The impact of the fifa world Cup 2010 on unemployment in South Africa Michiel Antoine Oosterbaan

0.107 
0.066 
(1.42) 
(2.41)* 
Trend 
-0.010 
-0.022 
(1.85) 
(6.13)** 
TrWC
i
-0.087 
0.007 
(2.86)** 
(0.64) 
TrPost
t
-0.019 
-0.002 
(1.40) 
(0.37) 
TrPostWC
i, t
0.245 
0.007 
(3.28)** 
(0.39) 
Constant 
-4.537 
-3.911 
-4.450 
(2.67)** 
(2.42)* 
(2.61)* 
Observations 
675 
675 
675 
R-squared 
0.79 
0.78 
0.79 
Robust t statistics in parentheses
* significant at 5%; ** significant at 1% 
Bold coefficients are for variable of interest to test for unemployment effects of the FIFA World Cup South 
Africa. 

---

20 
Table 4 shows over the period 2003 to 2011, that the higher the shares in the agriculture, 
hunting, forestry, logging and fisheries sector, the community service industry sector and the 
mining industry sector in 2003, the lower the level and growth rate of unemployment for the 
municipality in all three the equations. However the shares in the trade, hospitality industry 
and traffic sector and the manufacturing sector did not differ significantly from zero. Larger 
population in 2003 translates itself into a higher level and growth rate of unemployment for 
the municipality. Furthermore the negative significant value of the variables WC
i
and TrWC
i
, 
indicate that over the whole period of observation the host municipalities experience lower 
levels and growth rates of unemployment relative to non host municipalities. 
Column (1) of table 1 represents the first equation. The relevant variable PostWC
i, t
is not 
significant. This indicates that unemployment levels in the 9 host municipalities did not differ 
significantly within the observation period from those other municipalities in the aftermath of 
the FIFA World Cup relative to before. The significant negative value of the variable Post
t
shows for the whole sample a lower significant level of the unemployed after the FIFA World 
Cup compared to pre-FIFA World Cup period, a decrease of 8.7%.
The growth of the numbers unemployed through the FIFA World Cup, estimated by equation 
(2), is shown in column (2). The relevant variable TrPostWC
i, t
is significantly positive at a 
1% level. In other words, the host municipalities show in comparison with the non host 
municipalities a positive significant trend in the development of unemployment for the period 
after the FIFA World Cup relative to before, corresponding with 24.5%.
Equation (3) that takes into account both the changes in the levels and trends of 
unemployment corresponds to column (3). The variables of interest are TrPostWC
i, t
and 
PostWC
i, t
. The variable TrPostWC
i, t
does not differ significantly from zero, indicating that 
for the period after the FIFA World Cup, the hosting municipalities show in comparison with 
the non host municipalities no trend differing significantly from zero in the growth rate of 
unemployment. However the variable PostWC
i, t
is significant and positive. This indicates 
that, while controlling for the growth rate of unemployment, the levels of unemployment in 
the period after the FIFA World Cup relative to the period before have developed 
significantly more with 6.6 % in host municipalities compared to the levels of unemployment 
in the non host municipalities. Furthermore, the negative and significant value of the time 
trend variable indicates a 2.2% downward slope in unemployment over the entire sample. 

---

21 
The results show that if we take the whole period of observation into consideration the host 
municipalities are relative better off with lower levels and growth rates of unemployment. 
However, this changes in the aftermath of the FIFA World Cup. The levels and growth rates 
of unemployment increase compared to non host municipalities. Therefore, an effect of the 
FIFA World Cup on unemployment in the nine host municipalities can be demonstrated, 
although it is negative. 

---

22