but this must be interpreted carefully. We attributed the positive impact of ESS attendance on sporting success to efficient time use and other related factors. Since sporting success when young may not be a sufficiently valid criterion for talent identification and thus for the potential for long-term future sporting success, selection bias may influence the results.

The explanatory variables reflecting comparative advantage have no substantial association with long-term future sporting success. The reason for this might be that Olympic medals are scarce compared to the overall number of competitors. Although Germany might have a comparative advantage in form of a relatively high rate of women's participation in elite sports, different geographical and climatic conditions as well as financial power, this might also apply to the rest of the developed world. Many developed countries have a high participation rate of women in elite sports and even more favorable geographical prerequisites. Since in this study our success criterion was medal success, a few equal or stronger competitors are sufficient to cause nonsuccess. Our results moderately support theories regarding the promotion of junior elite sports that suggest an early start of training and exclusiveness.

The application of transaction cost theory appears promising in this case, particularly in the case of the effect of site specificity on sporting success. The sporting success of ESS students relative to non-ESS students is much higher in sports based on rare environmental conditions. However, this effect is weaker for physical asset specificity. The reason for this difference might be that geographical and climatic conditions are unchangeable, whereas means of production are linked to financial resources. This does not apply to all individual sports (e.g., road cycling) and the commercialization of certain individual sports has increased in the recent past, too. Thus, through private sponsorship, at least some clubs might have the financial means to be able to afford cost-intensive sports facilities exclusively built for elite sports.

The small difference in institutional effectiveness might also partly be attributable to track and field athletics, the site specificity of which is low and the physical asset specificity of which has--not unambiguously--been classified as high. Track and field is the Olympic sport with the largest number of disciplines and therefore the highest number of German participants in the Olympic Games, but the medal success of German track and field athletes was poor in 2004 (only two medals). This poor sporting success, in terms of the medals-participant ratio, is attributed mainly to high physical asset specificity in our empirical analysis, perhaps affecting the results.

This study could be extended along several dimensions. Issues associated with the classification of sports to discrete categories discussed above indicate more attention must be paid to identification of asset specificity in future research. In the case of physical asset specificity, this could be done by analyzing costs and the degree of alternative use of sports facilities required for each sport. Site specificity could be quantified by the proportion of area/time in which the prerequisites for a certain kind of sports are available per total area/year in a country.

Since the human capital of young athletes, as well as coaching ability and qualifications, are important inputs to the production of long-term future sporting success, this analysis could be extended to include human capital asset specificity. Since athletic talent is scarce, exploitation of the available athletic talent through scouting is crucial for future sporting success. An extensive state athletic talent spotting program was one of the factors explaining the GDRs Olympic success. Since the government cannot employ such policies in a free society, talented athletes have to be spotted by clubs with voluntary membership. Success in this is higher the higher the organization density, in terms of sports clubs per number of inhabitants, which again is connected to the degree of organization in a sport; the popularity of participation in specific sports should be considered as a covariate in future research.

Furthermore, the degree of commercialization should be considered in future research. Top-level clubs in popular spectator sports are in a position to generate significant financial resources. This reduces the requirement of youth sport promotion by the state even in sports with high physical asset specificity because many of those sports clubs invest large sums in talent development.

References

Arkes, H. R., & Blumer, C. (1985). The psychology of sunk costs. Organizational Behavior and Human Decision Processes, 35, 124-140.

Arrow, K. J. (1974). The limits of organization. New York, NY: Norton & Co.

Beckmann, J., Elbe, A.-M., Szymanski, B., & Ehrlenspiel, F. (2006). Chancen und Risiken: Vom Leben im Verbundsystem von Schule und Leistungssport. Cologne, Germany: Straufi.

Breuer, C., & Hallmann, K. (2011). Die gesellschaftliche Relevanz des Spitzensports in Deutschland. Koln, Deutschland: Straufi.

Campbell, L. M., Mixon, F. G., & Sawyer, W. C. (2005). Property rights and Olympic success: An extension. Atlantic Economic Journal, 33, 243-244.

Coase, R. H. (1937). The nature of the firm. Economica, 4, 386-405.

Deutscher Bundestag (2010). 12. Sportbericht der Bundesregierung. Bundestagsdrucksache 17/2880. Cologne, Germany: Bundesanzeiger Verlagsgesellschaft.

Dietl, H. M., Franck, E., Hasan, T., & Lang, M. (2009). Governance of professional sports leagues--Cooperatives versus contracts. International Review of Law and Economics, 29, 127-137.

Downward, P., & Dawson, A. (2000). The economics of professional team sports. London, England; New York, NY: Routledge.

Easton, S. T., & Rockerbie, D. W. (2005). Overtime! Rules and incentives in the National Hockey League. Journal of Sports Economics, 6, 178-202.

Eggertson, T. (1990). Economic behavior and institutions. Cambridge, MA: Cambridge University Press.

