National Culture and Home Advantage in Football

Abstract

This article examines home advantage (HA) in association football (soccer). HA is the tendency for teams to perform better when playing on their home ground than when playing away. National variations in HA are found to be related to national cultural and social characteristics. HA tends to be elevated in countries with high levels of collectivism and in-group favoritism, and in countries where governance is prone to corruption and where the rule of law is not strictly adhered to. These findings are consistent with the concept of HA as a social phenomenon that derives from the influence of spectators on the match officials. HA is also found to be elevated in countries with diverse terrain, but the effects of culture persist even when diversity of terrain is controlled for. On the other hand, the hypothesis that HA is elevated in the presence of large crowds or potentially violent spectators was not supported.

Keywords

sports national culture football soccer home advantage

Introduction

Home advantage (HA) is the tendency for sports teams to enjoy better results when playing at home than when visiting their opponents or playing at a neutral venue. HA is a robust phenomenon found in many sports, including association football (Pollard, 2006), American football (Jehue, Street, & Huizenga, 1993), ice-hockey and basketball (B. Schwartz & Barsky, 1977), rugby union (Thomas, Reeves, & Bell, 2008), Australian Rules football (Clarke, 2005), and cricket (Morley & Thomas, 2005). Jamieson (2010) conducted a meta-analysis of 10 sports, including 3 individual sports (boxing, tennis, and golf), and four eras (pre-1950, 1951-1970, 1971-1990, and 1991-2007). A significant HA was observed in all sports and across all eras; on average, the home team or home competitor won 60% of their contests. The extent of HA was however moderated by sport and era; association football showed the largest effect and baseball the weakest, and HA was higher before 1950 than in subsequent eras. Similarly, Gomez, Pollard, and Luis-Pascual (2011) examined nine different team sports played in Spain, and found varying degrees of HA in all of them, although the rank ordering of sports differed somewhat from that found by Jamieson.

In this article, we examine HA in association football (soccer). Football¹ is recognized as the world’s most popular sport. Two hundred and seven national associations are federated to football’s international governing body, the Fédération Internationale de Football Association (FIFA). FIFA estimate that approximately 265 million male and female players (of which 38 million are registered) and 5 million referees and officials are actively involved in the game of football (FIFA Communications Division, Information Services, 2007).

The sources of HA have been the subject of considerable research, and proposed explanations have included crowd effects, travel effects, familiarity, referee bias, territoriality, tactics, and psychological factors. Reviews by Carron, Loughhead, and Bray (2005), Nevill and Holder (1999), and Pollard (2008) may be consulted for further detail.

One antecedent of HA that has attracted the attention of researchers is the influence of the home spectators. Nevill and Holder (1999) suggested that the crowd is able to either raise the performance of home competitors, or subconsciously influence the officials to favor the home team.

It is the latter conjecture that concerns us here, and there is substantial evidence for both the existence of a referee bias and its origins in crowd behavior. First, several studies have confirmed the existence of a bias in favor of home teams among football referees. One manifestation of this bias is the tendency of referees (in both Spain and Germany) to add significantly more “extra time” at the end of matches when the home team is behind by one goal than when it is ahead by one goal or level, presumably giving the home team more opportunity to turn around a losing situation (Garicano, Palacios-Huerta, & Prendergast, 2005; Sutter & Kocher, 2004). Supporting the notion of home-team bias, Dawson, Dobson, Goddard, and Wilson (2007) found that after controlling for differences in team behavior, referees in the English Premier League tend to issue more disciplinary sanctions against away teams than against home teams.

