(incorporating the International Ex-Fetus Association)

Manufacturing Concern:

Chapter Four: Specific Cases of Coverage

Jim Boyce

The last chapter revealed a difference in the amount and type of headline coverage given to male and female victims of violence. This raises some obvious questions which I address in this chapter by examining some specific aspects of the headlines: chronological trends in headline coverage, front-page and multi-page headlines, headlines from only those CNI categories in the sample with gender-neutral titles, headlines from 1984 and headlines from the CNI category Child Abuse.

Chronological Trends

I will begin by examining chronological trends in coverage of male and female victims of violence. To determine how and whether this coverage fluctuated, I tabulated the coverage by year and season. The following chart contains the results:

Number and Percentage of Articles by Season and Year

These numbers indicate a 120 per cent increase in coverage from 1989 to 1991, with the result being that 65 per cent of the coverage appeared in 1991 and 1992. Each season saw a two to three-fold increase in coverage during the four years.” Winter, spring and summer were proportionally similar (although they varied in particular years) while the fall contained nearly twice as many articles as any other season and 41.5 per cent overall. This fall coverage was most pronounced in 1990 and 1991 when it represented nearly half of the yearly coverage compared to one-third in 1989 and 1992.³⁴

Similar trends occur among literal and contextual references to gender and violence. The results are as follows:

Number of Articles by Season and Year 

	1989	1990	1991	1992	Total
	Women Men	Women Men	Women Men	Women Men	Women Men
Winter	18     00	16     00	12     00	39     00	85     00
Spring	17     00	26     03	31     03	25     01	99     07
Summer	07     00	13     00	31     00	33     00	84     00
Fall	41     01	71     00	90     02	55     05	257    08
	82     01	126     03	164     05	152   06	525     15

Conclusions about coverage of male victims are difficult to make because of the small sample size. For example, the 1992 increase in coverage can be attributed to one Canadian Press article which appeared in five of the newspapers. Thus, one-third of coverage of men, as noted above, is found on one day of our sample. In the case of women, coverage doubled between 1989 and 1991 before tapering off. These increases were 25 per cent smaller than those in overall coverage which slightly increased in 1992. Coverage in the winter, spring and summer was near parity while that in the fall was triple that of any other season. This concentration in the fall is due to contextual references, where fall headlines account for 78 per cent of the total (104 of 133) compared to fall literal headlines which follow general trends at 39 per cent (153 of 392).

Fall coverage comprised at least half of yearly coverage in 1989, 1990 and 1991 and one-third in 1992, and like overall coverage, it peaked as a percentage of yearly coverage in 1991 and 1992.

To test for a link between fall coverage and coverage of the Montreal murders and their anniversaries, I calculated the number of headlines published during the two weeks following the murders on December 6, 1989, and the week before and after their anniversaries in 1990, 1991 to 1992. 1 also calculated the number of headlines which were gender-specific and the number associated with the murders rather than to other issues related to violence. The following results show the number of headlines in this period and, in the case of “Total Headlines,” their percentage of yearly coverage:

	1989	1990	1991	1992	Total
Total Headlines	39/21.9	44/17.0	57/14.5	37/08.9	177/14.3
Gender-Specific Headlines	31	36	36	19	122
Headlines on Montreal Murders	29	27	25	17	98

These weeks represent about four per cent of the sample but more than 14 per cent of the headlines. Headlines are two-and-one-half times more prevalent than average fall coverage and five times higher than the average coverage of any other season. Percentage-wise, coverage declined during the four years, particularly from 1991 to 1992. 

Sixty per cent of these headlines (122) are gender-specific, nearly double the average, and most of these are contextual (84) and linked to the Montreal murders (98). The highest percentage of headlines emphasizing the Montreal murders occurs during 1989 and 1990. Other issues related to violence begin competing with this coverage in the last two years although references to the Montreal murders still represent half of the headlines in 1992.

