Framing economic inequality in the news in Canada and the United States

The paper investigates how economic inequality was framed in Canadian and U.S. news between 2000 and 2014 and what determinants shaped those frames. Motivated by long-standing framing scholarship and the policy relevance of frames, the authors examine whether inequality is presented as a social problem, how causes and solutions are attributed, and how contextual factors—national context, newspaper political leaning, economic conditions (Great Recession), and social movement activity (Occupy)—influence framing. They note a gap in empirical documentation of inequality framing per se and aim to compare multiple determinants simultaneously to advance theory about how news frames emerge.

The authors review scholarship on economic inequality trends in Canada and the U.S. (inequality high in 1920s–30s, low around 1970, rising since; U.S. more unequal than Canada) and contextual differences (Canadian politics with stronger labor presence and higher expectations of state intervention). They discuss the Great Recession’s disproportionate negative impact on lower-income households and the emergence of the Occupy movement in 2011, which centered inequality with the 99%/1% rhetoric. They review determinants of frames: economic conditions (objective trends vs. constructed newsworthiness), national cultural repertoires and discursive opportunities (e.g., U.S. market logic vs. other nations’ solidaristic logics), social movement efforts (capacity to shape attention and sometimes frames), and media political orientation (liberal vs. conservative differences in attributing causes/solutions in prior studies). These yield focused research questions on whether the Great Recession or Occupy shifted attention and framing, whether Canadian news would more often frame inequality as a social problem with state solutions, and whether liberal-leaning outlets would differ from conservative ones.

Design: Mixed approach combining attention metrics and content analysis. Attention data were drawn from Lexis Nexis counts of articles by year under the subject keyword “economic inequality.” Framing data came from hand-coding a corpus of 2109 articles from four national newspapers (U.S.: New York Times [NYT], Wall Street Journal [WSJ]; Canada: Toronto Star [TS], Globe and Mail [GM]) published from 01/01/2000 to 12/31/2014. Sampling: Articles were retrieved via ProQuest using the keyword phrase “economic inequality.” Articles without explicit or implicit discussion of economic inequality were excluded. Coding: Six trained research assistants coded all articles in Dedoose. A subsample of 50 articles (from 2002–2004) was double-coded to assess reliability; average pairwise agreement exceeded 85%, and Krippendorff’s alpha ranged 0.68–0.82. Variables: Problem definition (binary: any mention of inequality as a social problem/negative consequences), causes (binary indicators for individual, corporate sector, state), and solutions (binary indicators for individual, corporate sector, state). Determinants/independent variables: national context (U.S. vs. Canada via newspaper grouping), political leaning (NYT, TS as liberal-leaning; WSJ, GM as conservative-leaning), Great Recession indicator (0 for articles up to 12/31/2007; 1 thereafter), and Occupy indicator (0 up to 09/30/2011; 1 thereafter). Analysis: Trends in attention and framing were described; logistic regression models (NYT as reference) estimated associations of determinants with each frame: problem identification, corporate-sector cause, individual-level solution, corporate-sector solution, and state-based solution. Attention was measured by yearly counts of “economic inequality” articles in Lexis Nexis and compared to “poverty” as a reference topic.

• Attention: Economic inequality coverage did not increase after the Great Recession, but spiked sharply in 2011 coincident with Occupy and remained elevated afterward. Poverty coverage was far more voluminous overall and rose more linearly; the poverty-to-inequality article ratio fell from about 200:1 (2000) to 60:1 (2014).
• Problem identification: No significant effect of the Great Recession; very strong effect of Occupy, with odds of identifying inequality as a problem nearly sixfold higher post-Occupy (OR ≈ 5.99, p<0.001). TS articles were more likely than NYT to identify inequality as a problem (OR ≈ 2.541, p<0.001); WSJ and GM did not differ significantly from NYT.
• Causes: Corporate-sector cause framing did not increase post-Occupy; only GM was significantly more likely than NYT to cite corporate causes (OR ≈ 1.573, p<0.01). Individual-level causes were rare throughout; state causes were initially more common than corporate causes but declined after mid-2000s, converging by 2014.
• Individual-level solutions: No significant associations with any determinant (including Occupy, national context, or political leaning); mentions were rare across years.
• Corporate-sector solutions: GM was much more likely than NYT to cite corporate solutions (OR ≈ 3.243, p<0.001). Mentions of corporate solutions increased post-Occupy (OR ≈ 2.596, p<0.001), contrary to Occupy’s critique of corporate responsibility.
• State-based solutions: Most commonly cited overall, but without clear temporal pattern; no significant differences for GM or WSJ vs. NYT. TS was significantly less likely than NYT to cite state solutions (OR ≈ 0.738, p<0.01). Occupy and the Great Recession had no significant effects.
• Overall framing pattern: By the end of the period, coverage reflects an “urgent but ambiguous” frame—inequality is widely recognized as a serious problem, but causes and solutions are diverse and lack consensus. Individual blame/solutions are uncommon.

The findings show that social movement activity—specifically Occupy—had a strong effect on news agendas and on problem identification, elevating both attention to and recognition of economic inequality as harmful. However, Occupy did not shift framing around causes (e.g., corporate responsibility) or preferred solutions (e.g., state intervention), suggesting a limited, surface-level framing impact. National context and political leaning had generally weak and inconsistent effects; the Canadian outlets did not uniformly emphasize state solutions, and liberal vs. conservative differences were limited. The pattern aligns with political economy critiques suggesting media’s structural ties may constrain uptake of frames that challenge corporate interests. The emergent “urgent but ambiguous” frame may influence public discourse by acknowledging inequality’s harms while diffusing clear attributions or policy pathways, potentially dampening transformative policy responses.

Empirically, the study documents how North American newspapers framed economic inequality from 2000–2014: growing recognition of inequality as a problem after 2011, sparse individual blame, and mixed, nonconvergent accounts of causes and solutions with state solutions most frequent but not ascendant. Conceptually, the study compares multiple determinants and finds social movement activity to be the most influential for agenda setting and problem recognition, outweighing economic shocks, national context, and political leaning, yet with limited impact on deeper causal and solution frames. Future research should broaden media outlet and country samples to disentangle national context from political orientation, examine diverse movements’ characteristics and tactics in shaping frames, and test longer post-2014 trajectories to see whether clearer causal/solution frames emerge.

• Limited media sample: Only four newspapers, two per country, may underrepresent variation in national contexts and political leanings; political leaning and national context may be confounded (e.g., GM’s conservatism relative to Canadian context).
• Scope of national comparison: Only Canada and the U.S. were compared; more divergent contexts might yield different patterns.
• Keyword-based sampling: Reliance on the “economic inequality” keyword may exclude relevant articles framed without that term, potentially biasing coverage of related issues.
• Movement characteristics not modeled: The study does not vary or analyze specific social movement strategies or attributes that might condition framing effects.
• Observational design: Causal inferences about determinants are based on temporal associations and comparisons, not experimental manipulation.