Second, various lines of evidence are consistent with the idea that home-team bias is generated by pressure from the home supporters. The famous line-judgment experiments of Asch (1951) demonstrated the power of groups to coerce the perceptual judgments of individuals, or at least to modify the way they were reported, in the direction of conformity with the opinion of the majority. In the context of a football match, official decisions to call a foul, issue a disciplinary sanction, or to declare a technical infringement such as an offside are frequently marginal, so the reaction of the crowd could be expected to influence the decision-maker. In fact, there is direct experimental evidence that referees’ judgments are influenced by the behavior of the spectators. Thus in a laboratory setting, Nevill, Balmer, and Mark Williams (2002) found that football referees judging video clips of infringements awarded fewer fouls to the home team when the crowd noise was played than when the clips were viewed in silence. Unkelbach and Memmert (2010) conducted a parallel study in which they manipulated the volume of the crowd noise. They found that referees awarded more yellow cards (disciplinary sanctions) against the away team in the high volume condition than in the low volume condition. Other evidence is indirect. For example, Unkelbach and Memmert found that in real matches, the difference between yellow cards awarded to the away team and the home team increased with crowd density, a finding that is consistent with the crowd influencing match officials in favor of the home team. Furthermore, Nevill, Webb, and Watts (2013) found substantial declines in HA in English and Scottish football since 1945, corresponding to significant changes in the training of referees, but the declines were less pronounced in the higher divisions with larger crowds. The authors concluded that although training had improved referees’ resilience to crowd pressure, large crowds still exerted an influence on their judgments.

HA has been observed both at the between-country level, (i.e., a national team performs better in its own country than abroad) and the within-country level (i.e., a team performs better in its own stadium than in the opposition stadium). This article studies within-country HA, partly because the socio-cultural effects I wish to explore are simpler to analyze when the teams, officials, and spectators all belong to the same culture, and partly because relevant data are readily available.

An interesting feature of within-country HA in football is that it displays substantial national variation. Pollard (2006) examined HA in the first-tier domestic football leagues of 72 countries over a period of six seasons. Quantifying HA by the percentage of total points gained at home, Pollard found that national HA ranged between 49% and 79%, with an overall mean of 61%, (SD = ±6%). HA was particularly high in the Balkans and the Andean nations of South America, which Pollard suggested was due to high levels of territoriality among the subpopulations inhabiting these regions. The idea that differences in HA are rooted in cultural differences is certainly credible. Of course, social and cultural differences are not the only possible cause of differential levels of HA between countries. HA is also high in countries where conditions (such as altitude) vary widely between stadia, and where away teams thus perform in unfamiliar conditions.

Nevertheless, to the extent that HA is a social phenomenon, dependent on interactions between the crowd and the players and officials on the field, we might well expect national differences in social behaviors and cultural values to influence the extent of HA in predictable ways. Until now, no quantitative studies of this conjecture have been conducted, and this article is the first to do so.

The rest of this article is organized as follows: The next section sets out the hypotheses, and is followed by a description of the measures used. I then report the analysis and results, and discuss the implications.

Hypotheses

HA and Individualism–Collectivism

Individualism–Collectivism is a bipolar dimension of national culture and measures the extent to which group identity and cohesion are practiced and valued in a society. According to Hofstede (2001), in individualistic societies, “ . . . the ties between individuals are loose . . . ” and everyone is expected to fend for themselves and their immediate families; conversely, in collectivist societies, people are “integrated into strong cohesive in-groups which . . . protect them in exchange for unquestioning loyalty.” (p. 225). There are perhaps two reasons why we might expect HA to be elevated in collectivist countries. First, individuals in collectivist countries tend to be more conformist. For example, Bond and Smith (1996) found that the level of group conformity in judging the lengths of lines was higher in collectivist countries than in individualist countries. This tendency to conformity suggests that in collectivist countries, referees’ judgments will be particularly susceptible to the influence of the majority (i.e., home) crowd. Second, as argued by Triandis, Bontempo, Villareal, Asai, and Lucca (1988), the divisions between in-groups and out-groups in collectivist cultures are often sharper than in individualistic cultures:

In collectivist cultures people share and show harmony within ingroups, but the total society may be characterized by much disharmony and nonsharing . . . In contrast, in individualistic cultures people define the ingroup . . . as “people who are like me in social class, race, beliefs, attitudes, and values.” (p. 326)

This suggests that home spectators in collectivist cultures may express more animosity toward visiting (out-group) teams, leading to intimidation of their opponents and indirectly exerting pressure on match officials to favor the home team. Thus, the first hypothesis is as follows:

Hypothesis 1 (H1): HA will be elevated in collectivist countries and depressed in individualist countries.

