Introduction
As the only game in the world that is played in every country and by people of every race and religion football is one of the few institutions that is as exceptional because the United Nations.
-Former U.N. Secretary Basic Kofi Annan

The rules of your modern day game of footballi as we know it right now had been formalized by the English Football Association in 1863 and spread world wide through the British Empire inside the identical style as rugby and cricket. As football became a working class sport close to 1870 Skelton 2000 it also grew more and more common in the rest of the globe and is today deemed to become probably the most common sport inside the globe. For the duration of the final Globe Cup held in Germany in the summer season of 2006 32 nations qualified out of the 207 members of FIFA Federation Internationale de Football Association. Extra than 1.one billion tv viewers watched the last involving Italy and France globally and when it comes to participation youll find nowadays much more than 200 million active players throughout the world as football is 1 of a number of sports which are played on all 5 continents www. Dobson telescop fifa.com.

In spite of the apparent planet wide appeal of football you will discover only two continents that may show proof of international accomplishment. The World Cups happen to be entirely dominated by European and South American national teams as the eighteen World Cups currently being held due to the fact 1930 only had South American or European winners. In reality there have only been seven one of a kind nations to ever lift the trophy some far more than the moment such as Brazil who have won it an impressive 5 instances table five.
Table 1- International World Cup Achievement

Group
Titles
Runners-up

Brazil
five 1958 1962 1970 1994 2002
two 1950 1998

Italy
4 1934 1938 1982 2006
two 1970 1994

Germany
3 1954 1974 1990
four 1966 1982 1986 2002

Argentina
two 1978 1986
two 1930 1990

Uruguay
two 1930 1950
-

France
one 1998
one 2006

England
1 1966
-

Netherlands
-
2 1974 1978

Czechoslovakia
-
2 1934 1962

Hungary
-
2 1938 1954

Sweden
-
1 1958
hosts
involves benefits representing West Germany involving 1954 and 1990
states that have because split into various independent nations

However the Globe Cup isnt the only way of measuring efficiency. In 1993 FIFA started publishing a ranking of how every one of the national teams within the globe compare against each other on a monthly basis. Following going by two large changes above the years considering that its introduction the FIFA ranking technique is right now primarily based on the benefits each and every group has achieved for the duration of the last 4 years in which value of match strength of opposition as well as the outcome with the match ascertain how many points every group obtains. But more investigation of the prime 20 on the FIFA ranking only confirms the scenario described above- On the top 20 countries in the planet as of 15th of February 2007 only one nation was African Cameroon one particular Central American Mexico although the rest were either South American or European table two.

Why is this Why are highly populated nations example- China ranked as 79 not ready to outperform scarcely populated nations example- Portugal ranked as 10 contemplating China has got 124 occasions far more people to choose from www.imf.org with regards to picking their best eleven players. In other words- What tends to make a nation very good at football This question was the authors initial determination for writing this paper and by utilizing an econometric approach based on previous work on the topic an attempt will be made to search out clues that may possibly aid answering the question.

Table 2- FIFA ranking as of 15th of February 2007
Ranking
Country
Points

one
Italy
1562

2
Brazil
1540

three
Argentina
1535

4
France
1496

five
Germany
1359

six
England
1330

7
Netherlands
1312

8
Portugal
1262

9
Czech Republic
1193

ten
Spain
1161

11
Ukraine
1018

12
Croatia
987

13
Greece
926

14
Switzerland
913

15
Romania
912

16
Sweden
894

17
Cameroon
893

18
Denmark
876

19
Mexico
857

20
Scotland
854
Source- www.fifa.com
2Literature evaluation
Only lately have economists taken an academic interest in football. In spite of the large public interest in each international and club football the quantity of literature on this subject is surprisingly scarce and for that reason the following section will give a short outline on which topics academics have focused on. The research that has been carried out has frequently had financial determination but this will need to come as no surprise contemplating the huge commercialisation the sport has gone by means of above the last 20 years Haugen and Hervik 2002. The key concentrate of the literature hitherto has been how clubs frequently European clubs have transformed from wanting to win football matches into large-scale firms holding a few of the worlds biggest trademarks--such as Actual Madrid and Manchester United worth 186.2mill and 166.4mill respectively Deloitte 2007.
2.1Sport- Micro degree
Financial theory on football is largely based mostly on the earlier theory written on sport in general and for that reason a short outline of this literature will also be integrated within the following.

The demand for sport generally has been thoroughly investigated among the earliest currently being Neale 1964 which concerned the demand for baseball in USA. Neale illustrates the -firm- in qualified sports because the league and suggests the -products- the -firm- is promoting is 1 the match and 2 the league standings. He claims that the closer the league standings the improved could be the item plus the increased is going to be the demand for the item.

The demand for football has also been cautiously studied. Hart et al 1975 identified that attendance to football matches in general was explained by ticket charges distance to travel for away fans the calibre of the opposing group and personal group efficiency. Bird 1982 employs time series data to investigate how attendance for English football matches have created as time passes and finds that demand for football is earnings inelastic making football an inferior very good. He also reports that the findings of ticket price tag inelasticity is predicted to be exploited by the market an insight that certainly materialized over time as ticket rates for prime division English football nearly have quadrupled due to the fact 1982 Dobson and Goddard 1999 pp75.
The demand for televised football has also been investigated Baimbridge et al. 1996 and Forrest Simmons 2002. The TV revenue for the 2006 Globe Cup reached 525 million and recent news of a new TV deal for the right to broadcast English football outside Britain really worth 625 million for the following three seasons exemplifies how much money is going into and in turn is expected to be created on the sport of football from a commercial point of view.

Another topic that has been given broad attention is how player contracts free agency and transfer payment have developed and changed with time most significantly following the Bosman Ruling in 1995. For examples see Sloane 1969 Speight and Thomas 1997 Charmichael and Thomas 1993 and 2002 Dilger 2001 and Fees and Muelhauser 2002. The primary findings on this topic are that the contract length increases because the property rights of a player are transferred from the club to the player as was the legal consequence of the Bosman ruling.
Other aspects of club football which have been researched include proof of discrimination Szymanski 2001 Reilly and Witt 1995 and Charmichael and Thomas 2000 and league reward systems such as if the 3 points for winning increased the attacking play which has been investigated by Palmino et al 1999 and Guedes and Machado 2002 amongst others.
2.2Sport- Macro level
Contrary to the financial and club degree referred to as the micro degree there is the macro financial degree which could be the principal focus in the following paper. The macro financial term is used when discussing how football plus the rest on the society are integrated. The book -France plus the 1998 Globe Cup- Dauncey and Hare 1999 gives a thorough analysis of how hosting a World Cup affects the country as a whole the two economically as a result of an increase in tourism but also on aspects this kind of as the -feel good factor- and how a well-organized tournament can influence politicians popularity with time. In other words the book explains how football has influenced the political and economic environment of a Planet Cup hosting country.
Most with the literature on footballsports functionality and economics at the macro level adjustments the causality from the Dauncey et al 1999 book into- Which countrys specific elements ascertain how well a nation performs in a given sport You will discover numerous articles thinking of which aspects determine sport performance implicitly and football overall performance explicitly. Most from the literature covering sport functionality helps make use of Olympic Games medalsparticipation because the dependant variable. The basic idea throughout the literature is to use gross nationaldomestic item GNPGDP per capita and weathertemperature as explanatory variables with small variations.

