Black protests about policing in 1990s and 2000s: Time trends

There were many US Black protests about policing before the Black Lives Matter protests after 2014. Drawing on archives of 950 articles including 310 from three mainstream newswire services and 640 from 37 Black newspapers as sources of information, we identified about 1,059 Black-relevant protest events about policing in the United States between 1994 and 2010. These 1,059 events were about 279 different specific issues in more than one hundred cities, including 103 separate cases of police killings (37% of all specific issues) and another 64 cases of non-fatal claims of police brutality (23% of all specific issues) in more than 70 different cities. Other kinds of policing issues were non-fatal instances of police brutality, protests about police brutality in general not tied to a specific case, other complaints about discriminatory policing, and miscellaneous policing issues. There were, on average, many more protest events about fatal policing issues than nonfatal policing issues. The number of protests about an issue increases the rate of news coverage of events about that issue.

All the graphs show a general rise in protests in the 1990s with a peak 1999 to 2001, then a drop off after the terror attacks of September 11, 2001, a low point in 2004, and then a rise in activity toward the end of the 2000s.

Our concept of “specific issue cluster” is all the events that were about the same specific thing such as the New York police killing of Amadou Diallo or the annual October 22 National Day of Protest Against Police Brutality. Most specific issue clusters include only one or two events, but large ones such as the Diallo or National Day protests include large numbers of events.

For these plots, events are distinguished by the types of news sources that covered them. (1) Black newspapers covering events in the state where the event occurred (BN instate), (2) Some Black newspaper coverage from sources outside the state with or without in-state coverage as well but no newswire coverage (BN outstate), (3) Both Black newspaper (instate or outstate) and mainstream newswire coverage (NW & BN), (4) Newswire coverage only (NW only).

The first table shows the proportion of fatal issues by coverage type for articles, events, estimated minimum number of participants, and clusters. Black newspaper articles are subdivided by those that discuss at least one event occurring in the state where the newspaper is published and those that are only discussing events in other places. No article can be both Newswire and Black news but events and clusters that are covered by more than one article may be covered by both Black newspapers and newswires. The table shows that the proportion of newswire articles and of events and participants in events covered by newswires that involved fatal incidents is much higher than for events covered only in Black newspapers. The cluster column reveals that specific issue clusters covered by both newswire and Black newspapers are especially likely to involve fatalities and for issue clusters covered by only one source type, newswires cover more issues involving fatalities than Black newspapers and non-local (outstate) coverage involves more fatalities than local (state) Black newspaper coverage. The higher fraction of participants for BN state than BN outstate is because there is a great deal of outstate Black newspaper coverage of events about the annual October 22 day against police violence. In 1999-2000, in the wake of the killing of Amadou Diallo in New York and Tyisha Miller in Riverside, California, and other cases, Black activists held many events and sought to build momentum for a movement against police brutality in general. These “brutality in general” events got much more coverage in Black newspapers than in newswires.

Proportion About Policing Issues Involving Fatalities

	Articles	Events	Participants	Clusters
BN state	59	53	48	27
BN outstate	74	56	27	38
NW+BN	.	86	84	72
Newswire	88	82	77	47
Total	73	69	59	46

Specific Issue Clusters

The first time plot shows is the number of specific issue clusters, the things people were protesting about, by year. Most specific issues garnered only one or two protests, but there were a lot of protests about a few issues. The event counts will be the next graph. Classification of coverage type is across all the events in the cluster. So, for example, if some events in a cluster got local Black newspaper coverage only (the most common type of event coverage) while one event also got newswire coverage, the cluster is categorized is getting both newswire and Black newspaper coverage. What I see in the graph: (1) The number of specific issue clusters covered only in Black instate sources is relatively consistent over the years although there are low points in 2004 and 2009 and a high point in 2007. (2) The number of issues covered by non-local Black newspapers or both Black newspapers and newswires seems substantially lower after 2000. (3) The distribution of issues across years is generally more even than for events or participants in the graphs below.

