Canadian Press Coverage of the Ethnic Chinese Community: A Content Analysis of The Toronto Star and the Vancouver Sun, 1970-1990

Jianming Ma (Univeristy of Windsor)

Kai Hildebrandt (Univeristy of Windsor)

Abstract: We examined the coverage of ethnic Chinese in The Toronto Star and the Vancouver Sun from 1970 to 1990. Coverage and the diversity of stories grew dramatically. However, the largely positive slant of most 1970 stories was later balanced by neutral and negative coverage. The reportage slant changed partly because cultural stories were augmented by crime and (increasingly negative) immigration stories. The decline of "stereotypically positive" stories may also mark the increasing integration of the Chinese community and their coverage into the mainstream.

Résumé: Nous avons éxaminé les reportages sur les chinois dans le Toronto Star et le Vancouver Sun de 1970 à 1990. Les reportages et la diversité des nouvelles ont beaucoup augmenté. Par contre, le biais plutôt positif des histoires des années 70 a par la suite cédé place à des reportages neutres ou négatifs. Ce changement s'explique par le nombre croissant d'histoires ayant trait à des crimes ou à des situations d'immigration (de plus en plus négatives). Le déclin d'histoires positives stéréotypées marque aussi le début de l'intégration croissante de la communauté chinoise.

Media reporting of ethnic minorities has become a sensitive issue in recent years. Although overtly demeaning or discriminating coverage of minorities has become rarer, many ethnic community leaders as well as media researchers remain critical of the media. In their view, the reporting of negatively loaded issues such as ethnic crimes and immigration may have replaced open ethnic slurs and condescending cultural reportage. Few systematic analyses of media content, however, have examined multicultural coverage in Canada. Our research provides a case study of the treatment of members of the ethnic Chinese community, examining the coverage over two decades in the two largest papers in Toronto and Vancouver, the two Canadian cities with the largest concentrations of ethnic Chinese.

Media Coverage of Minorities

Many recent content analyses on minority reporting have been conducted by American researchers who focused mainly on the image of Blacks and Native peoples. For example, an analysis of five selected US metropolitan papers (Martindale, 1985, 1990a, 1990b) explored the changing image of Blacks from the 1950s to the 1980s and concluded that "the old stereotypes of black Americans are beginning to lose their sway," while new stereotypes, some of them positive ones, are being introduced (Martindale, 1990a, pp. 48-49). There were positive changes in "the increased volume of coverage and greater attention to everyday life coverage of black Americans." Criminal activity by Blacks made up only a very small percentage of the coverage of the Black community during the 1970s; instead, the coverage of Blacks as protesters, politicians, sports figures, and entertainers rose. In the 1980s, however, there was a renewed increase of "stereotypic coverage" of problems such as Blacks involved in crimes and conflicts.

Entman has echoed Martindale's point and developed it into a criticism of "modern racism":

Because old-fashioned racist images are socially undesirable, stereotypes are now more subtle, and stereotyped thinking is reinforced at levels likely to remain below conscious awareness. Rather than the grossly demeaning distortions of yesterday, stereotyping of blacks now allows abstraction from and denial of the racial component. (1992, p. 345)

Entman's recent study found that local television aired a lot of news about Blacks with a generally fair and professional coverage and high visibility of Black anchors and reporters, perhaps conveying the illusion that racism was no longer a social problem. However, crime news about Blacks still abounded, and Black criminals were more often portrayed as lower class and violent than Whites. Black politicians are often treated as a demanding special interest group. These aspects "help to produce modern racism by denying the history of discrimination, whose residue--high crime, high and impatient demands for services--local television so graphically emphasizes" (Entman, 1992, p. 361).

Content analyses of minority reporting by Canadian researchers are rare; until the late 1970s, this remained largely a "blank area" (Ujimoto, 1980, pp. 6-7). Scanlon studied changes in the coverage of "The Sikhs of Vancouver" (1977, pp. 193-262) in the Vancouver Sun and the Province from 1945 to 1974. He found that the press reported "mainly intra-group conflict, conflict among Sikh factions at the temple, conflict and violence among family groups." Similarly, Singer's study of the image of native Canadians in the metropolitan press from 1971 to 1975 yielded negative results (1982, pp. 348-359). Native Indians were often conveyed as dependent on the government and engaged in aggressive land claims; 80% of native news fell into the theme "conflict-deviance." The image of the Inuit appeared less pronounced, but had the same negative features.

Not every study has been critical of the mainstream media. A US study of attitudes toward Native peoples in Alaskan newspapers in the early 1980s concluded that "overall attitudes toward Native people were found to be positive." Although non-Native papers were less favourable than Native ones, the difference was small, and "what [came] through most clearly [was] the pro-Native bias of the Native press" (Murphy & Avery, 1982, pp. 39-45). (The last comment illustrates, albeit inadvertently, the relative nature of "bias": the "pro-Native" bias of Native papers may be particularly visible when viewed against the "norm," i.e., the pro-White bias of the non-Native press!)