HA and Official Integrity

Referee bias involves a willingness to “bend the rules” to benefit a powerful constituency (in this case the supporters of the home team). Such behavior is presumably more acceptable and more common in societies where laws are not always strictly applied and officials are open to corruption. This leads to the following hypothesis:

Hypothesis 2 (H2): HA will be elevated in countries where the integrity of officials is low.

HA and In-Group Favoritism

Because home-team supporters outnumber away-team supporters, the home team can be considered an in-group and the visitors an out-group. In countries with a strong tendency toward in-group favoritism, match officials might therefore be expected to favor the home team in preference to the away team. This suggests the third hypothesis:

Hypothesis 3 (H3): HA will be elevated in countries with a culture of in-group favoritism.

HA and Intimidation

Match officials (and away-team players) are more likely to feel intimidated when a crowd is aggressive or threatening, and as home-team supporters generally outnumber away-team supporters by a considerable margin, it may be supposed that match officials would be induced to favor the home team in such situations. Intimidation is also likely to be stronger in front of larger crowds. This leads to the next hypothesis:

Hypothesis 4 (H4): HA will be elevated in violent societies and in front of large crowds.

HA and Terrain

In a study of international football matches, McSharry (2007) found that teams such as Brazil that are used to playing at low altitudes underperform when playing away against high-altitude teams such as Bolivia or Columbia. Low-altitude teams find it difficult to adapt to high altitudes, while high-altitude teams have no difficulty playing at low altitudes. Thus high-altitude teams score more goals and concede fewer with increasing altitude difference; each 1,000 m of altitude difference between the teams confers an advantage of about half a goal to the high-altitude team. This suggests that domestic HA might be higher in countries with extremes of altitude than in countries where the terrain is relatively uniform. Diversity of terrain is included in the analysis as a control variable, and the fifth hypothesis is

Hypothesis 5 (H5): HA will be elevated in countries where the altitude varies widely.

Method

Measure of HA

National measures of HA were taken from Pollard (2006). In a FIFA domestic football match, the winning team is awarded three points, and the losing team no points; if the match is drawn, both teams are awarded one point. Pollard calculated HA as the number of points won by teams playing at home divided by their total points both home and away. Pollard’s data spanned six complete seasons prior to 1st January 2004 in 51 European nations (with England, Wales, Northern Ireland, and Scotland treated as individual nations), 10 South American nations, and 11 other nations including the United States, 3 Far Eastern nations, and 2 Middle Eastern nations.

Measures of Individualism–Collectivism

National scores for Individualism for were taken from Hofstede (2001). Hofstede’s data set was developed during the 1970s and does not include any countries of the former Soviet Union. To increase national coverage, a Collectivism scale was constructed from five items in the combined World Values–European Values Survey (European and World Values Surveys Four-Wave Integrated Data File, 1981-2004, 2006). The scale items measured respectively the extent to which respondents felt (a) concern for fellow countrymen, (b) work is a duty toward society, (c) equality is more important than freedom, (d) parents must always be loved and respected, and (e) that being useful for society is important in a job. The scale had a reliability of .80, well above the recommended minimum of .70 for research scales, and correlated strongly with Hofstede’s Individualism scale in the expected direction (r = −.65, p < .001, n = 35).

Measures of Official Integrity

Official integrity was measured by two indicators of the perceived quality of national governance, taken from Kaufmann, Kraay, and Mastruzzi (2003). The first indicator, rule of law, measures the extent to which agents have confidence in and abide by the rules of society, and includes “ . . . perceptions of the incidence of crime, the effectiveness and predictability of the judiciary, and the enforceability of contracts.” The second indicator, control of corruption, measures perceptions of corruption, defined as the exercise of public power for private gain, and includes

. . . the frequency of “additional payments to get things done” . . . the effects of corruption on the business environment . . . “grand corruption” in the political arena . . . the tendency of elite forms to engage in state capture. (Kaufmann et al., 2003, p. 4)

Together, these indicators measure the degree to which fair and predictable rules form the basis for economic and social interactions in a society. The national scores for rule of law and control of corruption used in the present analysis were annual values averaged over the years 1998-2004.