Emrich, E., Frohlich, M., Klein, M., & Pitsch, W. (2009). Evaluation of the elite schools of sport--Empirical findings from an individual and collective point of view. International Review for the Sociology of Sport, 44, 151-171.

Emrich, E., & Gullich, A. (2005). Zur "Produktion" sportlichen Erfolges. Organisationsstrukturen, Forderbedingungen und Planungsannahmen in kritischer Analyse. Cologne, Deutschland: Straufi.

Emrich, E., & Pierdzioch, C. (2011). Konvergenz und Efizienz? Zur Ubernahme von Elementen des DDR-Sports in den bundesdeutschen Spitzensport. In K. Bohr & A. Krause (Eds.), 20 Jahre Deutsche Einheit. Bilanz und Perspektiven (pp. 105-122). Baden-Baden: Nomos.

Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100, 363-406.

Evers, J. (2003) Das Landessportgymnasium Leipzig zwischen Vision und Realitat. In K. Rost (Ed.), Schule und Sport - gemeinsam zur Spitze (pp. 55-66). Frankfurt, Germany: DSB.

Flatau, J., & Emrich, E. (2011). Die Organisation sportlichen Erfolges Zur Frage nach Markt oder Hierarchie im Spitzensport am Beispiel der Eliteschulen des Sports. Sportwissenschaft, 41, 100-111.

Humphreys, B. R., Johnson, B. K., Mason, D. S., & Whitehead, J. C. (2011). Estimating the value of medal success at the 2010 Winter Olympic Games (Working Paper No. 2011-20). Edmonton, AB, Canada: University of Alberta.

Kelly, T. (2004). Sunk costs, rationality, and acting for the sake of the past. Nous, 38, 60-85. Klein, B., Crawford, R. G., & Alchian, A. A. (1978). Vertical integration, appropriable rents, and the competitive contracting process. Journal of Law and Economics, 21, 297-326.

Malone, T. W., Yates, J., & Benjamin, R. I. (1987). Electronic markets and electronic hierarchies. Communications of the ACM, 30, 484-497.

Marburger, D. R. (2002). Property rights and unilateral player transfers in a multiconference sports league. Journal of Sports Economics, 3, 122-132.

Marburger, D. R. (2003). Does the assignment of property rights encourage or discourage shirking? Journal of Sports Economics, 4, 19-34.

Pierdzioch, C., & Emrich, E. (2013). A note on corruption and national Olympic success. Atlantic Economic Journal. Published online March 7th, 2013.

Pierdzioch, C., Emrich, E., & Klein, M. (2012). Die optimierende Diktatur--Politische Stabilisierung durch staatlich verordnetes Doping am Beispiel der DDR (Working Paper Series No. 123). Hamburg: Helmut Schmidt Universitat, Department of Economics.

Richartz, A., & Brettschneider, W.-D. (1996). Weltmeister werden und die Schule schaffen. Zur Doppelbelastung von Schule und Leistungstraining. Schorndorf: Hofmann.

Sanderson, A.R., & Siegfried, J.J. (2006). Simon Rottenberg and baseball, then and now: A 50th anniversary retrospective (Working Paper No. 06-W06). Nashville, TN: Vanderbilt University.

Shughart, W. F., & Tollison, R. D. (1993). Going for the gold: Property rights and athletic effort in transitional economies. Kyklos, 46, 262-272.

Teichler, H. J. (1999). Die Gesamtkosten des Leistungssports. In H. J. Teichler & K. Reinartz (Eds.), Das Leistungssportsystem der DDR in den 80er Jahren und im Prozefi der Wende (pp. 88-93). Schorndorf, Germany: Hofmann.

Teubert, H., Borggrefe, C., Cachay, K., & Thiel, A. (2006). Spitzensport und Schule. Moglichkeiten und Grenzen struktureller Kopplung in der Nachwuchsforderung. Schorndorf, Germany: Hofmann.

Voss, T., & Abraham, M. (2000). Rational choice theory in sociology: A survey. In S. R. Quah & A. Sales (Eds.), International Handbook of Sociology (pp. 50-83). London, England: Sage.

Whyte, G. (1994). The role of asset specificity in the vertical integration decision. Journal of Economic Behavior & Organization, 23, 287-302.

Williamson, O. E. (1971). The vertical integration of production: Market failure considerations. American Economic Review, 61, 112-123.

Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust implications. New York, NY: The Free Press.

Williamson, O. E. (1983). Credible commitments: Using hostages to support exchange. American Economic Review, 73, 519-540.

Williamson, O. E. (1985). The economic institutions of capitalism. London, England: Free Press.

Williamson, O. E. (1989). Transaction cost economics. In R. Schmalensee & R. D. Willig (Eds.), Handbook of Industrial Organization (pp. 135-182). Amsterdam, Netherlands: North Holland.

Williamson, O. E. (1991). Comparative economic organization. The analysis of discrete structural alternatives. Administrative Science Quarterly, 36, 269-296.