Much of the research on the treatment of minorities in the press has been conducted and written as a sub-topic of the issue of minorities in society at large. Discrimination, racism, and xenophobic attitudes have been understood to be the foundation of negative media coverage of minorities. The gradual decline of these forces would then be a major factor making media treatments more positive, and, conversely, an increase in negative media images would point to parallel developments in the attitudes of society. The focus of such discussions is naturally on the evaluations implicit or explicit in the coverage, whereas the extent (and variety) of coverage is of less interest.

In contrast to the "minorities perspective," which may want to draw inferences from the coverage slant to the views of reporters, editors, and /or the public, an alternative view suggests that as coverage increases, so will the variety of issues covered and the variety of formats in which the stories are told. One consequence might be a greater proportion of neutral items, many of which would be short and factual. In a study of newspaper coverage of Black Americans over four decades, Martindale found that as coverage grew, the proportion of stories showing Blacks in "everyday life" situations grew to between 60 and 74% of all stories (1990, p. 104). A greater variety of themes is also likely to result in a greater balance of evaluations: if the coverage historically contained mostly derogatory items, more (and certainly more varied) coverage will include a greater share of positive and neutral stories; if the traditional stories were mostly positive, they may be balanced by neutral or negatively valued material.

Given the rapid increase in the amount and variety of coverage as the point of departure for hypotheses and interpretations, the development of media coverage of the Chinese community in Canada may be seen as an example of what happens when a group (or community) is integrated into society, moving from marginal (and exotic) to mainstream. Trends in the coverage of ethnic Chinese thus should be parallel to similar paths for other groups, organizations, or perhaps even topics.

Background: Chinese Immigration to Canada

Canada has long been a country of immigrants. The most dramatic change since the enactment of the non-discriminatory immigration rules of 1967 has been the influx of Third World immigrants, especially those from Asia--who made up over half of all the newcomers to Canada in the 1980s (Logan, 1991). Together with an institutionalized multicultural policy, this influx will no doubt have a long-term impact on Canada's future. The case of ethnic Chinese is in many ways representative of Canada's immigration history: they were for many decades victims of "institutionalized discrimination" (Li, 1988, p. 33) and have in recent years emerged among the most prosperous under the multicultural policy; while now more visible, they also seem to be more often involved in controversies.

The biggest visible minority group in Canada, the Chinese community had a 1991 census population of nearly half a million, quadrupled since 1971. Due to mass immigration the proportion of Canadian-born Chinese dropped from 40% in 1971 to 25% in 1991. The burgeoning Chinese community is most visible in Ontario and British Columbia, where 80% have settled. Greater Toronto and Vancouver, in turn, have 80% of the respective provincial Chinese populations.

The earliest waves of Chinese immigrants to Canada came in the late nineteenth century for the gold rush in British Columbia and in response to the labour shortage in the construction of the Canadian Pacific Railway. But they had a different fate from the earliest European immigrants to North America. In 1885, the same year the CPR was completed, the Dominion government of Canada introduced a "head tax" levied exclusively upon the Chinese. Finally, in 1923 the federal Parliament promulgated an act that excluded Chinese immigration altogether for 25 years. The act was repealed in 1947, but it was not until 1967, when quotas were eliminated from the immigration law, that the Chinese entered again in large numbers.

A new phenomenon of the post-1967 era, besides the changing composition of immigration, has been the stronger economic presence of recent Asian immigrants. "The coming of Hong Kong money has buried forever the old image of Canada's Chinese as inferior, second-class citizens... they are no longer coolies" (Cannon, 1989, p. 15). Canada's new immigration policy put much emphasis on boosting employment and productivity. To qualify as an "investor" immigrant, for instance, a candidate should have a net worth of $500,000 or more, and commit $150,000 to $250,000 to business investment for at least three years, depending on the destination in Canada. To date, nearly half of the wealthy "entrepreneur" and "investor" immigrants have come from Hong Kong and Taiwan (Logan, 1991, p. 13). Hong Kong investment in Canada reached $2.4 billion in both 1988 and 1989, while the two-way trade between Canada and Hong Kong amounted to only $2.2 billion in 1989 ( Joint Centre, Spring 1990, p. 13).

In the 1986 census, ethnic Chinese caught up with the rest of the nation in managerial and clerical occupations for the first time, and they were overrepresented in professional jobs. They also made rapid strides in higher education. In the censuses of 1981 and 1986, the proportion of ethnic Chinese holding university degrees was approximately double the national level, suggesting further potential for upward social and economic mobility.

Methodology: Sampling and Coding

The two newspapers in this study, The Toronto Star and the Vancouver Sun, constitute a purposive sample. They are the largest dailies in the metropolitan areas with the two largest ethnic Chinese communities in Canada. They have similar circulations, similar prestige, similar layout characteristics, a similar length of around 70 pages per issue and similar sections. They also adopt similar ethnic coverage policies. The Vancouver Sun holds that "multicultural coverage should not be separated from other coverage, that the various ethnic communities should be covered as mainstream news." The Toronto Star declares: "We do not single out any community for specific coverage, but we do cover interesting and important events with the community and also important to Toronto as a whole."