Measures of In-Group Favoritism

Van de Vliert (2011) defined three types of in-group favoritism: compatriotism is favoritism shown to fellow nationals as opposed to immigrants, nepotism is favoritism shown to relatives, and familism as favoritism shown to one’s closest relatives in the nuclear family. Based on national surveys of these three facets of favoritism, Van de Vliert derived an indicator of general in-group favoritism in 120 countries.

Measures of Intimidation

We suppose that intimidation stems from the size of the home crowd and its perceived potential for violence. Crowd sizes were taken from Wikipedia (“List of Attendance Figures at Domestic Professional Sports Leagues,” 2013). The most obvious indicator of the level of aggressiveness in a country is perhaps the extent of violent crime. However, there are substantial difficulties in comparing the recorded crime statistics of different countries because of different practices in reporting and definition (Van Dijk, 1990; Vigderhous, 1978). For these reasons, national levels of aggressiveness were taken from the International Crime Victim Survey, which asks random samples of individuals about their own experience of being a victim of crime. The data set used covered 77 countries surveyed in several sweeps from 1989 to 2005, and is described in Van Kesteren (2007). For the present study, a nation’s aggressiveness was represented by the probability of an individual being a victim of assault in the previous year.

Measure of Altitude Diversity

Altitude measures were taken from the PLACE III geographical data set (Center for International Earth Science Information Network/Columbia University, 2012). For each country, this data set includes the area of terrain in each of 12 different altitude bands between 5 m and 5,000 m. These areas were used to calculate national values of altitude diversity. Diversity was measured by the Gini coefficient, first described by Gini (1912/1955, 1921). The Gini coefficient is an index of inequality frequently used in economics and ecology, and summarizes the dispersion of some quantity of interest among the members of a population. For a modern treatment of the Gini coefficient, see Milanovic (1997) or Yitzhaki (1998). The coefficient ranges between zero and one. A Gini coefficient of zero expresses maximum homogeneity, where the quantity of interest is allocated equally throughout the population; a Gini coefficient of one expresses maximum heterogeneity, where all the quantity of interest is allocated to a single member of the population, and none to the remaining members. In the present case, the quantity of interest is altitude, and each square meter of terrain represents a “member” of the population. Gini coefficients ranged between .06 (Andorra) and .68 (Colombia) with a mean of .45 (SD = .14).

Results

Table 1 shows the descriptive statistics and the correlations between the study variables. Correlations are shown below the diagonal and N sizes above the diagonal.

Table 1.

Descriptive Statistics and Correlations.

		n	M	SD	1	2	3	4	5	6	7	8	9	10	11	12
1	Home advantage (%)	74	61.4	5.9		46	42	67	67	54	64	38	66	54	35	69
2	Individualism	46	50.0	23.3	−.42**		35	45	45	42	46	32	46	38	28	46
3	Collectivism	42	0.009	1.5	.38*	−.65***		42	42	40	41	26	42	37	29	42
4	Rule of law	67	0.456	1.0	−.43**	.71***	−.50***		67	53	63	38	65	50	32	66
5	Control of corruption	67	0.486	1.1	−.36**	.71***	−.57***	.96***		53	63	38	65	50	32	66
6	Compatriotism	54	−0.154	1.0	.04	−.54***	.54***	−.41**	−.49***		54	33	54	42	31	54
7	Nepotism	64	−0.134	1.1	.41***	−.59***	.57***	−.72***	−.75***	.41**		38	64	50	32	64
8	Familism	38	−0.116	1.0	.33*	−.71***	.73***	−.74***	−.83***	.80***	.66***		38	33	22	38
9	In-group favoritism	66	−0.059	0.9	.24*	−.69***	.67***	−.73***	−.80***	.85***	.79***	.95***		51	32	66
10	Average attendance (000s)	54	10.01	8.9	−.13	0.14	−.06	.20	0.22	−.23	−.08	−.11	−.33*		30	52
11	Assault	35	3.0	1.3	−.28	.50**	−.33	.34	.37*	−.25	−.38*	−.44*	−.38*	−.13		32
12	Altitude diversity	69	0.45	1.4	.47***	−.32*	−.02	−.02	−.02	−.02	.09	.11	.00	.28*	−.15