Kuper and Sterken 2001 investigates Olympic accomplishment making use of information on winners from the winter Olympics where the major finding was that in group sports exactly where a referee is required there is going to be proof for a home team advantage in addition to GDP per capita which is also identified to be significant. Other articles on Olympic success are Johnston and Alis 2000 investigation of female participation rates in the Olympics and an article by Hoffmann et al 2002b taking into consideration how political systems influence a countrys Olympic good results.
2.3Football overall performance
The first article investigating determinants of football performance to the authors finest knowledge was written by Hoffmann Lee and Ramasamy 2002a entitled -The socio economic determinants of international soccer performance- hereon termed HLR. HLR utilizes the FIFA Planet ranking of January 2001 as the dependent variable to investigate which country certain aspects establish football performance. They discover that population as a single explanatory variable has no impact on overall performance. To more investigate the idea of whether a increased population increases the efficiency from the national football group HLR introduced a variable intended to indicate football tradition which was named LATIN. This variable was used as a dummy variable for countries from Spain Portugal Central America and South America. The justification for utilizing this variable were according to HLR underlying cultural things within the Luco- Hispanic culture that promotes male participation and support of sporting events in these nations. Even if this variable was insignificant when used on its own when multiplied with the population variable the final results became significant. This is explained by claiming that a larger population does not necessarily mean improved football players unless the extra people today actually play football. Other variables incorporated were GNP per capita temperature the countrys share of world population and a HOST dummy also intended to capture football traditions.

In another paper by Torgler 2004 the concentrate is on which factors decide female football efficiency. This paper is built upon the Hoffmann et al article. Torgler can make a number of amendments however namely the change in the temperature variable is taken as an average temperature for the country as a whole rather than from the capital of the country investigated. He also modifications the proxy from ranking points on the FIFA globe ranking to the actual ranking position every country has implicating a lower number equals improved functionality. The key finding in this paper is that the key determinants of female football performance are population football tradition and GDP per capita.Macmillan and Smith published a paper in December 2006 also seen as a response to HLR discussing econometric issues in the article this kind of as sample bias along with the poor fit of the model.

Just before the Planet Cup final in 2006 Gelade wrote an article entitled -Academics obtain Formula for the ideal international football team.- Utilizing 5 variables the paper claims to have worked out a new way of ranking international football teams but this ranking is ambiguous at very best. Making use of country distinct variables such because the number of men playing football number of years as a member of FIFA number of internationals playing abroad wealthand climate this formula ranks World Cup winners Italy as number three France as number four and Brazil as number 18 The reason given why these benefits are not realistic is because the formula isnt ready to quantify football passion a variable that if included is certain to boost poorer countries this kind of as Brazils score significantly.
3Method
In this section an outline with the dependent and explanatory variables determining international football performance will likely be given. As mentioned above this paper will take an econometric method based on data comprehensively available around the internet. Substantial effort has been made to collect the most recent data available mainly utilizing official sources this kind of as FIFA Globe Bank and International Monetary Fund. The following regression will draw partly around the perform of HLR as this could be the first most important article around the subject but will also use others input e.g. Togler 2004 Macmillan and Smith 2006 and Gelade 2006. In addition to previous result to two new explanatory variables not previously tested are integrated ELITE and HEALTH.
3.1Sample
HLR makes use of the 76 medal winning countries from Summer time Olympics in Sydney 2000 as their sample pool arguing that this would be a bias free sample. However MacMillan and Smith 2006 claim that this sampling method could cause sampling bias because the countries are not chosen randomly plus the countries are additional likely to be chosen from the upper end with the ranking table. To avoid this Macmillan and Smith run a regression which consists of 176 nations to avoid the problem of sample bias and within the regression below 179 nations are integrated in an try to avoid sampling bias. The excluded nations are characterized by non-published information for the explanatory variables nations this kind of as Cuba Iraq Somalia and North Korea. Additional England as the biggest UK nation is chosen to represent United Kingdom excluding Northern Ireland Scotland and Wales.
three.2Dependent Variable
The dependent variable would be the FIFA ranking score as of 15th of February 2007 located on www.fifa.com and widely published by the globe press. The method for calculating the FIFA score was altered following the 2006 Globe Cup and to the authors finest knowledge has no paper been published employing the new algorithm of determining the FIFA ranking. The ranking was first introduced in 1993 but has been heavily criticised for becoming both extremely complicated and for being unable to give a realistic ranking by the years. It was first altered in 1999 in an try to compensate for the enormous criticism but as there was still no evident outcome that the ranking mirrored the true performance of international teams a completely new algorithm were developed soon after the Globe Cup in Germany in July 2006. The new simplified algorithm works as follows- For every single match the national team plays 3 points is awarded for victory one point for draw and zero points if the group loses. Next the importance of the match is weighted by awarding one particular point for friendly two points for an intercontinental cup e.g. Euro Cup qualifying game three points playing in an intercontinental cup and 4 points for a World Cup match. Strength of opposition is calculated using the oppositions ranking in the existing FIFA ranking. Subsequent each and every conference is weighted where UEFAii is weighted at one.00 followed by CONMEBOLiii weighted at 0.98 and the many other confederations becoming weighted 0.85. So for each match played these components are multiplied to get a ranking score for both teams in any given international -A- match. Matches played within the final four years contrary to eight years that had been the case before the change are taken into account when calculating the month-to-month FIFA score. Additional the latest matches are weighted as follows- The last twelve months count in full where the previous year only count for half along with the games played three and four years ago have decreasing significance only 30 and 20 respectively. Two aspects that were removed inside the new calculation method were the number of goals scored and homeaway advantage in each match.