For story selection purposes, "ethnic Chinese" in this study refers to all Canadian citizens and permanent residents (including applicants in Canada) of claimed Chinese origin who were either born in Canada or whose apparent last country of residence was China, Hong Kong, or Taiwan. To be coherent with official statistics, the concept excludes ethnic Chinese originating from other countries. This exclusion is considered a necessary price for achieving higher reliability in story selection. Consequently, the estimates of coverage are conservative: if ethnic Chinese immigrants from other countries were included in official statistics or our coding, the reported figures for population, investment, crime--as well as incidence of visibility in newspaper coverage--would all increase.

Every 1990 issue of The Toronto Star and the Vancouver Sun was sampled and examined for stories involving ethnic Chinese. Since exploratory searches had shown the low sampling success rate for the earlier time points, only every second 1970 and 1980 issue was sampled. For the analysis, the difference in the sampling fractions was adjusted by weighting the data. The searching for newspaper items was done manually through microfilm copies. Given the large number of selection decisions, the possibility exists that some relevant items were missed. We intended to use the Canadian News Index as a control; however, the Index does not offer the category of "Chinese-Canadians" for issues prior to 1986. For 1990, the Index selected only 10 stories from both papers, compared to our 628. Obviously, the Index has a much narrower definition of ethnic Chinese coverage, and in addition does not include relevant pictures and cartoons.

Stories were coded on 15 dimensions, including items such as length, format, by-line, theme, level of topic, source, etc. Multiple coding was allowed for level and themes of stories to avoid loss of content information. The most difficult coding category was "slant," which attempted to capture the evaluation of ethnic Chinese expressed in the news items. To minimize subjectivity, coders were required to make their judgment only on the literal meaning of each item.

Findings

Overall, 783 stories, pictures, and cartoons with at least partial ethnic Chinese content were selected in The Toronto Star and the Vancouver Sun of 1970, 1980, and 1990 as fitting the selection criteria. Since only every second issue was examined from the 1970 and 1980 Sun and Star, the difference in sampling fractions was adjusted by doubling the numbers of 1970 and 1980 items: consequently, in 1970, we estimate that about 120 items in both papers combined had ethnic Chinese content, or a little more than one story per paper per week. The number increased to 190 in 1980, or nearly two stories per paper per week. Finally, 10 years later there were 628 items in total, or almost one story per paper for each publishing day (see Table 1).

Table 1 also documents the population growth of the ethnic Chinese communities in Toronto and Vancouver, and allows comparisons between population and coverage growth rates. Over the entire 1970-1990 time span, coverage of ethnic Chinese rose by 452% in the Star, lagging behind the 660% growth rate of Toronto's Chinese population. In Vancouver, the relative magnitude of the growth rates was reversed: coverage more than quintupled, and exceeded the Chinese population growth of 293%. In the 1970s, the coverage increase in both papers lagged behind the population growth: population more than doubled in both cities, while coverage increased by 54 to 61%. In the 1980s, the population grew only half as fast as in the preceding decade, but coverage more than tripled.

Individual stories also became much longer. The median length of Toronto Star stories in 1970 was just 6 column inches (standard: 6 columns per page), or about the size of a cigarette package. It grew to 14 column inches in 1980 and to 17 inches in 1990. The Vancouver Sun devoted a median length of 14 column inches to Chinese stories in 1970, 17 inches in 1980, and 24 inches in 1990. Similarly, the average Chinese story in the Star increased from 13 in 1970 to 20 and 22 column inches 10 and 20 years later; in the Sun, the comparable means were 18, 25, and 26 (Ma, 1992, p. 86). The maximum individual story length grew as well, in the Star from 44 column inches in 1970 to 117 column inches in 1990, and in the Sun from 55 to 184 inches, or to well over one and a quarter pages.

Table 1
Relationship of Chinese Ethnic Population and Reportage Growth
Toronto Vancouver
Year Chinese Pop. Star storiesa Chinese Pop. Sun storiesa
1970 26,285 48 36,405 72
1980 89,590 74 83,945 116
1990 199,720 265 143,010 363
% change
1970 to 1980 241 54 139 61
% change
1980 to 1990 123 272 70 213
% change
1970 to 1990 660 452 293 404
a.
Weighted to adjust for differences in sampling fractions.


Source for population data: Statistics Canada, Census. Actual census years are 1971, 1981, 1991.

The visibility of the ethnic Chinese community increased not only in frequency and size of coverage. In both The Toronto Star and the Vancouver Sun the layout visibility of ethnic Chinese stories grew over time, with larger headlines, more pictures, more upper- and front-page locations, and more prominent attributions.

Increased media coverage of specific population groups may be due to a variety of factors, including changes in editorial policy, the hiring of group members, and the growing importance of groups in terms of their numbers, political power, or economic clout. All these factors probably played a role in the observed greater visibility. For example, while neither paper has adopted explicit policies on multicultural coverage, both appeared sensitive to the issue when queried (Ma, 1992, pp. 172-173). In both papers, the percentage of stories with Chinese content that had ethnic Chinese by-lines increased from 3 and 4% in 1970 to between a fifth (The Toronto Star) and a quarter (the Vancouver Sun) in 1990. The population increase and the rising socio-economic level of the community were described above.

In most (if not all) of these instances, increased attention by the media is likely to also foster changes in the content of media coverage. As a group moves from the margins to societal mainstream, coverage is less likely to focus on the exotic and unusual, both because the group is seen as less exotic and because it is more involved in mainstream activities. This process can be examined through over-time analysis of story format, geographic focus, and, most importantly, the thematic content of press coverage.