Zaheer, A., & Venkatraman, N. (1994). Determinants of electronic integration in the insurance industry: An empirical test. Management Science, 40, 549-566.

Jens Flatau (1) and Eike Emrich (2)

[1] Christian-Albrechts University of Kiel

[2] University of the Saarland

Endnotes

(1) E.g.: What exactly are the effects that constitute the host advantage? A greater motivation of the host country's athletes? A relatively higher amount of public money spent for the promotion of elite sports?

(2) In the meanwhile, other authors have suggested further sorts of asset specificity (e.g., Malone et al., 1987; Zaheer & Venkatraman, 1994) that are not relevant in connection with this article.

(3) This is a hypothetical consideration. In fact, hardly any clubs exist far from the training sites required for a sport.

(4) Concerning this, we also differ from Flatau and Emrich (2011) where a trichotomous and a continuous success parameter were used.

Jens Flatau is professor of economics and sociology of sport at the Christian Albrechts University of Kiel. His research interests include sports organizations, voluntary engagement, promotion of elite sport, and research methods.

Eike Emrich is professor of sports economics and sociology of sports at the Saarland University. His research interests include sports economics, political economics, sociology of sports, and evaluation research.

Table 1: Allocation of Sports to Categories of Asset
Specificity (Flatau & Emrich, 2011)

                    Physical asset specificity

High (n=80)                 Low (n=68)
Biathlon                     Badminton
Bobsled                  Beach volleyball
Fencing                       Boxing
Canoe                         Curling
Athletics (a)             Figure skating
Heptathlon                 Speed skating
Nordic combined           Weight lifting
Equestrian                     Judo
Toboggan                    Gymnastics
Rowing                        Cycling
Sailing                      Wrestling

Skeleton               Rhythmical gymnastics
Surfing                      Shooting
Trampoline                   Swimming
Short track
Alpine skiing
Cross-country skiing
Snowboard
Triathlon
Diving

                        Site specificity

High (n=66)                 Low (n=82)
Biathlon                     Badminton
Bobsled                  Beach volleyball
Canoe                         Boxing
Nordic combined               Curling
Toboggan                  Figure skating
Rowing                     Speed skating
Sailing                       Fencing
Skeleton                  Weight lifting
Alpine skiing                  Judo
Cross-country skiing        Gymnastics
Snowboard                    Athletics

Surfing                     Heptathlon
                              Cycling
                            Equestrian
Wrestling
Rhythmical
  gymnastics
Shooting
Swimming
Short track
Trampoline
Triathlon                     Diving

(a) The problem of the assignment of many sports, some of
which comprise very different disciplines, becomes obvious
using athletics as an example. Here practically no capital
is required in order to produce sporting success in long-distance
running, as shown impressively by the bare-footed marathon
Olympic champion Bikila Abebe. However athletics was assigned
to the highly capital-intensive sports overall because of
the relatively high costs for sporting facilities (not least
for a track). Unfortunately, information on the disciplines
within the sports is not provided by the survey data.

Table 2: Probit Model of Future Sporting Success (n=95)

Variable                                          Estimate   Standard
                                                              error

Intercept                                          -1.567     0.417
Systematic training at junior age (in years)       0.051      0.043
Number different sports done at junior age         -0.171     0.240
Gender (0 = male; 1 = female)                      0.371      0.259
Duration attendance at ESS (in years)              0.078      0.037
Site specificity (0 = low; 1 = high)               0.339      0.362
Physical asset specificity (0 = low; 1 = high)     0.221      0.357

Variable                                            Z        P

Intercept                                         -3.756   <0.001
Systematic training at junior age (in years)      1.170    n. s.
Number different sports done at junior age        -0.711   n. s.
Gender (0 = male; 1 = female)                     1.432    n. s.
Duration attendance at ESS (in years)             2.100    <0.05
Site specificity (0 = low; 1 = high)              0.934    n. s.
Physical asset specificity (0 = low; 1 = high)    0.621    n. s.

Table 3: Interaction Effects Between ESS Attendance and
Asset Specificity on Olympic Medal Success (n=95)

Interaction                           Estimate   Standard
                                                  error

Site specificity (low) x duration      0.026      0.044
  Attendance at ESS (in years)
Site specificity (high) x duration     0.147      0.041
  Attendance at ESS (in years)
Physical asset specificity (low)       0.057      0.039
  x Duration attendance at ESS
  (in years)
Physical asset specificity (high)      0.135      0.044
  x Duration attendance at ESS
  (in years)

Interaction                             Z       P

Site specificity (low) x duration     0.591   n. s.
  Attendance at ESS (in years)
Site specificity (high) x duration    3.585   <0.001
  Attendance at ESS (in years)
Physical asset specificity (low)      1.462   n. s.
  x Duration attendance at ESS
  (in years)
Physical asset specificity (high)     3.068   <0.01
  x Duration attendance at ESS
  (in years)