To test for robustness of your specification the preferred regression will also be ran making use of the two Could 2006 data before the change in ranking method and on an alternative ranking system created by football enthusiasts around the Internet called the ELO ranking.
three.3Explanatory variables
The basic variables incorporated inside the specification are drawn from HLR and we will initially test the robustness of their regression working with updated data on the two dependent and explanatory variables. We will then investigate additional possible variables and we will also consider removing a few of the original variables to allow for improved fit with the model.
3.3.1Population
The first explanatory variable is every single countrys fraction of planet population which theoretically really should be positively correlated with football functionality as a larger pool of potential football players to draw from should increase the possibility of becoming successful inside the sport. Population has been shown to be significant when testing Olympic good results Hoffmann et al 2002 even if the final results have shown that this variable has become less significant as time passes Johnson Ali 2004.
3.3.2GNP and GNP2
The second explanatory variable hypothesizes initial development of football talent to be extremely relevant for the functionality degree from the national football group. Even if football is a low cost activity private access to equipment and available leisure time must be available and this is easiest to measure by making use of GNP per capita. However as earnings rises above a threshold degree the marginal effect of extra income becomes negative Hoffmann et al 2002. As HLR note you will find two possible explanations of this relationship. First football is an inexpensive sport to engage in when compared to sports this kind of as car racing golf sailing etc. For this purpose it is possible to argue that poorer people today will over invest in playing football simply because there are handful of other sporting alternatives available to them. The second cause is similar in fashion to the first stating that as gnp per capita increases not only will other sporting activities act as substitutes for football but also other activities children get engaged in this kind of as video games DVD satellite TV and other indoor activities may perhaps compete with the time children spend playing football. To capture this we use gnp alone and as a quadratic function GNP2 expecting a u shaped curve and a negative coefficient in which there also will exist an optimal degree of gnp for developing football abilities. Data on population and gnp per capita have been gathered from the World Bank www.worldbank.org 3.three.3Temperature
Following temperature is used as an indication of climatic conditions in every single county. Earlier studies regarding Olympic accomplishment Hoffmann et al 2002 have shown that the ideal yearly aggregate temperature is approximately 14C and any deviation from this temperature is expected to have negative impact on sporting efficiency. It is easy to imagine how nations with extreme temperatures will have difficulties performing any sporting activity at an optimal level. To model this HLR use a variable that picks up any deviation from the preferred temperature given by Temp-14two expected to have a negative coefficient.

Other researchers have used climate as a variable to replace temperature Gelade 2006 but inside the following the variable controlling for exogenous weather conditions could be the deviations from the optimal 14C temperature described above.
Temperature data is collected from www.geographyiq.com and makes use of the yearly average in every single nations capital. Making use of temperature gives a precise estimate of your exogenous weather conditions even if the temperature is only estimated in the capital. The cause why this method is chosen is because a nation can have huge domestic variations in yearly temperature. The capital is however normally the highest populated city in each country plus the temperature here will affect always a massive proportion on the population.
three.3.4Latin
HLR also incorporated two variables regarding cultural things promoting a countrys football achievement. The first variable is included intending to capture how well-known football is in a given country both when it comes to numbers of spectators and amount of active players in every single country. Directly popularity when it comes to a lot more individuals watching football matches live or on TV increases financial and status incentives for players. Indirectly higher degree of player rewards and increased popularity may well over time increase the pool of potential national players.
To find a variable thats capable to illuminate the cultural aspect which affects football efficiency is a challenging task as this relationship has proved to become a complex one particular Archetti 1999 Giulianotti 1999 Lever 1995 as cited by HLR. Lacking alternative variables proven to show significant benefits working with the LATIN variable as reported by HLR as one of their key findings appears as a decent proposal. The way HLR justifies the use of this variable is by investigate the prime ten around the FIFA ranking as of January 2001 noting that eight of these nations have predominantly catholic population the only exceptions being Germany and England. HLR further notes that a romantic language referred to as LATIN by the American Heritage Dictionary AHD is spoken in seven of the leading ten countries and only two prior World Cup winners does not share these features again Germany and England. Even if religion does play an important role in shaping a countrys culture HLR makes use of language as a proxy for cultural attributes that increase international football functionality. In turn the romantic languages differ internally and HLR chooses the Luco-Hispanic languages as defined by the US library of Congress- All Spanish and Portuguese speaking nations inside the planet to become precise thats the nations in Central and South America plus Spain and Portugal. According to HLR these nations share underlying cultural elements that support the high popularity of mens football the two as spectator and participation within the sport. Within the 2007 sample the nations in the top ten have changed somewhat but as eight out of ten nations is still defined as LATIN according to AHD this variable is incorporated in the following specification as 1 measure of football traditions.
three.three.5Host
HLR also uses a variable called HOST a dummy set to among the country has previously hosted a World Cup and zero otherwise. This variable is included as a second variable for football culture. By indicating that a country has hosted the Globe Cup it need to give an indication that this distinct nation has long football traditions implicating that a lot more folks are likely to play football in that nation and therefore increase the chance of a greater FIFA ranking score.
The source of this information came from the www.fifa.com.

The abovementioned variables have been the ones incorporated in the specification of HLR. For the remainder of this paper we will investigate the robustness of these variables working with 2007iv information. New variables will also be added to investigate how this effects the regression and its final results regarding to fit and significance.
3.three.6History
As mentioned by most authors on this subject by far the most tricky variable to find a proxy for could be the so called -football tradition- proxy. HLR chose the variables LATIN and HOST and whilst the more use of LATIN has been justified above hosting the Globe Cup as a measure of football tradition is possibly not the preferred indicator of football traditions. As there has only been 14 special hosts out on the 207 member nations including this variable is bound to reject valuable information. This is also mentioned by Macmillan and Smith along with the solution they found was to drop the variable HOST and replace it with a variable in which there is information available for every one of the countries within the sample namely the number of years the country have been an affiliate of your FIFA. This variable is named HISTORY and is expected to have a positive effect on international football efficiency as longer membership need to mean longer traditions. Again data had been collected from the FIFA webpage.
3.three.7Republic
Macmillan et al do however mention one particular problem in making use of the history variable referring to the number of -new- states appearing in Eastern Europe immediately after the fall of the Soviet Union the splitting of Czechoslovakia and post war changes to former Yugoslavia in the early nineteen nineties. The problem this elevates is that the new nations will only have a short FIFA history. A single way of addressing this problem is to create a dummy variable called REPUBLIC set equal to one for each of the nations which have been a former republic. This variable is intended to reduce the point loss occurred to the affected Eastern European countries of only having approximately 15 years of history when the fact is these countries do have long football traditions as part of other nations.
three.three.8Elite
A absolutely new variable intended to become included is one measuring football skills applying the percentage of internationals currently playing in the top division inside the prime 5 leagues in Europe. According to UEFA ranking www.uefa.com prior to UEFA Champions League 2006-2007 these have been Spain La Liga England Premier League Italy Serie A France Ligue 1 and Germany Bundesliga This variable is called ELITE. A different version of this variable is included in Gelade et al 2006 however that was a variable of how quite a few players playing abroad. Merely playing abroad does not necessarily increase the quality on the national group as there could be numerous reasons why a player is playing club football in a different nation than his native a single. A single cause why players play abroad could simply be because of prior generations emigration or players going abroad to a lower league club for the -experience-.