Most important among the format changes were small (5%) increases in the share of opinion pieces (from 3% to 8%) and of hard news items in both papers (from 45% to 50%) between 1970 and 1990, although the hard news share had dipped to one-third in both papers in the middle time point, 1980. In The Toronto Star, the growth came at the expense of an 8% decline in features, which dropped from 38% to about 30%. In the Vancouver Sun, the features' share actually grew by 9% to 37%; instead, the Sun scaled down its use of free-standing pictures from 28% to reach the level of the Star (about 8%). In both papers, about half of all 1990 stories contained pictures, presumably increasing their impact.

Coverage continued to be dominated by stories with a local focus, but to a less overwhelming degree (down from 87% to 76%). In contrast, the percentages of stories with home-province interest (24% in 1990) or other provincial (17% in 1990), national (38% in 1990), and international interest (42%) all increased. In addition, the share of items with multiple levels doubled in the 20-year span, reflecting perhaps a greater complexity of stories and the impact of immigration.

The most important indicators of coverage content are, of course, themes or topics, since they reflect what the press staffs perceive as newsworthy about a group. In 1970 the press seemed primarily interested in the cultural aspects of ethnic Chinese. Chinatown stories dwelt upon such things as restaurants, food, arts, and lifestyles. The 1980 and especially the 1990 stories covered a wider variety of topics, including economics or business, crime, immigration, and politics. Cultural themes remained popular, but were no longer the sole facet of the Chinese image.

The over-time analysis of thematic changes (Table 2) showed that cultural themes dropped overall by a substantial 30%, from 73% in 1970 to second place in 1990 (43%). The topical change was more dramatic in the Sun (from 81% to 39%) than in the Star (63% to 49%). In the Star, cultural stories retained a narrow lead over economic stories, but it is common to both papers that the decline of cultural items happened only during the 1980s when economic themes took over the lead, climbing in the combined figures from 35% to 53%. This reversal of cultural and economic themes may reflect the greatest image update for the Canadian-Chinese community throughout its history. For the first time in well over a century, ethnic Chinese have won wide public attention for their increasing economic power--rather than for allegedly exchanging cheap "coolie labour" for a share of Canadian wealth.

Table 2
Changes in Reportage Themes
Topic/ Theme
Paper Year Political Cultural Immigration Crime Economic Total
Star 1970 25%a 63% 33% 8%b 33% 40
1980 24% 40% 16% 11% 32% 63
1990 39% 49% 28% 30% 45% 223
Sun 1970 36%a 79%b 17% 6%a 36%b 61
1980 24% 85% 19% 19% 43% 96
1990 38% 39% 28% 19% 59% 303
Both 1970 32%b 73%b 23%a 7%b 35%b 100
1980 24% 75% 18% 16% 3% 159
1990 39% 43% 28% 23% 53% 526
a.
Change ( sn2>) significant at p < .05.
b.
Change ( sn2>) significant at p < .01.


Note: Percentages sum across rows to > 100%, because nearly 60% of items were coded as touching on more than one theme.

Such enhanced economic representation may help explain the simultaneous if slower growth of the political theme. In 1970, fewer than one-third of all stories about the Chinese contained references to political issues or politicians. By 1980, the political content had diminished even further (to 24%), especially in the Sun, but it increased to 39% in 1990. While in 1970 political items were linked with economics, in 1980 and 1990 political stories were more likely immigration stories as well; the immigration theme showed parallel trends with political topics over the three time points.

While The Toronto Star's coverage of immigration fluctuated, the Vancouver Sun showed constant growth in immigration stories from 17% in 1970 to 28% (equal to the Star's share) in 1990. Curiously, between 1970 and 1990 Chinese immigrants shifted their preferred landing place from the West Coast to Ontario. In contrast, immigration topics were most prevalent (in 33% of all stories) in the Star in 1970, but declined when the growth of the Chinese community reached its peak in Toronto. This is one example of the extent of coverage not simply following the population numbers. However, the Chinese community in Toronto constituted a smaller portion of the metro population (5.3% in 1990) compared with Vancouver's 6.6% in 1980 and 9% in 1990. Moreover, with its location on the Pacific Rim and its traditionally developed Asian market, Vancouver seemed to have attracted a stronger influx of immigrating wealth.

Among all coded themes, the fastest growth occurred in the area of crime, perhaps mirroring general social trends as much as (or more than) ethnic realities or perceptions. In 1970, only 6% and 8%, respectively, of the Sun's and Star's Chinese stories involved crimes. By 1980, the crime theme rose to 19% and 11% in the two papers, and by 1990, crime was a topic in 19% of the Sun's and fully 30% in the Star's items.

Since many stories reflected more than one theme (and were so coded), we calculated the correlations among themes over time to determine which themes tended to co-occur (data not shown). In 1970 economic and political topics were significantly correlated at .33. There were few purely economic or business stories involving ethnic Chinese, and Chinese economic activities that did attract public attention were often linked with political issues. With the increase of normal economic and business stories involving Chinese, the association between economic and political themes disappeared from 1980 on.