Note. Correlations below diagonal. N sizes above diagonal.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

The negative correlation between HA and individualism and the positive correlation between HA and collectivism provide support for H1, that HA is elevated in collectivist countries. The negative correlations between HA and the rule of law and control of corruption indicate that HA is elevated in countries where official integrity is low and thus provide support for H2. The positive correlations between HA and three of the four measures of favoritism (nepotism, familism, and overall in-group favoritism) support H3. There is no evidence of a correlation between HA and either average attendance or assault probability, indicating that contrary to H4, intimidation does not seem to influence national levels of HA. Finally, the positive correlation between HA and altitude diversity provides support for H5.

Stronger evidence for the research hypotheses would be the existence of a relationship between the socio-cultural variables and HA that persists after accounting for other factors known to affect HA. There are at least two such factors available for all the countries in the sample. The first is altitude diversity as described above. The second is the FIFA rating of the country’s national team. FIFA ratings are constructed from the results of competitive matches between national teams. FIFA ratings were obtained from the rankings and statistics pages of the FIFA website (FIFA n.d., Men’s and Youth: Ranking and Statistics.) and averaged over the period 1999 (when the current points system was introduced) to 2005; averaging is justified because annual ratings are highly stable over this period, with a mean correlation between years of .96. The correlation between the FIFA rating and HA is .28 (p = .017).

Because of missing data, a single regression model involving all the socio-cultural variables would have reduced the effective sample size too much to draw any robust inferences. Separate regressions were therefore calculated for each of the socio-cultural variables that were significantly correlated with HA. In each case, HA was regressed simultaneously on altitude diversity, FIFA rating, and one of the socio-cultural variables. The results are reported in Table 2, but to summarize, six of the seven socio-cultural variables added significant explanatory variance to the regression, the exception being Familism. On average, the incremental explained variance was 12%.

Table 2.

Socio-Cultural Variables Regressed on National Home Advantage Controlling for Altitude Diversity and FIFA Rating.

Model	n	Standardized regression coefficients			Total R²	ΔR²
Model	n	Altitude diversity	FIFA rating	Socio-cultural variable	Total R²	ΔR²
1. Individualism	46	.15	.27	−.37*	.28	.12
2. Collectivism	42	.05	.50**	.38**	.42	.14
3. Rule of law	66	.36**	.11	−.44***	.34	.17
4. Control of corruption	66	.36***	.10	−.39***	.30	.14
5. Nepotism	64	.34**	.10	.39***	.27	.14
6. Familism	38	.25	−.26	.22	.20	.05
7. In-group favoritism	66	.39**	.04	.24*	.22	.05

Note. FIFA = Fédération Internationale de Football Association.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

One problem with this approach is the high correlations between the socio-cultural predictors; for instance, individualistic countries tend to have low levels of corruption and in-group favoritism, while adhering strongly to the rule of law. This means that the individual regressions are not really independent; disentangling the effects of the various predictors remains problematic, and the use of seven separate regressions, as opposed to one, raises the issue of inflated levels of significance.

One technique for dealing with correlated predictors in regression is to construct a summated scale, for example by factor analysis, and regress the dependent variable on that. In the present case however, this is not straightforward, because there are only 28 complete cases, which is not really enough to factor analyze seven variables reliably. I therefore adopted a multiple imputation procedure (Little & Rubin, 1989; Rubin, 1987). In this procedure, missing values are predicted from non-missing data and each set of imputations is used to construct a complete data set; the completed data sets are then analyzed using standard statistical procedures and the outcomes pooled.