By only counting the players inside the prime five leagues within the world the players simply playing abroad can be distinguished from the players in demand from the very best clubs in the globe- Specialized football clubs are ran like any other enterprise and a team in one of many leagues mentioned above spends huge resources on scouting networks intended to bring within the worlds greatest talent from all over the globe. As having a lot more players playing around the highest degree should make the overall national team efficiency improved this variable is seen as extremely relevant. Another purpose why high profile international footballers can increase the overall performance in the national team as time passes will be the increased interest in this player in his home country giving young players inspiration to become skilled footballers.
The information on how lots of players playing abroad have been discovered at www.national-football-teams.com.
3.3.9Health
The final variable is a new proxy intended to quantify how the degree of health in a nation affects football overall performance. The arguably finest indicator of a countrys health is to gather data on healthy life expectancy Robine Romieu and Cambois 1999. A healthier population will need to increase the pool of potential football players which in turn theoretically will need to increase the overall football overall performance of that country. Data on healthy life expectancy had been gathered from the web pages from the Planet Health Organization www.who.org.
4Results four.1Robustness check of initial regression
First making use of 2007 data we will replicate the original regression from HLR employing equation one and investigate and examine the benefits to check how robust the model is. Overall lower coefficients for the HLR study are anticipated because the score points obtained under the old technique were lower than it is today Italy is number 1 in 2007 with 1562 points while Brazil only had 821 points as number 1 in 2001.

Y 1GNPi 2GNPi2 TEMPi-142 POPi x LATINiHOSTi i1
Table 3- Robustness of HLR estimates standard errors in brackets
Personal estimates
HLR estimates
Sample size- 179
Sample size- 76

Variable
Estimate 2007
t-value
Estimate 2001
t-value

Constant
408.2982
ten.34616
492.5865
19.2582
39.4637

25.578
GNP
0.016409
3.624019
0.0107
2.3742
0.0043

0.0045
GNP2
-2.7210-8
-2.957301
-2.4510-7
-1.6875
6.8710-8

one.451910-7
TEMP-142
-1.292244
-4.759471
-0.4895
-1.9848
0.2715

0.2466
POP x LATIN
20286.55
2.406350
8587.4616
two.1828
8430.42

3934.1495
HOST
492.5931
five.492097
81.0510
one.8238
89.6913

44.4407
Adjusted R2
0.4707

0.3180

Note- and denote significance at 10 5 and 1 respectively

As can be observed from the left hand side regression is that all of the variables are significant at the 5 degree most even at a 1 level and that the signs are as expected. The GNP coefficient is only 0.0164 and can be interpreted as a 1000 dollar increase in GNP per capita will increase the FIFA score by 16 points. The sign for the quadratic relationship of GNP2 is negative as anticipated and also significant which indicates that there is an optimal degree of GNP per capita to increase football efficiency. To search out the optimal degree of GNP per capita the first derivative with respect to GNP is taken and set equal to zero. The optimal level is according to this regression 30163 which is considerably increased than the outcomes from HLR 21836.

The temperature coefficient is negative as expected and also highly significant. The variable measures the deviations from the optimal degree of 14C and can be interpreted within the following manner- A deviation from the optimal temperature level decreases the FIFA score along with the additional you get away from the optimal temperature degree the larger is the decrease in football functionality. The relationship is quadratic and a high deviation from the optimal temperature could possible mean a substantial decrease in estimated FIFA ranking points.

Even though LATIN and POPULATION were tested individually only significant benefits can be observed for the LATIN variable. However the variable LATIN POPULATION proves to become significant just as in HLR. This means that the size of a population should not increase football efficiency significantly unless the nation is of Latin origin as defined above. The coefficient on the LATIN POPULATION variable is significant in both studies but has a much lager impact in our study compared to HLR. If a countrys population relative to the rest of your world grows by 1 86 points will probably be added in HLR specification working with 2001 data when in our new specification 203 points are going to be added for a 1 increase in relative planet population making use of 2007 data.

The Host variable according to this regression have a huge impact on football functionality and also the truth that a country have hosted the Planet Cup within the past must increase the points gained by that respective country by 492.six points. This is a lot increased than the outcome obtained by HLR but this is as expected with background inside the relative difference in point calculation.

The fit in the model is fair with an adjusted R2 of 0.4707 and even if it is larger than the outcome HLR located 0.31 it does suggest there to be other potential aspects determining football efficiency. To further test for specification errors the model has also been tested utilizing Ramsey RESET test 1969. This test utilizes additional variables called proxies intending to act as replacement for missing variables. If these proxies do indeed increase the fit of the model this means that there is a high possibility of omitted variables within the model. The F score in the RESET test is one.45 and is lower than the critical value 2.ten which mean we cannot reject the null hypothesis of no specification error. However as the R2 are relatively low we will make some amendments to the regression in an try to improve the fit of your model. Also the Durbin Watson test statistic DW testing for first order serial correlation in this case is 1.798 which means that the test is inconclusive at the five percent degree for 6 coefficients and a sample size of 179 1.69841 one.81311v which gives us another cause for doubting the specification.
4.2The new model explained
The first change we will make to the model is to substitute the HOST dummy variable by a new variable controlling for football traditions. The variable is inspired by Macmillan and Smith currently being the HISTORY variable measuring how many years every nation has been a member of FIFA. The advantages of employing this variable compared to the HOST variable is that we have information on all of the 179 countries for this variable plus the fact that this variable is also in a position to measure relative difference in football passion as is a continuous variable and not just a dummy. As mentioned above we also include the dummy variable REPUBLIC to measure the impact of beneficial football nations with a short football history in Eastern Europe. The sign from the HISTORY coefficient is expected to become positive as a longer affiliation with FIFA ought to mean that the nation have stronger football traditions. The sign of the REPUBLIC variable is also predicted to be positive as a youll find numerous examples of great -new- football nations this kind of as Ukraine Czech Republic Serbia etc.

To additional investigate which other possible country certain indicators that determines football performance the above-mentioned variable called ELITE is integrated to investigate how footballers playing inside the top rated leagues inside the planet affect a countrys football overall performance.

As mentioned above we also wanted to test for healthy life expectancy to see if this has the hypothesised positive effect on football performance. Running test adding the HEALTH variable to the preferred specification above HEALTH was identified to become insignificant at all levels and also to have an unexpected negative sign. This outcome suggests that healthy life expectancy according to the new model does not affect football performance to any greater degree. Another variable that has been tested will be the variable of how substantial the fraction of GDP that is certainly spent on health services. The problem employing this variable is that it has been investigated and discovered to become a poor indicator of the degree of health in a nation Filmer and Pritchett 1999. As there are no information supporting other measures of healthy living without going into complicated calculations Filmer Hammer and Pritchett 1997 the following regression will not include a variable on health. The minimum degree of health needed to perform as a football nation need to anyway be captured by the GNP2 variable integrated above.
Equation 2 summarizes our new model of international football achievement with the variables summarized in table four. The regression is displayed in table 5 and also compares the new model to HLRs model with the two regressions working with the FIFA ranking score from 2007 as dependent variable.