Another theme consistently correlated with politics was immigration, at about .23 at each time point. State intervention clearly played a crucial role throughout Canada's immigration history, and immigration policy always implied political choices about what kind of society the Canadian state wanted. By 1990, immigration stories also often involved crime or economics and business themes, as indicated by significant coefficients of .16 and .10.

Summing up our thematic analysis, we see changes in these aspects: economic themes replaced cultural topics as the centre of media attention, especially in Vancouver, as the Chinese community became more influential and visible; and perhaps more controversially, the media increasingly selected political, immigration, and especially crime issues in their coverage of the Chinese community. Immigration became less of an abstract political issue and more linked with economics and crime.

Changes in the media agenda--and especially changing linkages of issues with groups--often imply more than shifting topics, since most societal and political issues are not neutral but evoke positive or negative associations. Since many issues imply a specific evaluation ("valence issues," Clarke, et al., 1979, p. 244), agenda shifts often mean changes in evaluations as well.

The analysis of coverage slant is at the heart of many analyses of the media treatment of minorities or marginalized groups. We determined for each selected news item whether it had a positive, negative, or neutral slant. The default code was neutral: it was assigned if the item overall showed neither an obvious positive nor an obvious negative slant, or if there was a balance between positive and negative aspects. Positive slant usually meant references to co-operation, confirmation, and adaptation or catering to the values and views of North American society. Negative slant meant strangeness, disconfirmation of North American values, difficulty of understanding, isolation, or alienation. Of course, the confirmation or rejection of a society's dominant values is not automatically positive or negative. However, most members of society and most (mainstream) media accept the dominant social values as the norm, and evaluate the "other" positively or negatively to the extent that the "other" exhibits the norm.

Following the literature, we had expected gradual change in the slant of ethnic Chinese coverage from more negative to less negative and to more positive evaluations. Table 3 shows a very different picture, however. It shows a trend from largely positive to fewer positive evaluations in both The Toronto Star and the Vancouver Sun. At the same time, both the number of negative and neutral stories grew. Positive stories still outnumbered negative stories at each time point and in each paper, but negative stories overall reached 25% in 1990.

Table 3
Summary of Reportage Slant
Slant Percentage
Difference
Paper Year Positive Neutral Negative Index (PDI) (n)
Star a 1970 45.8% 29.2% 25.0% +20.8 (40)
1980 48.6% 45.9% 5.4% +43.2 (63)
1990 35.1% 38.5% 26.4% +8.7 (223)
Suna 1970 50.0% 44.4% 5.6% +44.4 (61)
1980 32.8% 50.0% 17.2% +15.6 (96)
1990 26.2% 50.1% 23.7% +2.5 (303)
Botha 1970 48.3% 38.3% 13.3% +35.0 (100)
1980 38.9% 48.4% 12.6% +26.3 (159)
1990 29.9% 45.2% 24.8% +5.1 (526)
a.
Change ( sn2>) significant at p < .01.

In 1970 in The Toronto Star, 45.8% of the stories were positive compared with 25% negative. The negative percentage was four times higher than the Sun's, perhaps as a reflection of Toronto's difficulties in absorbing the first wave of the Chinese immigration flow that had shifted eastward. In 1980, there were hardly any negative stories (5.4%), while almost equal proportions of the Chinese stories were positive and neutral. By 1990, however, the perception of ethnic Chinese in the Star was more controversial, as positive and neutral reportage each declined to between 35 and 40%, while negative stories made up 26.4%, mainly focusing on Chinatown crimes.

The Vancouver Sun showed a steadier trend. In 1970, half of its stories were positive, slightly fewer neutral, and only 5.6% were negative. In 1980, half the Sun's items were neutral, positive stories declined to one third, and negative stories went up to 17.2%. In 1990, half the stories were again coded neutral, while the positive and negative percentages were nearly even at about a quarter each.

Perhaps the most important overall trend in coverage slant was a big increase in neutral stories, replacing positive items as the mainstay of coverage--in line with the professed values of Western journalism. Similarly, clearly positive items were almost balanced by negative ones in the overall assessment, although in the Star positive stories still outnumbered negative ones by almost 10%.

Combined, the decline in positive items, the increase in negative ones and the preponderance of neutral stories signified a shift away from clearly positive evaluations. We can summarize the "balance" between positive and negative slant by calculating a Percentage Difference Index. The PDI is established by subtracting the negative from the positive percentages. The Star had a positive PDI of 20.8 in 1970, which declined to 8.7 in 1990--a change of 12.1 points. The change in the Sun was even more dramatic: from 44.4 to just 2.5, a drop of 41.9!

Since over the study period the balance of story slants became less positive--in contrast to the rising trends in story number, story length, layout effect, and content variety--we need to ask why the reporting slant turned so much less positive as coverage grew and "multiculturalism" was widely promoted. Before engaging in speculations, we can find out what stories specifically were receiving negative reporting and, in contrast, what stories tended to be neutrally or more positively written. If there were patterns, the shifting thematic composition of the coverage may help account for the overall change in evaluations.

