Professional Sports and Domestic Violence
Tables and Graphs for UC Berkeley Presentation February 23, 2006
David Card University of California Berkeley Gordon Dahl University of Rochester
Table 1. Top 50 Television Programs for 2004. Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 21 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 40 42 43 44 45 46 47 48 49 50
Program Super Bowl XXXVIII Academy Awards Friends Clipshow Special MLB ALCS Game 7 Summer Olympics Thu Prime 1 Summer Olympics Tue Prime 1 World Series Game 4 Survivor: All-Stars Premiere Summer Olympics Wed Prime 1 Countdown To Oscars 2004 Golden Globe Awards Summer Olympics Mon Prime 1 Summer Olympics Mon Prime 2 Frasier Finale AFC Divisional Playoff-Sa American Idol-Premiere Special World Series Game 2 Summer Olympics Sun Prime 2 Grammy Awards Summer Olympics Tue Prime 2 World Series Game 3 MLB ALCS Game 6 Summer Olympics Sun Prime 1 AFC/NFC Playoff GM2 Summer Olympics Wed Prime 2 American Idol-Tuesday Sugar Bowl Pre-Game Summer Olympics Opening Ceremony Summer Olympics Fri Prime 1 American Idol-Wednesday Amer Idol-Wed Perf Show Summer Olympics Thu Prime 2 Desperate Housewives NBA Finals-GM 5 World Series Game 1 CSI Summer Olympics Sat Prime 2 CSI – Thanksgiving Friends 1/22 Friends 4/29 Survivor All-Stars Reunion MLB NLCS Game 7 Apprentice Survivor All-Stars Finale Friends 4/22 NBA Finals-GM 4 Law And Order 9/22 Summer Olympics Fri Prime 2 Survivor: All-Stars American Idol-Special
Network CBS ABC NBC FOX NBC NBC FOX CBS NBC ABC NBC NBC NBC NBC CBS FOX FOX NBC CBS NBC FOX FOX NBC ABC NBC FOX ABC NBC NBC FOX FOX NBC ABC ABC FOX CBS NBC CBS NBC NBC CBS FOX NBC CBS NBC ABC NBC NBC CBS FOX
Total US Household Rating 41.43 26.03 22.01 19.41 19.28 18.26 18.18 17.91 17.29 17.28 16.93 16.61 16.39 16.32 15.93 15.88 15.87 15.78 15.69 15.66 15.65 15.65 15.39 15.36 15.30 14.91 14.80 14.56 14.38 14.06 13.88 13.81 13.79 13.78 13.71 13.59 13.56 13.46 13.12 13.08 13.08 13.04 13.01 13.00 12.94 12.68 12.52 12.47 12.41 12.21
Source: Nielsen Media Research, Galaxy Explorer, 1/4/04-12/31/04. The total number of television households in the United States in 2004 was estimated at 108,400,000. Each national rating point represents one percent of the total, or 1,084,000 television households.
Table 2. Major NIBRS Reporting States as of 2002. State Colorado Connecticut Delaware Idaho Iowa Kansas Massachusetts Michigan Nebraska New Hampshire North Dakota Ohio Oregon South Carolina South Dakota Tennessee Texas Utah Vermont Virginia West Virginia
Date Testing Begun
Certification Date
Number of Agencies
Percent of State Pop.
Percent of State Crime
5-1991 12-1996 6-1991 12-1990 5-1991 3-1995 8-1990 2-1994 5-1991 12-1993 1-1989 12-1992 2-1994 11-1989 7-1995 11-1993 7-1992 5-1990 10-1992 10-1992 7-1994
11-1997 7-1999 7-2001 7-1992 8-1992 2-2001 8-1995 2-1996 2-1997 not certified 2-1991 1-1999 not certified 1-1991 2-2001 7-1998 7-1998 4-1994 4-1994 11-1994 9-1998
193 44 67 138 243 354 194 755 16 210 106 128 50 442 73 518 59 80 57 416 363
59% 30% 100% 100% 100% 85% 50% 81% 14% 74% 87% 20% 21% 100% 39% 100% 9% 60% 83% 100% 100%
51% 33% 100% 100% 100% 51% 45% 70% 9% 66% 77% 22% 20% 100% 24% 100% 9% 65% 91% 100% 100%
Notes: States in bold can be linked to NFL football teams. NIBRS accepts crime data submitted prior to certification and includes it in the official database once the quality of the reporting agency’s data is verified. Testing begins several years before data is accepted. States with a very low number of reporting agencies are excluded from the table. The number of reporting agencies within states has generally increased over time.