The regression-
Y 1GNPi 2GNPi2 TEMPi-142 POPi x LATINi HISTi REP ELITEi itwo

From table 5b observe that each of the variables remain significant and that the sign for the two expected negative variables remain negative.The magnitude of each and every of your remaining variable does not change a great deal so this could indicate that the host variable isnt omitted inside the new regression. The new variable HISTORY is extremely significant and has a positive relationship with the FIFA score. The interpretation of HISTORY is that one extra year as a member of FIFA increases the FIFA score by three.72 points. Currently being a former republic with only 15 years history is going to be compensated by the enormous lump increase in FIFA ranking points of 166 which will be the equivalent of approximately 44 years affiliation.

Next consider the effects of including the ELITE variable. Observe that the new variable ELITE is also extremely significant with a coefficient of ten.14. This is a rather massive number and can be interpreted by saying that a 1 percentage point increase within the number of footballers exported to the worlds elite clubs will increase the FIFA ranking score by approximately ten points. Adjusted R2 increases and indicates that 74 of football performance can be determined inside this model suggesting that this model is far better specified than the 1 suggested by HLR. The optimal degree of GNP per capita is in this model 18678 which is close to HLRs estimate.
Table four- Regression Variables

Variable
Description

Y
FIFA ranking points February 2007

GNP
GNP per capita of nation i

TEMP
Average annual temperature in country Is capital

POP
Nation Is share of planet population

LATIN
Latin dummy

HIST
Number of years country I has been an affiliate of FIFA

REP
Republic dummy

elite
Number of players in national squad playing in an elite club

Error term

1 2
Parameters

Table 5- Regression results standard error in parenthesis
a Old regression

b New regression

Sample size- 179

Sample size-179
Variable
Est. HLR07 information
t-value
Est. NEW07 information
t-value

Constant
427.9397
9.96
144.5484
three.0254
42.97

47.77751
GNP
0.015252
three.0941
0.0065
1.9768
0.00493

0.003267
GNP2
-2.5710-7
-2.5684
-1.74E-07
-2.6802
one.0010-7

6.5110-08
TEMP-14two
-1.3075
-4.4231
-0.6575
-3.0808
0.2956

0.213417
POP x LATIN
20074.75
2.1871
16273.28
two.9527
9178.504

0.835036
HOST
504.3909
5.1653
97.65
HISTORY
3.729681
5.5093
0.676978
REP
165.6135
2.7829
59.51191
ELITE
10.141
12.2655
0.826782
Adjusted R2
0.4247

0.7400

Note- and denote significance at ten five and 1 respectively
4.3Testing the model
Running the RESET test on this specification increases the F score to 3.63 which is above the critical value of 2.10 indicating rejection with the null hypothesis of no specification error. This could potentially be a rather severe problem as we according to the RESET test is forced to assume this specifications just isnt correct contrary to the regression making use of HOST as an indication of football culture. Theory suggests however that when comparing two or additional specifications Akaikes criterion AIC and Schwartz criterion SC are recommended more than the RESET test to test for specification error Studenmund 2001. Although the RESET test ran above does suggest specification error within the last regression econometric theory suggest that employing a variable which might be available for all of the countries within the sample will need to be a improved measure of football tradition than a dummy. Note that the HOST dummy affects less than 8 of your sample as only 14 out from the 179 countries have ever hosted the Planet Cup though there is information available on all nations regarding FIFA affiliation. Comparing the AIC and SC from the two specifications supports this view and it is observed that the AIC and SC falls from 13.97 to 13.26 and 14.07 to 13.41 respectively.

Also the DW test statistic is 1.98 making it unreasonable to reject the null hypothesis of no serial correlation. These three tests Adjusted R2 AICSC and DW increase the confidence of improved ability in the new specification to estimate every countrys FIFA ranking score compared to HLR.

Equation three shows the new model including the coefficients.
Points 144.55 0.0065 gnp - 0.000000174 gnp two - 0.6575 temp- 14two
16273.28 poplatin 3.7230 history 165.6135 republic ten.141 elite i3
5Discussion and Robustness checks five.1Discussion
So how well does the new model perform with regards to determining how several points any given country receives around the FIFA ranking To investigate this the regression will be run around the data available about each and every nation. When performing this test there might be many cases where the chosen specification does not explain how well a nation performs and additional investigations are going to be created to search out any patters that could be the basis for further studies to improve the model.
When testing the regression on actual information it becomes clear that the point estimated by the regression did rather well overall estimating 125 with the 179 70 countries within one particular standard deviation from the genuine FIFA score as of February 2007. This is illustrated in figure 1 which shows the residual plot of all the nations integrated inside the study. The two dotted lines indicate the standard deviation and simply by eyeballing the information we can see that a fair amount of the variables are within a single standard deviation from the true value.
Figure 1- Residual plot

What seems striking is that Table 6 shows how within the leading 20 only eight nations 40 have been predicted a score within a single standard deviation form the actual score. This suggests that the specification is lacking elements that can emphasize how well the nations at the top rated on the ranking really perform.

Table 6- Checking how the model performs

Nation
Points
Est. points

Difference
1 SD from actual
Two SD from true

Italy
1562
1457
105
one
one

Brazil
1540
1549
9
1
one

Argentina
1535
1430
105
1
one

France
1496
1557
62
one
one

Germany
1359
1550
191
0
1

England
1330
1533
204
0
1

Netherlands
1312
855
456
0
0

Portugal
1262
958
303
0
0

Czech Republic
1193
999
194
0
one

Spain
1161
1694
533
0
0

Ukraine
1018
406
611
0
0

Croatia
987
627
360
0
0

Greece
926
706
220
0
1

Switzerland
913
805
108
1
one

Romania
912
591
320
0
0

Sweden
894
703
190
0
one

Cameroon
893
565
327
0
0

Denmark
876
865
11
1
1

Mexico
857
792
65
one
1

Cte dIvoire
853
718
134
one
1

SD176.27

The Standard Deviation SD in case is 176.27 and if we look at countries such as Netherlands Portugal Spain Ukraine Croatia Romania and Cameroon we can see that these countries scores are a lot more than two SDs from the true score which shows that the model is hardly predicting these countries at all. Ukraine and Spain are the two outliers predicted 611 less and 533 points more than what they really have respectively. Within the overall study utilizing all the countries these two nations had been the highest outliers above all and will later be the subject of a case study to examine where the regression failed.