Best practice currently recommends using 20 imputations (Enders, 2010). The fully conditional MCMC algorithm in the SPSS statistics package was used to create 20 imputed data sets. I then factor-analyzed the socio-cultural variables individualism, collectivism, rule of law, control of corruption, compatriotism, nepotism, and familism in the imputed data sets. A strong single factor which explained on average 70% of the variance emerged consistently. This factor had positive loadings (pooled estimates in brackets) on compatriotism (.64), nepotism (.84), familism (.90), and collectivism (.76), and negative loadings on individualism (−.82), control of corruption (−.93), and rule of law (−.91), and accordingly was labeled unethical collectivism. Factor scores were then computed, and HA was regressed on altitude diversity, FIFA rating, and unethical collectivism in each of the imputed data sets. The pooled regression results from the imputations are reported in Table 3.

Table 3.

Pooled Regression Results.

Pooled imputed regression	Model 1		Model 2
Pooled imputed regression	B (SE)	β	B (SE)	β
(Constant)	52.5 (2.2)		52.7 (2.0)
FIFA ranking	1.1 (1.1)	.12	1.9 (1.0)	.20
Altitude diversity	17.6 (5.1)	.40***	15.7 (4.7)	.36***
Unethical collectivism				.40***
Total R²	.22		.37
ΔR²			.16

Note. B = unstandardized regression coefficient; β = standardized regression coefficient.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

The results show that altitude diversity and FIFA rating together account for 21% of the variance in HA scores, and the unethical collectivism factor accounts for a further 16%, a statistically significant increase.

Discussion

The finding that national HA is unrelated to average attendance is consistent with the observations of Pollard (1986) and Clarke and Norman (1995), both of whom noted that HA was almost identical in the four divisions of the Football League in England, despite large differences in crowd size between the divisions. Together with the finding that national levels of aggressiveness are also unrelated to HA, it can be concluded that intimidation of match officials (or the visiting team) can be ruled out as a source of national differences in HA. However, it should be noted that the analysis of crowd size and violence reported here is “ecological”; that is, the statistics are calculated at an aggregate level (in this case the competitive league) as compared with the individual level (i.e., the team or match). It is well-known that statistics calculated at the aggregate level may be quite different to the same statistics calculated at lower levels, an observation which dates back at least as far as Robinson (1950). The lack of influence of crowd size and violence on HA reported here does not therefore imply a similar lack of influence would be found at the team or match level within a league. The overall findings reported here provide credible evidence that HA in football is, at least in part, a social phenomenon, whose extent is influenced by socio-cultural factors. The general picture is that of elevated HA in collectivist, in-group, and corruption-prone societies where the law is not strictly followed. This is a plausible conclusion, yet the detailed mechanisms underlying this phenomenon remain to be clarified. One possibility is that typical crowd behaviors at sporting occasions differ between countries. A second possibility is that the normative values of referees from different cultures may differentially determine their willingness to please the home crowd. Because referees and crowds in this study hail from the same country, we cannot distinguish between these competing explanations, or assess their relative importance.

As previously noted, HA differs between sports, and a natural extension of this research would be to investigate sports other than football. As Jamieson (2010) remarked, football is known for having rowdy fans who chant songs and jeers throughout a game, whereas baseball, which has considerably lower levels of HA, is known for a less intense atmosphere in which fans routinely leave before the game is over. We would predict that socio-cultural differences would be less evident in sports such as baseball, where the crowd is less salient, and in sports where results are less dependent on the subjective judgments of officials who might be influenced by home spectators.

On a more general note, football is played in almost every country in the world under a common regulatory framework, and football statistics thus provide interested researchers with a rich pool of comparable cross-cultural data. For example, Miguel, Saiegh, and Satyanath (2011) showed that the history of civil conflict in a player’s home country predicts his propensity to behave violently on the soccer field. Such studies however remain comparatively rare, and football appears to be under-researched in the literature of cross-cultural psychology. Yet in the same way that studies of teachers across the world allowed Schwartz to create and validate a system of cultural values (e.g., S. H. Schwartz, 1992), comparative studies of football players, spectators, and officials from different countries could contribute to our cross-cultural understanding of a wide range of group and individual behaviors.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Garry A. Gelade received a BA and a master’s in natural sciences from Cambridge University, where he majored in experimental psychology. He was awarded a PhD from Nottingham University for his thesis on visual perception, and subsequently worked at Oxford University where he collaborated with Anne Treisman on the feature-integration theory of attention. He has published in a number of fields including psychophysics, cognitive psychology, organizational psychology, and cross-cultural psychology. He lives in London and consults to leading business and sports organizations.