Table 3. Incidence of Domestic Violence by NIBRS Reporting Agency Size, 1995-2002. Rates per 100,000 Spousal Assault
Family Assault
Spousal Assault, with Alcohol
Fraction of Agencies
Population Weighted Fraction
pop ≤ 20,000 20,000 < pop ≤ 50,000 50,00 < pop ≤ 100,000 100,000 < pop
0.73 0.87 1.08 1.28
1.12 1.13 1.69 1.94
0.54 0.23 0.25 0.24
0.81 0.13 0.04 0.02
0.32 0.27 0.19 0.22
All population sizes
0.96
1.47
0.26
1.00
1.00
Population
Notes: Population refers to the city or county population of the NIBRS reporting agency. The rates broken down by population size are unweighted; the rates for all population sizes are weighted.
Table 4. NFL Football Teams Matched to NIBRS Agencies.
New England Patriots (MA) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
1995
AFC East 1996 1997
1998
1999
2000
2001
2002
6-10 -70 1.2
11-5 SB 94 1.7
10-6 DP 111 1.9
9-7 WC 125 2.2
8-8 -143 2.4
5-11 -184 3.2
11-5 SB 199 3.3
9-7 -215 3.8
1995
AFC North 1996 1997
1998
1999
2000
2001
2002
3-13 -9 0.7
4-12 -25 1.0
4-12 -38 1.3
6-10 -55 1.7
2-14 -68 1.9
2-14 -44 0.7
3-13 -59 0.9
7-9 -92 1.5
9-7 WC 129 2.0
1998
1999
2000
2001
2002
8-8 -174 2.1
13-3 SB 290 3.6
13-3 DP 342 5.1
7-9 -343 5.1
11-5 CP 343 5.2
AFC West 1996 1997
1998
1999
2000
2001
2002
12-4 SB 134 2.1
14-2 SB 133 2.2
6-10 -129 2.2
11-5 WC 126 2.3
8-8 -124 2.3
9-7 -116 2.2
7-9 -300 2.3
6-10 -311 2.3
8-8 -303 2.3
Cincinnati Bengals (OH) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions) Cleveland Browns (OH) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
1995
AFC South 1996 1997
Tennessee Titans (TN) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
1995 Denver Broncos (CO) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions) Kansas City Chiefs (KS) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
Table 4, continued. NFL Football Teams Matched to NIBRS Agencies.
1995
NFC East 1996 1997
Dallas Cowboys (TX) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions) Washington Redskins (VA) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
Detroit Lions (MI) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
Carolina Panthers (SC) Regular Season Record Postseason Play Number of reporting agencies Population coverage (millions)
6-10 -15 0.4
9-7 -66 0.9
8-7 -101 1.6
1998
1999
2000
2001
2002
10-6 WC 38 1.1
8-8 WC 47 1.3
5-11 -50 1.6
5-11 -53 1.8
5-11 -56 2.1
6-10 -156 2.2
10-6 DP 173 3.1
8-8 -181 4.6
8-8 -195 4.8
7-9 -195 4.8
1995
NFC North 1996 1997
1998
1999
2000
2001
2002
10-6 WC 259 2.8
5-11 -378 5.0
5-11 -439 6.5
8-8 WC 460 7.0
9-7 -475 7.8
2-14 -488 8.2
3-13 -495 8.4
1995
NFC South 1996 1997
1998
1999
2000
2001
2002
7-9 -211 3.7
12-4 CP 221 3.7
4-12 -223 3.8
8-8 -227 3.9
7-9 -227 4.0
1-15 -219 4.0
7-9 -228 4.1
9-7 WC 414 6.1
7-9 -224 3.8
Notes: “Postseason Play” indicates whether the team advanced to the playoffs; “WC” stands for wild card playoffs, “DP” stands for divisional playoffs, “CP” stands for championship playoffs, and “SB” stands for Super Bowl. For the two Ohio teams, NIBRS agencies are assigned based on whether the center of the reporting agency’s county is closer to Cincinnati or Cleveland. All agencies in Virginia are assigned to the Washington Redskins. Reporting agencies can be either cities or counties.