Cultural topics and crime were at two opposite extremes of reporting slant: cultural stories were highly positive with an overall Percentage Difference Index between positive and negative slants of +39.0, while crime stories were the most negative with an overall PDI of 49.7. This was not surprising, as we had earlier described cultural stories as a traditional niche of ethnic Chinese coverage, while the issue of crime usually evokes negative evaluations. Other themes varied to lesser degrees: economic stories were moderately positive at +9.1, political issues basically neutral ( 0.9), and immigration had a slightly negative PDI of 6.7. Similar analyses of the slants of different formats and levels of geographic interest show that the most "positive" formats (pictures alone and features) and levels (local) have declined over time, while the increasingly frequent formats (hard news, opinion) and levels (higher than local) are evaluated less positively.

Our analysis of the changing coverage showed that over time traditional stories about ethnic Chinese (such as features on food and lifestyles, local news, and stories on cultural themes) gave way to hard news and opinions, wider areas of interest, and other themes--economics and crime in particular. These changes contributed significantly to the decline in the positive slant of ethnic Chinese coverage. Shifting topical agendas and other reportage developments, however, are not the only reason for changes in slant. Issues themselves change over time in their connotations and significations. To explore this possibility, we examined the slant of the coded themes over time (Table 4).

Table 4
Slant Changes by Theme (Both Papers Combined)
Slant Percentage
Difference
Theme Year Positive Neutral Negative Index (PDI) (n)
Culturala 1970 56.8% 38.6% 4.5% +52.3 (73)
1980 46.5% 46.5% 7.0% +39.5 (118)
1990 49.3% 36.0% 14.7% +34.8 (227)
Economicb 1970 52.4% 33.3% 14.3% +38.1 (35)
1980 40.5% 43.2% 16.2% +24.3 (62)
1990 26.0% 50.1% 23.9% +2.1 (280)
Politicalb 1970 42.1% 42.1% 15.8% +26.3 (31)
1980 26.1% 65.2% 8.7% +17.4 (38)
1990 24.3% 42.8% 32.9% 8.6 (203)
Immigrationb 1970 57.1% 21.4% 21.4% +35.7 (23)
1980 41.2% 52.9% 5.9% +35.3 (29)
1990 24.0% 30.3% 45.7% 21.7 (146)
Crime 1970 0 25.0% 75.0% 75.0 (8)
1980 26.7% 20.0% 53.3% 26.6 (25)
1990 11.6% 23.8% 64.6% 53.0 (122)
a.
Change ( sn2>) significant at p < .05.
b.
Change ( sn2>) significant at p < .01.

All themes, except possibly crime, exhibited an obvious and steady decline in positive slant; not surprisingly, crime remained throughout the most negatively evaluated theme with PDIs between 26.6 and 75. The most positive cultural theme had an extremely positive PDI of +52.3 in 1970 which dropped to +39.5 10 years later and to +34.8 in 1990, a 20-year decrease of 17.5 points.

Economic stories were also evaluated quite positively in 1970 (PDI = +38.1). By 1990, they had become more "business-like," with a virtually neutral PDI of +2.1. Political themes actually turned to a negative balance, from quite positive PDIs of +26.3 in 1970 and +17.4 in 1980 to 8.6 in 1990. The most dramatic shift occurred, however, with the issue of immigration, which plummeted from very positive PDI levels of +35 in both 1970 and 1980 to 21.7 in 1990. Such a reversal suggests a wholesale change in the meaning of the issue as covered by the press.

Over half of the stories included more than one theme, and themes differ in the evaluation (slant) they evoke. Therefore, it is unclear which of several themes should be considered "responsible" for the slant of a multiple-topic item. Since the crime theme was the most one-sided topic, the removal from the analysis of all stories in which crime is at least one of the themes would give a clearer assessment of the valuation of the other themes.

The exclusion of stories that included the crime theme raised the PDI (i.e., made the balance more positive) for all other themes to varying degrees. The cultural theme was the least affected, since few of the cultural stories involved crime. It remained the most positive issue (+39.4 in 1990), but the trend toward less positive coverage even of cultural topics also persisted. Stories with economic themes became more positive by 7 PDI points for 1970 and by around 14 points in 1980 and 1990, so that the 1990 balance was +15.4. In 1970 and 1980 few stories with political themes overlapped with crime topics, but in 1990 almost a quarter of the political items also dealt with crime; when these items were deleted, the slant balance for political items changed by 10%, returning to a positive +1.6.

The exclusion of the crime theme had the most dramatic impact on the theme of immigration--and almost exclusively in 1990. Over one-third of the immigration stories in that year dealt with crime as well; removing the crime-related stories turned the slant balance by 20 points from a PDI of 21.7 to 1.7, since over half of the negative immigration stories (but only 20% of the positive and 14% of the neutral ones) involved crime. Still, even disregarding the crime issue, the press treatment of immigration had changed from largely positive in 1970 and 1980 to neutral and even slightly negative hues. It is beyond the scope of this study to determine whether the negative treatment of immigration topics was limited to ethnic Chinese or was part of a more general apparent turn in public opinion against the entry of non-White immigrants to Canada.