Clusters about fatal issues

This graph and the next separate clusters by whether they involve fatalities or not. The scale is kept constant between the two graphs to help convey visually the fact that the majority of clusters are about non-fatal issues and those were primarily covered by Black newspapers reporting on events in the state where the newspaper is published.

Clusters about nonfatal issues

Events

The next graph is counts of events by coverage type. What I see i the graph: (1) Much more year-to-year variability than for clusters, which is due to the years in which the big clusters appear, especially 1999-2000 in New York around the killing of Amadou Diallo and 2001 in Cincinnati around the killing of Timothy Thomas. (2) Again the umber of events covered only by local Black newspapers varies less from year to year than the events covered by newswires or non-local Black newspapers. (3) There is a peak in events 1999-2001.

Events about fatal issues

Unlike clusters, a majority of events are about fatalities and the large clusters involving dozens of events produce large peaks in protests especially in 1999 to 2001.

Events about nonfatal policing issues

Number of Participants

In our project, we ask coders to provide estimates of the minimum and maximum number of participants in each event from all of the available contextual information in news articles. Here I present the graphs using the minimum participation estimates again by the type of news coverage the event got. The main thing I see in the graph is that events covered by both newswires and Black newspapers or by non-local Black newspapers account for most of the participants. And participation varies greatly with a few peak years 1997, 1999, 2000, and 2006 being much higher than other years.

Participants in events about fatal issues

The graph of participants about fatal issues is very similar to the total graph as the fatal issues draw more participants.

Participants in events about nonfatal incidents

Similar to prior graphs, participants in protests about non-fatal incidents are mostly covered in local Black newspapers. The main exception is 1997 which included the protests about New York police officers torturing Abner Louima in a police station bathroom often covered by both newswires and Black newspapers and the non-local coverage of protests about police brutality in general.

Articles

These plots show the number of articles by news source type. Any given article is either newswire or Black newspaper so the “NW & BN” category does not appear in these graphs. The main pattern is that newswire articles about police protests are especially common in 1999 to 2001 and all article types exhibit the 2004 drop.

Articles about fatal issues

The plot of articles about fatal issues looks very similar to the total plot because fatal issues both have more events and get more coverage.

Articles about nonfatal issues

The majority of articles about nonfatal issues appear in Black newspapers and there are fewer articles about nonfatal issues.

Tables

The tables below include the numbers underlying the graphs above. They are included primarily for accessibility for people using screen readers.

The first table is clusters.

Clusters

	Total				Fatal				Nonfatal
	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only
1994	6	4	2	5	0	1	1	3	6	3	1	2
1995	7	3	3	3	2	2	2	2	5	1	1	1
1996	9	3	3	1	3	0	2	0	6	3	1	1
1997	12	5	2	2	4	2	1	1	8	3	1	1
1998	17	4	1	0	6	2	1	0	11	2	0	0
1999	13	8	7	1	2	2	5	0	11	6	2	1
2000	13	2	6	4	5	0	4	2	8	2	2	2
2001	15	2	2	1	4	1	2	0	11	1	0	1
2002	9	2	3	1	3	0	2	1	6	2	1	0
2003	12	2	1	1	4	1	1	1	8	1	0	0
2004	3	0	0	1	0	0	0	1	3	0	0	0
2005	4	3	2	4	2	1	1	0	2	2	1	4
2006	8	1	2	2	1	1	2	1	7	0	0	1
2007	17	4	0	1	5	1	0	0	12	3	0	1
2008	6	2	1	1	1	1	1	1	5	1	0	0
2009	2	1	1	1	0	1	1	0	2	0	0	1
2010	7	7	0	1	1	4	0	1	6	3	0	0
Total	160	53	36	30	43	20	26	14	117	33	10	16