Table 5. NFL Football Games and Domestic Violence. Fixed-Effects Poisson Regressions
(Team Play=1)*Game Day (Team Play=0)*Game Day (Team Play=1)*Playoff (Team Play=0)*Playoff (Team Play=1)*Super Bowl (Team Play=0)*Super Bowl Martin Luther King Day President’s Day Memorial Day 4th of July Labor Day Columbus Day Veteran’s Day Thanksgiving Christmas Eve (Dec. 24) Christmas Day New Year’s Eve (Jan. 31) New Year’s Day
Number of observations Number of groups Log Likelihood
Spousal Assault
Family Assault
Spousal Assault, Alcohol Involved
.0105 (.0071) -.0087 (.0065) .0150 (.0319) -.0070 (.0118) .2666** (.0736) .1083** (.0236) .0594** (.0304) .0002 (.0311) .3080** (.0238) .3154** (.0232) .3377** (.0254) -.0563* (.0315) -.0089 (.0313) .1512** (.0291) .2595** (.0265) .1086** (.0283) .5495** (.0233) .2886** (.0257)
.0211** (.0058) -.0018 (.0052) .0202 (.0265) .0040 (.0095) .2139** (.0633) .0942** (.0192) .0453** (.0205) .0378* (.0207) .2676* (.0160) .2929** (.0158) .3211** (.0168) -.0110 (.0204) -.0327 (.0210) .1719** (.0192) .2403** (.0178) .1140** (.0189) .5502** (.0156) .2805** (.0176)
-.0119 (.0161) -.0341** (.0146) .0785 (.0640) -.0551** (.0251) .8660** (.1268) .3009** (.0494) -.0120 (.0632) .0667 (.0597) .5014** (.0424) .5115** (.0373) .5289** (.0453) -.0711 (.0632) -.0010 (.0628) .3044** (.0513) .0291 (.0459) .5697** (.0437) 1.019** (.0342) .3468** (.0456)
4,010,603 2,183 -1,270,304
4,034,705 2,207 -1,677,229
1,966,675 1,962 -427,596
**significant at the 10 percent level, *significant at the 5 percent level. Notes: Each observation is a daily count for the number of crimes committed in a NIBRS-agency’s jurisdiction (city or county). The “fixed effects” are the NIBRS agencies. Outcome variables are measured as counts per 100,000 population. “Team Play” indicates whether the corresponding city or county’s NFL team played in the football game, and “Game Day,” “Playoff,” and “Super Bowl” are indicators for a regular season, a playoff, or a Super Bowl game day. All regressions also include dummies for day of the week, month of the year, year, and a quadratic in population size. Standard errors reported in parentheses.
Table 6. Sunday Games, Monday Night Football, and Domestic Violence. Fixed-Effects Poisson Regressions Spousal Assault
Family Assault
Spousal Assault, Alcohol Involved
A. Sunday Football Games (Comparison Group is Tuesdays, Wednesdays, and Fridays) Sunday Game Bye Sunday (Team Play=1)*Playoff (Team Play=0)*Playoff (Team Play=1)*Super Bowl (Team Play=0)*Super Bowl
Number of observations Number of groups Log Likelihood
.0098 (.0081) .0063 (.0146) .0332 (.0410) -.0012 (.0161) .2783** (.0737) .1146** (.0239)
.0282** (.0065) .0035 (.0118) .0506 (.0336) .0205 (.0129) .2308** (.0634) .1050** (.0174)
-.0274 (.0183) -.0285 (.0341) .1282 (.0868) -.0775** (.0365) .8773** (.1272) .3006** (.0501)
2,257,783 2,130 -716,142
2,274,211 2,158 -947,477
1,097,804 1,840 -235,624
B. Monday Night Football Games (Comparison Group is Tuesdays and Wednesdays) MNF Game (Team Play=1)*MNF Game
Number of observations Number of groups Log Likelihood
.0021 (.0091) .1040** (.0252)
.0024 (.0072) .0783** (.0205)
.0024 (.0212) .1777** (.0559)
1,662,547 2,072 -497,642
1,689,935 2,123 -673,431
807,152 1,739 -150,749
**significant at the 10 percent level, *significant at the 5 percent level. Notes: “Sunday Game” and “MNF Game” are indicators for a Sunday or Monday Night Football game day. “Bye Sunday” is a dummy variable equaling one for Sundays during the regular season when city’s NFL team did not play in a football game. See notes to Table 5 for other definitions and other included covariates (all regressions include dummies for the holidays appearing in Table 5 and all of the other variables listed in the footnote to Table 5). Standard errors reported in parentheses.