Comparing the true points with the estimated points figure two show that the model predicts the lower scores better than the higher scores.

Figure two- Comparing how several points needed for each and every ranking position
Later additional investigations will likely be made inside the case of Brazil which has been leading from the ranking for pretty much four and a half years prior to the February ranking and which was estimated only 9 points away from the true score. Gambia will also be examined as it was the only country estimated with the exact score 163 vs. 162.96 and discussions will likely be produced to see if this was due to chance or if it can be contributed to the effectiveness from the new model.
5.1.1Ukraine-
Ukraine was by far the most underestimated nation of all of the nations in the sample. As the model catches 70 of all the countries scores what is it with Ukraine that tends to make the new model miscalculate this country so grossly

The GNP of Ukraine is only 1532 per capita which is far from the optimal level in our specification 18678. By inserting these values into the specification for GNP we discover that Ukraine only wins 9.52 points for their GNP per capita.

Next the average temperature in Ukraine is 8.two degrees Celsius which is off the ideal temperature by practically six degrees and actually creates a loss of 22 ranking points. Ukraine does not have a Latin population and are therefore not awarded any points for their 25 million inhabitants 0.36 of world population according to this model. Following Ukraine became independent from the Soviet Union in 1991 giving them only 17 years of membership to boost their score. This is far shorter than a lot of other nations that Ukraine can compare them selves too such as Russia Former Soviet and Poland which has been an affiliate considering 1912 and 1923 respectively. As a former republic Ukraine does however benefit from the Republic dummy which adds 165 points to the total score in addition to the 63 points won for having 17 years of membership.The last coefficient may be the ELITE coefficient and our data shows that only 5 with the current national team plays their club football in among the list of worlds very best leagues adding only 50 points producing it a total of 611 points.

In short Ukraines low wealth cold climate short independent history and couple of players playing within the best leagues inside the planet are the principle reasons for the low score they received calculated by the new specification.
5.1.2Spain-
Following investigations of how Spain managed to get the highest score of all countries utilizing our specification are performed. The mystery of Spain underperformance is well known within the football world and several have asked why a country which such a vast amount of talented footballers have not won any main trophies considering their 1964 European Championship Ball 2003.

First a short look at the data The GNP per capita is 26458 which is well above the optimal level yet it increase the FIFA score by 49 points. The subsequent variable could be the temperature variable along with the average temperature of Spain is 14.3 Degrees Celsius which is very close to optimal and therefore does not deduct any points from the total.

Spain has a Latin population as defined by HLR and a population of 41 million 0.62 of planet population contributes to increase the overall score by 100 points. As Spain was among the list of countries who established FIFA in 1904 they have the longest history as an affiliate of 103 years which adds a further 383 points to the total score. The final relevant variable could be the ELITE variable and this information confirms what mentioned above about the number of quality players the national team have at its disposal 100 of your players within the current national squad plays inside the top rated division in the principal leagues in Europe where a huge majority plays their football domestically in La Liga. This truth adds a massive 1014 points to the total score and is probably the principle purpose for giving Spain the highest estimated score calculated by our regression. When investigating the Elite variable additional by looking at table three England Germany and France all have had their scores overestimated by the formula. As these countries in addition to Spain and Italy have been the only nations to have 100 with the national players in the 5 top leagues it is possible that the Elite coefficient is overestimated and must be adjusted down to achieve a a lot more precise specification. For the moment however no adjustments will likely be created to this variable.
5.one.3Brazil
Brazil has been ranked on prime on the FIFA ranking for 55 months consecutively before they were knocked down from the best spot by Italy in February 2007. This is a country that the new regression managed to estimate exceptionally well with only a 9 point difference between the true along with the estimated score. Why the specification seems to pick up so much on the cultural and economic factors in the Brazilian society will probably be inspected next.

Brazil has a GNP per capita of only 3448 which is far from optimal and only adds 20 points to the estimated score. Following the average temperature in Brazil is 21.8 degrees Celsius which also is about eight degrees increased than the optimal temperature and decreases the estimated score by 40 points. Brazil is however the largest Latin nation in the globe with a staggering 187 million inhabitants two.9 of world population. From this variable Brazil gains 460 points this is about 13 of their total score. Brazil has also got long football traditions getting an affiliate of FIFA for 84 years which additional increase the predicted score by 312.five points. Even if the Brazilian national group normally consists of international superstars according to the database we used around the ELITE variable only 64 of your national squad plays their club football within the planet finest clubs which wins them another 640 points.

This confirms what mentioned inside the criticism of Gelandes paper that Brazil do rely heavily on football traditions as few other components identified can predict them to become among the worlds best football countries.

As can be observed from table three all these variables added up gives Brazil an estimated score of 1549 even though the true score have been 1540. five.one.4Gambia-
The only nation estimated exactly employing the new specification of football performance were Gambia ranked as number 126. The reason for this will likely be examined briefly below- Gambia has an extremely low gnp of only 248 per capita per year. According our model that will only add one.six points towards the FIFA score. The average temperature in Gambia is 28 degrees which is twice as much as the optimal temperature and will decrease the score Gambia gains in our estimation by 128 points as temperature is non linearly and negatively correlated with the dependent variable. Gambia isnt a Latin country so population does not matter in this case. Gambia has been an affiliate of FIFA for 39 years and this fact will add 145 points to the total score which counts for nearly all the points this nation receives from the regression estimation in addition to the constant coefficient. Lastly Gambia does not have any players playing inside the top rated 5 leagues so nothing additional is added to the regression.
In other words the reason why Gambia isnt a very good football team is because the wealth of your country is extremely low the climate is just not ideal for football does not have any football culture and therefore the potential pool of football players total population does not affect the efficiency. Gambia does have some football tradition and an affiliate of FIFA because 1968 but no international starts does nothing to support increase the interest and quality of the national football group.
5.one.5Overall
To illustrate the overall fit of the model Figure 3 shows which nations are estimated within 1 standard deviation from the true value.
As figure 3 shows the model predicts pretty much the many countries in the Americas correctly. The model also performs adequately for Africa and South East Asia. Reversely the estimation does not seem to become a model of Europe because the FIFA ranking is at the moment nor the Middle East. This can be worth keeping in mind for more studies to try and pinpoint why the distribution the correct outcomes are as indicated above.