The second table is events

Events

	Total				Fatal				Nonfatal
	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only
1994	19	7	2	8	6	1	0	6	13	6	2	2
1995	30	11	5	4	21	10	4	3	9	1	1	1
1996	30	24	3	5	22	5	3	4	8	19	0	1
1997	38	15	7	9	16	8	1	6	22	7	6	3
1998	29	7	0	1	11	2	0	1	18	5	0	0
1999	54	39	25	17	32	26	24	14	22	13	1	3
2000	73	22	31	43	46	12	27	34	27	10	4	9
2001	35	9	26	39	19	8	26	34	16	1	0	5
2002	32	16	9	6	12	7	7	6	20	9	2	0
2003	22	5	3	6	9	3	2	6	13	2	1	0
2004	6	2	0	1	2	0	0	1	4	2	0	0
2005	7	6	3	8	3	2	0	3	4	4	3	5
2006	42	4	3	6	23	3	3	5	19	1	0	1
2007	47	13	2	3	24	4	2	2	23	9	0	1
2008	36	11	10	9	23	10	10	9	13	1	0	0
2009	8	7	5	6	3	6	5	5	5	1	0	1
2010	21	19	4	4	9	15	4	4	12	4	0	0
Total	529	217	138	175	281	122	118	143	248	95	20	32

The third table is participants

Total Participants, minimum

	Total				Fatal				Nonfatal
	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only	BN instate	BN outstate	NW & BN	NW only
1994	1,339	3,545	51	544	785	20		294	554	3,525	51	250
1995	15,523	1,455	162	165	14,781	1,445	62	140	742	10	100	25
1996	4,210	2,842	320	511	2,885	561	320	511	1,325	2,281		0
1997	5,361	22,205	14,880	2,510	935	1,480	500	1,145	4,426	20,725	14,380	1,365
1998	2,684	454		80	366	4		80	2,318	450
1999	7,089	6,781	19,724	7,039	2,666	4,327	19,624	6,837	4,423	2,454	100	202
2000	22,090	3,639	26,889	15,363	6,979	1,976	22,289	14,323	15,111	1,663	4,600	1,040
2001	3,694	457	2,890	7,088	1,365	437	2,890	1,985	2,329	20		5,103
2002	4,913	3,860	603	531	3,800	334	173	531	1,113	3,526	430
2003	2,783	256	401	273	2,356	201	201	273	427	55	200
2004	705	0		200	35			200	670	0
2005	1,326	243	165	401	1,150	101		215	176	142	165	186
2006	18,535	112	51,400	1,075	6,865	102	51,400	975	11,670	10		100
2007	8,092	2,439	640	35	2,009	1,050	640	30	6,083	1,389		5
2008	7,747	86	2,000	150	2,327	86	2,000	150	5,420	0
2009	1,648	125	1,800	346	121	125	1,800	341	1,527	0		5
2010	4,900	1,662	1,710	122	4,310	1,495	1,710	122	590	167
Total	112,639	50,161	123,635	36,433	53,735	13,744	103,609	28,152	58,904	36,417	20,026	8,281

The last table is articles

Participants (minimum)

	Total			Fatal			Nonfatal
	NW	BN outstate	BN instate	NW	BN outstate	BN instate	NW	BN outstate	BN instate
1994	7	5	9	3	0	2	4	5	7
1995	3	5	14	3	4	9	0	1	5
1996	5	9	19	5	7	14	0	2	5
1997	16	18	28	8	9	14	8	9	14
1998	1	1	24	1	0	8	0	1	16
1999	47	22	72	45	15	55	2	7	17
2000	58	17	55	49	15	37	9	2	18
2001	75	27	30	72	24	12	3	3	18
2002	9	16	35	8	11	14	1	5	21
2003	13	7	22	11	5	11	2	2	11
2004	3	1	5	3	0	2	0	1	3
2005	7	5	6	2	3	2	5	2	4
2006	10	5	35	9	5	25	1	0	10
2007	10	13	34	8	7	23	2	6	11
2008	18	20	34	18	18	23	0	2	11
2009	11	5	13	10	5	7	1	0	6
2010	17	9	20	17	8	10	0	1	10
Total	310	185	455	272	136	268	38	49	187