Regional effects of professional sports franchises: causal evidence from four European football leagues

Figures & data

Figure 1. Relegation status by bins of size one.

Note: Average treatment rates in equally sized bins of 1 point are shown. The vertical line represents the difference to the first non-relegation rank.

Source: Author’s own illustration.

Table 1. Sample properties of pre-relegation characteristics.

	Regions with relegated teams			Regions with non-relegated teams
	N	Mean	SD	N	Mean	SD
Attendance (club level)	216	22.009	11.490	999	29.956	15.456
Population (NUTS-3 region)	218	884.1	609.9	984	1156.4	955.1
Active population (NUTS-3 region)	218	416.0	287.8	984	553.6	466.1
Sectoral regional employment (NUTS-3 region)	218	119.4	89.5	984	164.0	149.2
Total regional employment (NUTS-3 region)	218	405.1	282.0	984	561.7	501.2
Regional sectoral GVA (NUTS-3 region)	218	6268	5459	984	9782	10,501
Total regional GVA (NUTS-3 region)	218	24,241	19.588	984	37,550	38,893

Note: Mean attendance per game, population (overall and active) and employment in football-related sectors are calculated in thousands. Gross value added (GVA) in football-related sectors is calculated in millions of euros. N may vary for selected indicators because of data availability (e.g., Monaco). The two-group mean-comparison test (not reported) shows that the differences between regions with relegated teams and regions with non-relegated teams are statistically significant.

Source: Author’s own illustration.

Figure 2. Density distribution by difference in points to the first non-relegation rank.

Source: Author’s own illustration.

Table 2. Testing for regional differences around the cut-off before treatment.

Differences in outcomes in the year before relegation	Attendance	Population/size of region I	Active population/size of region II	Regional sectoral employment	Total regional employment	Regional sectoral GVA	Total regional GVA
Relegation	−1.461	87.778	24.737	−1.065	−8.231	−409.960	−2257.900
	(3.293)	(172.610)	(65.097)	(17.677)	(54.292)	(1094.700)	(3999.100)
N	1215	1202	1202	1202	1202	1202	1202
Observations (l|r)	178|1037	180|1022	180|1022	180|1022	180|1022	180|1022	180|1022
Effective number of observations (l|r)	84|332	95|374	95|374	95|374	107|374	107|374	95|374
Order local polynomial (p) (l|r)	1|1	1|1	1|1	1|1	1|1	1|1	1|1
BW local polynomial (h) (l|r)	5.8|5.8	6.1|6.1	6.7|6.7	6.6|6.6	7.8|7.8	7.1|7.1	6.6|6.6
BW bias (b) (l|r)	10.7|10.7	11.0|11.0	11.5|11.5	11.5|11.5	14.5|14.5	12.4|12.4	11.7|11.7
Rho (h/b) (l|r)	0.5|0.5	0.6|0.6	0.6|0.6	0.6|0.6	0.5|0.5	0.6|0.6	0.6|0.6

Notes: BW, bandwidth; GVA, gross value added.

Standard errors are shown in parentheses, *p < 0.1, **p < 0.05, ***p < 0.01.

Source: Author’s own calculation.

Figure 3. Regression discontinuity (RD) plots using club-level attendance data as an outcome.

Source: Author’s own illustration.

Table 3. Effects of relegation on the demand for live attendance in the stadium.

Outcome	Change attendance (absolute)	Change attendance (relative)
Relegation	−4.254***	−0.287***
	(1.472)	(0.049)
N	1292	1292
Observations (l|r)	188|1104	188|1104
Effective number of observations (l|r)	74|299	101|397
Order local polynomial (p) (l|r)	1|1	1|1
BW local polynomial (h) (l|r)	4.6|4.6	6.3|6.3
BW bias (b) (l|r)	8.6|8.6	10.4|10.4
Rho (h/b) (l|r)	0.5|0.5	0.6|0.6

Notes: Outcome is the change in club-level attendance between the season of relegation and the immediate next season. BW, bandwidth.

Standard errors are shown in parentheses, *p < 0.1, **p < 0.05, ***p < 0.01.

Source: Author’s own calculation.

Table 4. Estimation of the effects of relegation on football-related sectors in the region.

Outcome	Regional sectoral employment (NUTS-3)	Regional sectoral GVA (NUTS-3)
Relegation	−0.027**	−0.030**
	(0.013)	(0.014)
N	1202	1202
Observations (l|r)	180|1022	180|1022
Effective number of observations (l|r)	95|374	83|328
Order local polynomial (p) (l|r)	1|1	1|1
BW local polynomial (h) (l|r)	6.3|6.3	5.1|5.1
BW bias (b) (l|r)	11.2|11.2	9.1|9.1
Rho (h/b) (l|r)	0.6|0.6	0.6|0.6

Notes: Outcome is the mean growth of regional sectoral employment and gross value added (GVA) between t − 1 and t + 1, where t is the year of relegation. BW, bandwidth.

Standard errors are shown in parentheses, *p < 0.1, **p < 0.05, ***p < 0.01.

Source: Author’s own calculation.

Figure 4. Regression discontinuity (RD) plots using regional sectoral development as an outcome.

Source: Author’s own illustration.

Figure 5. Regression discontinuity (RD) plots using total regional development as an outcome.

Source: Author’s own illustration.

Table 5. Estimation of the effects of relegation on football-related sectors in the region.

Outcome	Overall regional employment (NUTS-3)	Overall regional GVA (NUTS-3)
Relegation	−0.013*	−0.006
	(0.007)	(0.009)
N	1202	1202
Observations (l|r)	180|1022	180|1022
Effective number of observations (l|r)	107|400	107|400
Order local polynomial (p) (l|r)	1|1	1|1
BW local polynomial (h) (l|r)	7.1|7.1	7.5|7.5
BW bias (b) (l|r)	13.8|13.8	13.0|13.0
Rho (h/b) (l|r)	0.5|0.5	0.6|0.6

Notes: Outcome is the mean growth of overall regional employment and gross value added (GVA) between t − 1 and t + 1, where t is the year of relegation. BW, bandwidth.

Standard errors are shown in parentheses, *p < 0.1, **p < 0.05, ***p < 0.01.

Source: Author’s own calculation.