References

Asch

S. E.

(1951). Effects of group pressure upon the modification and distortion of judgments. In Guetzkow

Harold

(Ed.). Groups, leadership, and men (pp. 222-236). New York: Russell & Russell.

Bond

Smith

P. B.

(1996). Culture and conformity: A meta-analysis of studies using Asch’s (1952b, 1956) line judgment task. Psychological Bulletin, 119, 111-137.

Carron

A. V.

Loughhead

T. M.

Bray

S. R.

(2005). The home advantage in sport competitions: Courneya and Carron’s (1992) conceptual framework a decade later. Journal of Sports Sciences, 23, 395-407.

Center for International Earth Science Information Network/Columbia University. (2012). National aggregates of geospatial data collection: Population, landscape, and climate estimates, Version 3 (PLACE III). Palisades, NY: NASA Socioeconomic Data and Applications Center. Retrieved from http://sedac.ciesin.columbia.edu/data/set/nagdc-population-landscape-climate-estimates-v3

Clarke

S. R.

(2005). Home advantage in the Australian football league. Journal of Sports Sciences, 23, 375-385.

Clarke

S. R.

Norman

J. M.

(1995). Home ground advantage of individual clubs in English soccer. The Statistician, 44, 509-521.

Dawson

Dobson

Goddard

Wilson

(2007). Are football referees really biased and inconsistent?: Evidence on the incidence of disciplinary sanction in the English Premier League. Journal of the Royal Statistical Society: Series A (Statistics in Society), 170, 231-250.

Enders

C. K.

(2010). Applied missing data analysis. New York, NY: Guilford Press.

European and World Values Surveys Four-Wave Integrated Data File, 1981-2004. (2006). Surveys designed and executed by the European Values Study Group and World Values Survey Association (File VS\XWVSEVS_1981_2000_v20060423.sav) [Data file]. Madrid, Spain: ASEP/JDS, and Tilburg, The Netherlands: Tilburg University [File producers]; Cologne, Germany: ASEP/JDS and GESIS [File distributors]. Retrieved from http://www.wvsevsdb.com/wvs/WVSData.jsp

10.

Fédération Internationale de Football Association. (n.d.). Men’s and youth: Ranking and statistics [Web pages]. Retrieved from http://uk.soccerway.com/teams/rankings/fifa/?ICID=TN_03_05_01

11.

Fédération Internationale de Football Association Communications Division, Information Services. (2007). Big Count 2006. Retrieved from http://www.fifa.com/mm/document/fifafacts/bcoffsurv/bigcount.statspackage_7024.pdf

12.

Garicano

Palacios-Huerta

Prendergast

(2005). Favoritism under social pressure. Review of Economics and Statistics, 87, 208-216.

13.

Gini

(1921). Measurement of inequality of incomes. The Economic Journal, 31, 124-126.

14.

Gini

(1912). Variabilità e mutabilità C. Cuppini, Bologna (Reprinted in Memorie di metodologica statistica, Pizetti

Salvemini

, Ed., 1955, Rome, Italy: Libreria Eredi Virgilio Veschi)

15.

Gomez

M. A.

Pollard

Luis-Pascual

J. C.

(2011). Comparison of the home advantage in nine different professional team sports in Spain 1. Perceptual & Motor Skills, 113, 150-156.

16.

Hofstede

(2001). Culture’s consequences: Comparing values, behaviours, institutions and organizations across nations (2nd ed.). Thousand Oaks, CA: Sage.

17.

Jamieson

J. P.

(2010). The home field advantage in athletics: A meta-analysis. Journal of Applied Social Psychology, 40, 1819-1848.

18.

Jehue

Street

Huizenga

(1993). Effect of time zone and game time changes on team performance: National Football League. Medicine & Science in Sports & Exercise, 25, 127-131.

19.

Kaufmann

Kraay

Mastruzzi

(2003, June 30). Governance matters III: Governance indicators for 1996-2002 (World Bank Policy Research Working Paper, 3106). Available at SSRN: http://ssrn.com/abstract=405841

20.