When we went one step further and excluded both crime- and immigration-related stories, the number of remaining stories became small for some categories and years, but the results were suggestive. Much of the decline of positive reporting and the increase of negative reporting appeared to be due to the themes of crime and immigration and their increasing linkage with other themes. For stories with cultural and political themes, the remaining slant change was small and not statistically significant. In stories with business or economic topics, positive stories were replaced by neutral stories (barely significant); only 6.8% of the 1990 economic or business stories retained a negative slant once immigration- and crime-related stories were deleted. Aside from the negatively loaded issues of crime and (in 1990) immigration, the change in slant was partly due to a more critical attitude on the part of the press, but mainly the result of a tendency to report more professionally (and thus neutrally) about Chinese topics. The increasing number and percentage of Chinese-Canadian journalists contributed significantly to the growth of neutral items, as their items were more likely to be neutral than stories with non-Chinese or no attributions (Ma, 1992, p. 134).

In comparing the two papers, the Vancouver Sun became less positive than the Star, but the change could be attributed mainly to the Sun's larger proportion of neutral stories. The Star had both more positive (cultural) items and more negative stories, the latter involving both crime and immigration reportage.

Conclusion

Once again the Chinese community is receiving rapidly increasing attention from the press, the first time since after the early waves of Chinese immigration in the late 1880s (Lai, 1988). For nearly a century after that, the Chinese community kept a low public profile. What coverage there was appeared most likely in cultural sections, like restaurant and food reviews or life and family columns. The Chinese community appeared in the press as a social enclave whose persistent existence in Canada was tolerated, whose culture was mostly appreciated, but whose interaction with other parts of society was very limited. Individual stories may have had a positive slant, but such limited and exoticized stories could be interpreted overall as very negative coverage, as the media provide models for viewers and readers; if the coverage of minorities is limited, minority members are denied a full range of possibilities (Gist, 1990, p. 57).

Our study of these three snapshots (1970, 1980, and 1990) shows that there have been dramatic changes: newspaper coverage more than quintupled in each city, and during the 1980s The Toronto Star's coverage of the Chinese community more than doubled the city's Chinese population growth rate, while in the Vancouver Sun the number of stories grew four times as fast as the local Chinese population. The Chinese of the 1990s are no longer only part of Chinatowns or certain neighbourhoods. Hong Kong immigrants and investors have such a presence in metropolitan daily life that Vancouver has been nicknamed by some as "Hongcouver." And just as they are changing the physical appearance of Vancouver and Toronto, ethnic Chinese of the 1990s are also reaching out beyond the newspapers' cultural pages. They are more frequently seen in hard news and crime stories, in the business section and heard with a growing voice from the opinion pages.

While gaining a much greater volume of publicity in recent years, the Chinese community was less positively and more neutrally depicted in the press. In 1990, there were almost as many negative as positive Chinese stories. Thus, the broader coverage ends years of "positive stereotyping" (Martindale, 1990a) in the patronizingly positive if sporadic coverage of the Chinese community as an alien culture of largely ornamental value. Such coverage approached Chinese topics with a partially appreciative but generally narrow perspective, as many other themes were bypassed. In this view, less overwhelmingly positive coverage was structurally the normal and encouraging result of the enormous increases in coverage quantity and content variety.

Yet the trend toward less positive coverage slant can also be viewed with concern, as negative attitudes toward ethnic Chinese (and maybe other visible minorities) appear to be on the rise and are finding their expression in the media, not in explicitly racist comments but through the increasing numbers of crime stories and the reversed slant of immigration stories, in an example of what Entman (1992) and others have termed "modern racism."

Either of these perspectives is consonant with our empirical findings. The two processes differ, however, in their predictions for future coverage. The "minorities perspective" draws upon public attitudes (and prejudices), and current readings from Canada as well as abroad would warn us that at least as long as economic prosperity is not restored (and perhaps even beyond), racism and anti-foreign sentiments are likely to persist, even if the affected groups are legally fully part of domestic society. On the other hand, the alternative perspective would imply that the relative balance of coverage, including a large share of neutral stories, is likely to hold unless there are major open conflicts pitting minorities against the dominant (as seen and embodied by the media) culture.