These chronological observations necessitate a few words of caution. We are relying on the selection policies and practices of CNI. If articles are selected by CNI based on “significance,” it is possible that those concerning the anniversaries of the Montreal murders, or other events, are considered less significant in some years than others, particularly in later years when the total amount of coverage is higher. In addition, it is beyond the scope of this thesis to determine the degree to which CNI’s selection represents overall coverage of violence and gender or whether this degree fluctuates between years. Finally, changes in the number of connotative headlines in any given year, as well as gender-specific events I have not considered to be contextual, may have an effect. Even if there were moderate distortions in the selection of headlines, however, it seems safe to conclude that coverage of gender and violence increased between 1989 and 1992, that there is significantly more coverage in the fall than in other seasons and that this fall coverage was particularly high in 1990 and 1991 and in the weeks surrounding the Montreal murders and their anniversaries.

Front-Page and Multi-Page Headlines

Another question that might be asked is whether the trends we have found characterize headlines appearing on front pages, the most visible part of the newspaper. Since many of these front-page articles continue on other pages of the newspaper, studying them allows us a chance to examine whether the headlines accompanying these continued portions are consistent with our findings.

One hundred and seventy-five headlines from the sample appear the front pages, including 21 in 1989, 47 in 1990, a peak of 70 in 1991 and 37 in 1992. They fall into the following groups (of these, one literal headline from 1990 and 1991 emphasizes men as victims):

	1989	1990	1991	1992	Total
1. Literal	07	18	22	09	56
2. Contextual	04	08	09	04	25
3. Connotative	05	10	18	13	46
4. Neutral	00	07	13	08	28
5. Other	05	04	08	03	20
	21	47	70	37	175

Forty-six per cent (81 of 175) of front-page headlines directly refer to the gender of victims, women being emphasized forty times (79 to two) more frequently than men. These findings are similar to those of the overall sample. The peak years for gender-specific headlines as a percentage of yearly coverage and in absolute numbers were 1990 (47 or 56 per cent) and 1991 (70 or 47 per cent).

One hundred and twenty nine of these headlines continue onto other pages. To determine if the secondary headlines of these articles, where they exist, contain additional information on the gender of victims, I examined every fourth one. In each case, I considered whether the secondary headlines emphasized women, men or neither. In those cases where new information was presented, whether a minor variation on the original headlines or a new point, I considered whether it would have changed how I originally grouped the headline. After this analysis, I concluded there were no headlines which I originally had considered a direct reference that I would have changed. There were three cases where I would have redesignated a non-direct reference as a direct one (these all concerned women as victims). Secondary headlines, then, do not appear to lessen the focus on gender or on women.

Headlines From CNI Categories With Gender-Neutral Titles

The headlines in my sample which most clearly emphasize gender are those from the CNI categories Battered Women and Violence Against Women, categories whose titles explicitly identify the gender of victims. I have examined the remaining categories --Crime and Criminals: Statistics (CC), Crime and Criminals -- Media Coverage (MC), Family Violence (FV), Victims of Crime (VC) and Violence (VI) -- to determine the degree to which they emphasize the gender of victims (the number of headlines emphasizing male victims is listed in brackets):

	1989	1990	1991	1992	Total
1. Literal	11.5(1)	20.0(1)	21.5(3)	18.0(6)	71.0(11)
2. Contextual	25.0	0.0	1.0(1)	0.0	26.0(01)
3. Connotative	11.0	21.5	50.0	31.5	114.0
4. Neutral	44.0	59.0	73.5	121.5	298.0
5. Other	20.0	14.5	21.0	20.5	76.0
	111.5(1)	115.0(1)	167.0(3)	191.5(6)	585.0(12)

The number of headlines in these categories increased 41 per cent in these four years compared to 141 per cent overall. Twelve per cent of these headlines are literal references to gender compared to 34 per cent overall

while contextual references account for five per cent compared to 14 per cent overall. These contextual headlines are virtually non-existent when we consider that 25 of the 26 were published in 1989 and 23 of these concerned the Montreal murders. Like later articles related to the murders, these would likely have been found in the category Violence Against Women had it existed in 1989. The percentage of headlines emphasizing male victims is, at 12 per cent, eight times higher than average and the ratio of headlines emphasizing women to those emphasizing men is eight to one (the lowest to which this ratio drops is three to one in 1992, although five of the six headlines emphasizing men in this period appeared on one day).

I have broken down the data to see whether these trends are found in a particular CNI category:

	CC	MC	FV	VC	VI	Total
Literal	10.0(1)	0.0	47.5(9)	4.5(1)	9.0	71.0(11)
Contextual	0.0