List of attendance figures at domestic professional sports leagues. (2013, July 27)Wikipedia: The Free Encyclopedia. Retrieved from http://en.wikipedia.org/w/index.php?title=List_of_attendance_figures_at_domestic_professional_sports_leagues&oldid=5660232

21.

Little

R. J.

Rubin

D. B.

(1989). The analysis of social science data with missing values. Sociological Methods & Research, 18, 292-326.

22.

McSharry

P. E.

(2007). Effect of altitude on physiological performance: A statistical analysis using results of international football games. British Medical Journal, 335, 1278-1281.

23.

Miguel

Saiegh

S. M.

Satyanath

(2011). Civil war exposure and violence. Economics & Politics, 23, 59-73.

24.

Milanovic

(1997). A simple way to calculate the Gini coefficient, and some implications. Economics Letters, 56, 45-49.

25.

Morley

Thomas

(2005). An investigation of home advantage and other factors affecting outcomes in English one-day cricket matches. Journal of Sports Sciences, 23, 261-268.

26.

Nevill

A. M.

Balmer

N. J.

Mark Williams

(2002). The influence of crowd noise and experience upon refereeing decisions in football. Psychology of Sport and Exercise, 3, 261-272.

27.

Nevill

A. M.

Holder

R. L.

(1999). Home advantage in sport: An overview of studies on the advantage of playing at home. Sports Medicine, 28, 221-236.

28.

Nevill

Webb

Watts

(2013). Improved training of football referees and the decline in home advantage post-WW2. Psychology of Sport and Exercise, 14, 220-227.

29.

Pollard

(1986). Home advantage in soccer: A retrospective analysis. Journal of Sports Sciences, 4, 237-248.

30.

Pollard

(2006). Worldwide regional variations in home advantage in association football. Journal of Sports Sciences, 24, 231-240.

31.

Pollard

(2008). Home advantage in football: A current review of an unsolved puzzle. The Open Sports Sciences Journal, 1, 12-14.

32.

Robinson

W. W.

(1950). Ecological correlations and the behaviour of individuals. American Sociological Review, 15, 351-357.

33.

Rubin

D. B.

(1987). Multiple imputation for nonresponse in surveys. New York, NY: John Wiley.

34.

Schwartz

Barsky

S. F.

(1977). The home advantage. Social Forces, 55, 641-661.

35.

Schwartz

S. H.

(1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. Advances in Experimental Social Psychology, 25, 1-65.

36.

Sutter

Kocher

M. G.

(2004). Favoritism of agents—The case of referees’ home bias. Journal of Economic Psychology, 25, 461-469.

37.

Thomas

Reeves

Bell

(2008). Home advantage in the six nations rugby union tournament. Perceptual & Motor Skills, 106, 113-116.

38.

Triandis

H. C.

Bontempo

Villareal

M. J.

Asai

Lucca

(1988). Individualism and collectivism: Cross-cultural perspectives on self-ingroup relationships. Journal of Personality and Social Psychology, 54, 323-338.

39.

Unkelbach

Memmert

(2010). Crowd noise as a cue in referee decisions contributes to the home advantage. Journal of Sport & Exercise Psychology, 32, 483-498.

40.

Van de Vliert

(2011). Climato-economic origins of variation in ingroup favoritism. Journal of Cross-Cultural Psychology, 42, 494-515.

41.

Van Dijk

J. J.

(1990). Experiences of crime across the world: Key findings from the 1989 International Crime Survey. Brill Archive. Deventer, The Netherlands: Kluwer.

42.

Van Kesteren

J. N.

(2007). Integrated Database from the International Crime Victims Survey (ICVS) 1989-2005, codebook and data. Tilburg, The Netherlands: INTERVICT.

43.

Vigderhous

(1978). Methodological problems confronting cross-cultural criminological research using official data. Human Relations, 31, 229-247.

44.

Yitzhaki

(1998). More than a dozen alternative ways of spelling Gini. Research on Economic Inequality, 8, 13-30.