Notes

1
We are indebted to Bruce Burton, Christopher King, Walt Romanow, and Stuart Surlin for their help at various stages in this project or for their comments on earlier drafts.
2
The tax was $50 per person in 1885, $100 in 1900, and $500 from 1903 on. In comparison, the annual income for Chinese labourer in 1885 was $225, and between 1900 and 1905 the cost of living for average British Columbia family was between $600 and $700. See Li (1988) and the Royal Commission on Chinese Immigration (1885) Report, Ottawa, as cited in Li (1988).
3
Scott Honeyman, Managing Editor of the Vancouver Sun, personal communication, November 11, 1992: "We do not have a written policy about ethnic issues. We do have a multicultural advisory panel that... bring[s] to our attention where we are considered to be insensitive or lacking. Partly as a result of consultations with the panel, we have decided that multicultural coverage should not be separated from other coverage, that the various ethnic communities should be covered as mainstream news. We are developing contact lists that give us experts in all areas from ethnic communities, so that the faces and names in our newspaper won't be so predominantly male and Caucasian."
4
Mike Pieri, The Toronto Star's Assistant Manager of Special Projects, personal communication, December 2, 1992: "[W]e insist that no reference should be made to a person's race, color or religion unless it is pertinent to the story. In crime stories, particularly, we insist that great care must be exercised.... We do not single out any community for specific coverage, but we do cover interesting and important events with the community and also important to Toronto as a whole. Asian gang warfare is a subject that requires special vigilance. Youths from Southeast Asia do, occasionally, cause trouble in Toronto's four Chinatowns. Ethnic Chinese are invariably victims. Unfortunately, less responsible media often unwittingly fail to make the distinction, with the result that the wrong impression can be given to the public."
5
Articles were selected if they contained Chinese names, references to Chinese ethnicity, or if persons or other subjects were identifiable as Chinese from pictures. If the text contained references to origins other than those defined as "ethnic Chinese" by Statistics Canada, the article was excluded.
6
An inter-coder reliability test was administered among six independent coders who were graduate students at the University of Windsor; their co-operation is hereby gratefully acknowledged. They represented a mix of three Caucasian Canadians, two ethnic Chinese Canadians, and one East Indian Canadian. An overall inter-coder reliability of .96 was achieved, resulting largely from the simplicity of dichotomous "mentioned /not mentioned" decisions for most variables. The inter-coder reliability for the most "subjective" variable, story slant, scored lower at .83, though still within acceptable limits.
7
For the subsequent analyses following Table 1, the data were weighted to adjust for the different sampling fractions. To avoid an artificial inflation of the significance levels, the total n for the weighted data was set equal to the number of actually coded stories (783). The weighting does not alter the composition of the yearly totals ("Both" in subsequent tables): consistently, three-fifths of the combined coverage comes from the Vancouver Sun, two-fifths from The Toronto Star.
8
After coding five indicators of visibility (headline size; picture usage; position on page; position in paper; and source or attribution) into high, medium, low, and no visibility, a visibility index was constructed by summing the indicators for each news item. The average index grew from 7.4 in 1970 to 9.7 in 1990 for the Star, and from 7.9 to 9.0 for the Sun. For further details, see Ma (1992, p. 98).
9
The PDI is therefore a measure of the balance between positive and negative items. Since the percentage "neutral" does not enter the calculation, the PDI has the effect of overlooking this group. This may misrepresent the full distributions if the share of the neutral group varies between themes or over time. Since we are explicitly only interested in the balance of slant, omitting "neutral" is not a problem.

References

Cannon, Margaret. (1989). China tide: The revealing story of the Hong Kong exodus to Canada. Don Mills: Harper and Collins.

Clarke, Harold, Jenson, Jane, LeDuc, Lawrence, & Pammett, Jon. (1979). Political choice in Canada. Toronto: McGraw-Hill Ryerson.

Entman, Robert M. (1992, Summer). Blacks in the news: Television, modern racism and cultural change. Journalism Quarterly, 69(2), 341-361.

Gist, M. E. (1990, Summer). Minorities in media imagery: A social cognitive perspective on journalistic bias. Newspaper Research Journal, 11(3), 52-63.

Joint Centre for Asia Pacific Studies. (1990, Spring, Fall). Canada and Hong Kong update. Toronto: York University.

Lai, David Chuenyan. (1988). Towns within cities: Chinatowns within cities in Canada. Vancouver: University of British Columbia Press.

Li, Peter S. (1988). The Chinese in Canada. Toronto: Oxford University Press.

Logan, Ronald. (1991, Spring). Immigration during the 1980s. In Statistics Canada, Canadian social trends. Ottawa: Ministry of Supply and Services Canada.

Ma, Jianming. (1992). Canadian press coverage of the ethnic Chinese community: A content analysis of The Toronto Starand the Vancouver Sun 1970-1990. University of Windsor: MA thesis.

Martindale, Carolyn. (1985, Summer). Coverage of black Americans in five newspapers since 1950. Journalism Quarterly, 62(2), 321-328.

Martindale, Carolyn, (1990a, Winter). Changes in newspaper images of black Americans. Newspaper Research Journal, 11(1), 40-50.

Martindale, Carolyn. (1990b, Summer). Coverage of black Americans in four major newspapers, 1950-1989. Newspaper Research Journal, 11(3), 96-112.

Murphy, James E., & Avery, Donald R. (1982, Fall). A study of favourability toward natives in Alaskan newspapers. Newspaper Research Journal, 4(1), 39-45.

Scanlon, Joseph. (1977). The Sikhs of Vancouver: A case study of the role of the media in ethnic relations. In UNESCO, Ethnicity and the media: An analysis of media reporting in the United Kingdom, Canada and Ireland. New York: UNESCO.

Singer, Benjamin D. (1982, August). Minorities and the media: A content analysis of native Canadians in the daily press. Canadian Review of Sociology and Anthropology, 19(3), 348-359.

Statistics Canada. (1971, 1981, 1986). Census of Canada: Ethnic groups, immigrant population. Ottawa: Ministry of Supply and Services Canada.

Ujimoto, K. Victor. (1980). Visible minorities and multiculturalism: Asians in Canada. Toronto: Butterworth.



  •  Announcements
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Current Issue
    Atom logo
    RSS2 logo
    RSS1 logo
  •  Thesis Abstracts
    Atom logo
    RSS2 logo
    RSS1 logo

We wish to acknowledge the financial support of the Social Sciences and Humanities Research Council for their financial support through theAid to Scholarly Journals Program.

SSHRC LOGO