Table 7. Fixed-Effects Poisson versus Fixed-Effects Negative Binomial Regressions for NIBRS Agencies which Report Consistently Throughout the Year. Spousal Assault
(Team Play=1)*Game Day (Team Play=0)*Game Day (Team Play=1)*Playoff (Team Play=0)*Playoff (Team Play=1)*Super Bowl (Team Play=0)*Super Bowl
Number of observations Number of groups Log Likelihood
Family Assault
Spousal Assault, Alcohol Involved F.E. F.E. Poisson Neg. Bin.
F.E. Poisson
F.E. Neg. Bin.
F.E. Poisson
F.E. Neg. Bin.
.0051 (.0082) -.0066 (.0074) .0140 (.0379) -.0109 (.0138) .3177** (.0915) .1103** (.0275)
.0053 (.0087) -.0075 (.0079) .0151 (.0403) -.0110 (.0147) .3231** (.0964) .1133** (.0292)
.0163** (.0066) -.0019 (.0060) .0194 (.0314) .0014 (.0112) .2902** (.0778) .0885** (.0224)
.0189** (.0077) -.0010 (.0070) .0215 (.0365) .0038 (.0130) .2830** (.0903) .0939** (.0262)
-.0161 (.0191) -.0329* (.0173) .0747 (.0776) -.0737* (.0304) .8210** (.1609) .2878** (.0596)
-.0158 (.0202) -.0311* (.0183) .1012 (.0807) -.0702** (.0321) .8048** (.1931) .2860** (.0631)
776,666 520 -748,386
776,666 520 -747,087
776,666 520 -946,787
7766,666 520 -935,377
740,213 493 -276,385
740,213 493 -275,491
**significant at the 10 percent level, *significant at the 5 percent level. Notes: Sample restricted to NIBRS agencies which report crime data for at least 350 days out of the year; this sample is largely composed of agencies covering large populations. See notes to Table 5 for definitions and other included covariates. Standard errors reported in parentheses.
Table 8. NFL Football Games and Other Crimes. Fixed-Effects Poisson Regressions
(Team Play=1)*Game Day (Team Play=0)*Game Day (Team Play=1)*Playoff (Team Play=0)*Playoff (Team Play=1)*Super Bowl (Team Play=0)*Super Bowl
Number of observations Number of groups Log Likelihood
Gambling
Non-Family Assault
Crimes Involving Alcohol
Other Crimes (i.e., Burglary)
.3055** (.1356) -.2047* (.1154) -.4903 (1.0056) .4558** (.2285) -10.0515 (336.6285) .6931 (.4222)
-.0457** (.0098) -.0232** (.0080) .0889** (.0448) -.0402** (.0170) .1916* (.1075) -.0023 (.0352)
-.0442** (.0082) -.0429** (.0072) .0308 (.0333) -.0679** (.0125) .6045** (.0736) .2362** (.0263)
-.0067** (.0023) .0006 (.0019) -.0244** (.0112) -.0197** (.0041) -.1243** (.0299) -.0755** (.0090)
663,952 319 -13,048
4,073,447 2,248 -1,258,329
3,976,866 2,166 -1,323,330
4,113,401 2,301 -5,028,629
**significant at the 10 percent level, *significant at the 5 percent level. Notes: See notes to Table 5 for definitions and other included covariates. Standard errors reported in parentheses.
Table 9. Regular Season Game Characteristics, NIBRS-NFL Matched Sample 1995-2002. Mean
Std. Dev.
Minimum
Maximum
0.88 0.06 0.04