Figure 3- Globe map

The countries marked in red were estimated one particular standard deviation or less from the genuine score
Figure created on 5.2Robustness tests
In this section the regression is tested for robustness making use of the old FIFA ranking method and an alternative ranking method called the ELO ranking as dependent variables five.2.1May 2006 information
Probably the most recent information applying the previous FIFA ranking method is from May perhaps 2006 as there is no publication of FIFA ranking for the duration of World Cups.
The regression result employing May well 2006 information as dependent variable is in table 7.
By eyeballing the data we observe that GNP and GNP2 are no longer significant at a 10 degree of significance. Compared to the 2007 information it is obvious from the adjusted R2 that the fit with the model has decreased. Most of your coefficients have also decreased but as mentioned above this is as expected. However for the prime 20 teams we observe that 75 from the nations are within one standard deviation from the true score table eight-
Table 7- Regression benefits using -old ranking method- standard error in parenthesis

Pre WC 2006 ranking method
Sample size- 179
Variable
Est. May well 2006 information
t-value

Constant
282.0300
six.759077
41.72612
GNP
0.0039
1.369397
0.002854
GNP2
-7.7110 -7
-1.3570
five.69E-08
TEMP-14two
-0.386534
-2.0739
0.186386
POP x LATIN
9786.713
2.0333
4813.194
HIST
two.982407
five.0444
0.591233
REP
120.6496
two.321334
51.97427
ELITE
one.784541
two.471446
0.722064
Adjusted R2
0.39759

Note- and denote significance at ten 5 and 1 respectively

Table 8- Making use of May well 2006 information to calculate ranking points

May FIFA ranking
Nation
Points
Estimated
One particular sd from the true value1

1
Brazil
827
913
1

two
Czech Republic
772
682
one

three
Netherlands
768
681
one

4
Mexico
758
705
one

five
Spain
756
879
1

6
United States
756
615
1

7
Portugal
750
665
1

8
France
749
810
one

9
Argentina
746
777
one

ten
England
741
801
one

11
Denmark
736
663
one

12
Nigeria
736
428
0

13
Italy
728
742
1

14
Turkey
726
570
0

15
Cameroon
722
437
0

16
Sweden
709
646
1

17
Egypt
708
526
0

18
Japan
705
473
0

19
Germany
696
806
one

20
Greece
694
601
one

SD- 144.37

Again Spain is overestimated however this time the final results are within one SD from the actual value. Additional inspection of the table above shows that the only countries that were not correctly estimated were the ones from Asia and Africa. This could be the opposite result compared to the ones we identified making use of the new tanking formula but as two with the variables were not significant in this latter regression we pick out not to put too much emphasise on these final benefits.
5.2.2ELO ranking
There is also an alternative ranking method called the ELO ranking published regularly on the Internet- This ranking is primarily based on the ELO rating system developed initially for the ranking of Chess players. The ELO ranking for international football teams bases its outcomes around the following criteria- Status of match number of goals scored results of every single match multiplied with the expected result of your match to capture the opposition strength. The ranking method takes into account all matches played by every nation from 1872 onwards when calculating the last score. The advantage of this compared to by far the most recent four years as would be the case in the FIFA ranking is that you will get the notion of football culture incorporated into the model. The problem is that the final results are not likely to change much as time passes because the time span is extremely large.

The regression is nevertheless compared to this ranking to test the robustness from the new model and below is a brief outline of our results-

Running precisely the same regression as above only substituting the dependent variable for the ELO ranking points are presented in table 9.
Table 9- Regression benefits ELO points standard error in parenthesis
ELO ranking method

Sample size- 179
Variable
Est. ELO data
t-value

Constant
1099.564
19.55087
56.2412
GNP
0.0067
one.746035
0.0038
GNP2
-1.6810 -7
-2.186437
7.6610-8
TEMP-142
-0.4088
-1.627330
0.2512
POP x LATIN
9143.235
1.409353
6487.539
HIST
5.0920
six.389771
0.7969
REP
272.6909
three.892564
70.0543
ELITE
4.943897
5.079810
0.9732
Adjusted R2
0.5303

Note- and denote significance at ten 5 and 1 respectively

Observe that the signs stay as expected but a few of the variables have become less significant particularly the POP x LATIN and TEMP coefficients which now are the two insignificant. The fit with the model has also decreased. The Durbin Watson score is 1.99 indication that there is no severe problem of serial correlation in the model.

Further tests on how the nations inside the sample rank employing the new coefficients follows below.

Table ten monitors the true ranking according to the ELO system as of 9th of March 2007 compared to the estimated points obtained working with the regression above. As we observe within the top 20 ranking 80 in the nations are predicted within 1 standard deviation form the actual score which is far far better in comparison to the final results making use of the 2007 FIFA points regression. Overall 125 nations have been estimated within one particular standard deviation form the true value which was exactly exactly the same value as within the FIFA ranking regression. Yet again we can observe that Spain is estimated with the highest score of the many nations which was exactly the same outcome because the two other models predicted. This shows that the model is rather robust ignoring the fact that two of the variables inside the new regression became insignificant.

Table 10- Utilizing ELO ranking information to calculate ranking points
ELO ranking
Country
Points
Estimated
A single SD from the genuine value1

one
Brazil
2034
2099
one

2
France
2021
2150
1

3
Italy
1992
2090
one

four
Argentina
1971
2058
1

5
Netherlands
1965
1801
1

six
Germany
1955
2145
one

7
England
1917
2127
0

eight
Portugal
1890
1809
one

9
Croatia
1873
1598
0

10
Spain
1872
2236
0

11
Czech Republic
1866
1870
1

12
Denmark
1856
1766
1

13
Russia
1839
1605
0

14
Uruguay
1819
1806
1

15
Switzerland
1811
1695
one

16
Sweden
1800
1713
one

17
Greece
1795
1673
1

18
Turkey
1795
1613
one

19
Mexico
1793
1711
one

20
United States
1781
1638
1

SD- 207
five.3Suggestions for more testing
Another interesting comparison would be to test the model making use of bookmakers odds of winning the Globe Cup because the dependent variable linking the method to that used on international stock markets. Odds are supposed to capture all of the current information about group strength at any moment in time inside the exact same fashion as the stock market predicts the value of a company at any given time by means of share prices and need to therefore always change their odds depending of every single teams strength. One particular potential problem could be that a lot of weaker football countries are given the identical or no odds of winning the World Cup because the possibility of them even qualifying is close to zero.

An explanatory variable that would be interesting to test would be to include information on how several foreigners play inside the top rated league of every single nation. A lot of pundits have blamed the big favourites failure to perform within the large tournaments among them England and Spain on the substantial fraction of foreign players playing in their national best division. There happen to be claims that because from the large import of foreign players younger national players are not given a chance to prove themselves at the highest degree resulting in lower efficiency from the national group as time passes. To my knowledge there has been no testing done on in this field which must make it even much more attractive for future studies.

If we again investigate figure 2 which is a planet map that indicates how well the model performs world wide observations are made that the Middle East is poorly estimated. Additional studies might want to keep this in mind and try to pinpoint why this will be the case. One particular suggestion if further investigation from the Middle East is conducted is that cricket can be seen as a competitor to football as the most well known national sport in this area. As cricket shares many on the characteristics of football as a -sport for the masses- this kind of as low cost and high availability this could be an interesting notion to try and implement in any more studies.
6Conclusion
This paper investigates which nation certain things determining a countrys international football overall performance. It has been demonstrated that per capita wealth is important however beyond a certain level of earnings increase football performances actually declines. We have also shown how temperature affects football efficiency within the sense that any deviation from the optimal temperature of 14C harms a countrys football performance.

Length of FIFA membership does also play a part when attempting to estimate any given countrys FIFA score. The purpose for this is that a longer affiliation with the FIFA indicates a larger level of football culture which again gives a increased probability of additional active international football players. That quality of domestic players measured by their demand from the world best clubs very influences a countrys performance in the FIFA ranking which ought to come as no surprise. Subsequent we have also tested the impact a countrys population plays on international football functionality and located that population size only matters if the nation in question has a strong football culture indicated by a Latin population. No final results have been found to support any significant health variables.

Overall the model performs well even if youll find some concerns above the low estimation ability for the higher ranked countries.

7References

Baimbridge Mark Samuel Cameron and Peter Dawson 1996 -Satellite television plus the demand for football- A whole new ball game- Scottish Journal of Political Economy 43-three 317-333.

Ball P. 2003 -Morbo The story of Spanish Football- WSC Books Limited

Bird Peter 1982 -The demand for league football- Applied Economics 14- 637-649.

Carmichael Fiona and Dennis Thomas 1993 -Bargaining in the transfer marketplace- theory and evidence- Applied Economics 25-12 1467-76.

Carmichael Fiona and Dennis Thomas 2000 -Institutional Responses to Uncertainty- Proof from the Transfer Marketplace- Financial Issues five-one 1-19.

Carmichael Fiona Dennis Thomas and Robert Ward 2001 -Production and efficiency in Association Football- Journal of Sports Economics two-three 228-243.

Daunchey Hugh and Hare Geoff 1999 -France along with the 1998 Planet Cup- Frank Cass publishers

Dilger Alexander 2001 -The Ericson Case- Journal of Sports Economics 2-two 194-200.

Dobson Stephen and Goodard John 1999 -The economics of football- Cambridge University Press

Feess Eberhard and Gerd Mhlheusser 2002 -Economic Consequences of Transfer Fee Regulations in European Football- European Journal of Law and Economics 13 221-237.

Filmer D. and Pritchett L 1999 -The impact on public spending on Health does money matter- Social science and Medicine Harvard University Press

Filmer D. Hammer J. and Pritchett L. 1997 -Health Policy in Poor Nations- Weak Links within the Chain- Washington DC- World Bank Policy Study Working Paper No.1874.

Forrest David Robert Simmons and Patrick Feehan 2002 -A spatial cross-sectional analysis with the elasticity of demand for soccer- Scottish Journal of Political Economy 49-three 336-355.

Gelade G. 2006 -Determinants of Results in International Football- Cass Business School

Guedes J. C. and F. S. Machado 2002 -Changing rewards in contests- Has the three-point rule brought much more offence to soccer Empirical Economics 27 607-630.

Hart R. J. Hutton and T. Sharot 1975 -A statistical analysis of association football attendances- Journal of your Royal Statistical Society Series C 24 17-27.

Haugen Kjetil and Arild Hervik 2002 -Estimating the value of the Premier League or the worlds most profitable investment project- Applied Economics Letters 9 117-120.

Hoffman R. Lee C.G. and Ramasamy B. 2002a -Socio Financial Determinants of International Soccer Performance- Journal of Applied Economics Vol five No. 2- 253-272

Hoffman R. Lee C.G. and Ramasamy B. 2002b -Public Policy and Olympic Good results- Applied Economics Letters 9- 545-548

Johnson D. K. N. and Ali A. 2004 -A Tale of Two Seasons- Participation and Medal Counts at the Summer season and Winter Olympic Games- Social Science Quarterly 85 974 - 993.

Johnson Daniel K. N. and Ali Ayfer 2000 -Coming to play or coming to win- Perticipation and success within the Olympic Games- Wellesley College Working Paper 2000-10.

Kuper Gerard and Elmer Sterken. -The Olympic Winter Games- Participation and Performance.- University of Groningen mimeo December 2001.

Macmillan P. and Smith I. 2006 -Explaining International Soccer Ratings- University of St. Andrews

Neale W. 1964 -The peculiar economics of skilled sports- Quarterly Journal of Economics 78 1-14

Palomino Frederic Luca Rigotti and Aldo Rustichini 1999 -Skill strategy and passion- an empirical analysis of soccer- operating paper.

Reilly B. and R. Witt 1995 -English League transfer prices- is there a racial dimension- Applied Economics Letters 2 220-222.

Robine JM Romieu I Cambois E. 1999 -Health expectancy indicators.-Bulletin of the Globe Health Organization 77two- 181-18

Skelton C A Passion for Football- Dominant Masculinities and Primary Schooling - Sport Education and Society 2000 - Taylor Francis

Sloane P. 1969 -The labour market place in skilled football- British Journal of Industrial Relations 7 181-199.

Speight Alan and Dennis Thomas 1997a -Arbitrator decision-making in the transfer market place- an empirical analysis- Scottish Journal of Political Economy 44 198-215.

Studenmund A. H. 2001 Utilizing econometrics A Practical Guide 4th ed. Addison Wesley Longman Inc.pp 195

Szymanski Stefan 2001 -Income inequality competitive balance along with the attractiveness of team sports- some proof and a natural experiment from English soccer- Financial Journal 111 F69-F84.

Torgler B 2004. -The Determinants of Womens International Soccer Performance- CREMA Operating Paper 2004-19 Basel Centre for Reseach in Economics Management as well as the Arts

i Football in this context is what the rest in the English speaking globe knows as soccer Moore 2006. Soccer is an abbreviation for Association Football. The rest of this paper will use the word football when referring to Association Football.

ii Union of European Football Assosiations

iii CONfederacin sudaMEricana de FtBOL South American Football Confederation

iv Where 2007 information had been not available we included quite possibly the most recent information

v Durbin Watson 5 critical values table- Dobson telescop I have traveled over the world. I have seen the Canadian and American Rockies this Andes the Alps and the Highlands of Scotland but for simple splendor Cape Breton outrivals them all-
Alexander Graham Bell
Cape Breton Islandisa rugged littleisland approximately 175 km very long by 135 kilometer at its biggest point and is linked to the mainland of Nova Scotia from the Canso Causway. Thereisabout 150000 of us Cape Bretoners livingon this island paradise poker.
The island is one of the most breathtaking places in the world.Global travel magazinessuch as Conde NasteTravel related and National Regional Traveller have votedCape Breton Islandthe Third most beautiful Island on earth and 1 in the Mark vii US and Nova scotia.

Tagged with:

Filed under: Astronomy News