The Index of Economic Disparity: Measuring trends in economic disparity across Canadian Census Subdivisions and rural and urban communities

Territorial inequalities have long been a subject of study and concern in Canada. In the face large structural changes such as industrial shifts decarbonization our economies, there is an urgency to understand design effective policy interventions for those places facing persistent economic decline. This paper shares novel composite index that measures disparity across Canadian Census Subdivisions (CSDs) using data from 2001 through 2016 2011 National Household Survey. Named “Index Economic Disparity,” it comprised equally weighted average four sub-indices assign percentile rankings all CSDs based on whether they experience substantial decline key areas: population, labour force outcomes, working-age share diversity. The variation outcomes geographies—urban rural—highlights importance place-based policies. Les inégalités territoriales sont depuis longtemps un sujet d'étude et de préoccupation au Face à d'importants changements structurels tels que la désindustrialisation décarbonisation nos économies, il est urgent comprendre ces concevoir des politiques efficaces pour les territoires confrontés déclin économique persistant. Cet article présente nouvel indice qui mesure disparités économiques entre subdivisions recensement canadiennes (SDR) l'aide données recensements l'Enquête nationale auprès ménages 2011. Appelé « disparité », composé d'une moyenne également pondérée quatre sous-indices attribuent rangs centiles toutes SDR selon l'intensité leur dans domaines clés: conséquences sur le marché du travail, part population en âge travailler diversité industrielle. La résultats zones géographiques - urbaines rurales souligne l'importance mettre place axées spécificités locales. Increasing territorial inequality its potential undermine cohesion competitiveness growing (Alasia, 1996, 2004, 2005; Alasia al., 2008). Across countries Organisation Co-operation Development (OECD), gaps small regions increased since 2000, with divides between cities rural communities (OECD, 2020). Canada, study, including deepening rural-urban east-west (Breau & Saillant, 2016). Mechanisms overcome are baked into federation—from equalization payments Canada Health Social transfers. However, economies (that will bring significant socio-economic change different regions) divergences address “left behind places”—those (Krawchenko Gordon, 2021; Martins, 2021). contributes this literature by sharing Survey (NHS) (see Appendix). final output comes two variations: i) provincially/territorially scaled highlights within-province/territory disparity, ii) nationally pan-Canadian disparity. Both outputs use Index Remoteness (Alasia 2017) explore relationship remoteness construction involves great deal choice about what geographies/scales which variables include. Disparity offers synthesized way depict relative trends aims be useable policymakers community members. term “disparity” used indicate differences contrast “inequality” implies unfairness injustice. proceeds parts: (i) brief review relevant order situate index, (ii) methodology intended Disparity, (iii) results analysis, (iv) limitations conclusions. challenge entrenched makes daily headlines main regional development today (Vodden 2019). There notable divide, parts demonstrating poorer range indicators compared their urban counterparts 2016; Desjardins, 2011). Much scholarship topic has tended focus in-depth case studies disparities transitions specific (Brannen 2009; Frank 2014; Zadek, At same time, within systems, hollowing out downtown cores (Hollander 2018) phenomenon shrinking linked resource (Hartt, Jakar Dunn, 2019; Martinez-Fernandez 2012). These complex phenomena variability terms how these play geographies. Pan-Canadian (covering jurisdictions) less common very few indices (Oliveira 2022). Some recent examples include Marginalization (Matheson 2012); Community Wellbeing (Government 2020); Rural Capacity (Memorial University, 2022); Multiple Deprivation (Statistics 2019); Vulnerability Each serves purposes—none directly understanding disparities, but some (e.g., dependency, material resources, activity income). Territorially can usefully provide snapshot conditions help inform policymaking. grounded basic communities. It conceptually underpinned resilience—a multi-dimensional concept applied geographic scales seeks nature (shocks stressors) responses (recovery) (Dinh Pearson, 2015; Martin Sunley, 2015). measure many debates operationalize it. creation limited availability required trade-off depth breadth possible dataset. resulting outlined below, focusses trends, outcome industry diversity province/territory average. interpreted stressor where (the following section explains each sub-index detail). overall purpose facilitate comparisons ranking, aggregating, simplifying measurements, no single indicator capture created simplified robust picture intra-provincial/territorial differences. meant easy replicate, keep up date, communicate decision makers. captures divergent at CSD level. chosen scale reasons. First, align locally determined municipalities, well Indigenous communities; therefore, geographically boundaries levels government. Second, preferred among Remoteness, measured level 2017). indicates where, other CSDs, community's well-being trending. As performance, factors important: magnitude consistency trends. comprise constructed mind. percent variable weighs number times experienced one census period next. Whereas over weights determine caused anomalous period, or consistent trend. We opt more heavily weight we believe adaptive failure insofar and/or not realized reverse trend, even temporarily. particular, wish identify characterized places” (Martins, weighting reduces census-year dependence scores. Without weight, boom/bust may upper lower extremes score distribution depending most census-period was taken during favourable unfavourable year. designed national CSDs. assigned ranks desired geography (province/territory national), then score, again ranked percentiles ensure uniform distribution. objective localities experiencing relatively important than others, justifies rather raw values. variations Disparity: province/territory, provinces/territories; ranks. total All normalized 0 1 Disparity) assigning integer 5. short-form long-form Population (2001, 2006, 2011, 2016), NHS. Only present least year included index. For completeness, “present” both missing any necessary cannot terminus sub-index. compares so needed. Every effort made choose standardized periods, although certain cases, measurement name. best proxy available previous censuses. merged proximity centres, services, absolute A higher remote. paper, classified above threshold. robustness, consider multiple thresholds. longevity decreases either province/territory. decrease 1996 2016. Trend Sub-Index Ψ csd ${{\rm{\Psi }}}_{{csd}}$ inputs final, X ∈ ( 0,1 ) ${X}_{{csd}}\in (\mathrm{0,1})$ strongly consistently dependency ratio increased. that, rest province country, outside increasingly increasing CSD. Λ ${{\rm{\Lambda }}}_{{csd}}\in continuity CSD's LFP increases unemployment scores highly strong country. Industry Diversity describes degree specialization give diversified trending away diversification. encompasses considerations: dominant industry, domination, trend share, dependent becoming sector economy, diversifying. While specialized economy contribute growth, also risk factor fragility because volatility (O'Hagan Cecil, 2007). small, rely portions ill-equipped transition new industries, especially if infrastructure alternative industries limited. Particularly agriculture-based exogenous shocks commodity prices weather cause negative impacts welfare when employ laid-off workforce. reason, unlike sub-indices, specialization, just consideration measuring robustness. Φ ${{\rm{\Phi magnitude, persistency If high, means combination towards further force. differs marginally therefore requires additional justification. mentioned, others S ¯ $\bar{{S}_{{csd}}})$ , associated high particularly fortunes (Baldwin Brown, 2004). C ind ${C}_{{csd}}^{{ind}}$ M xmlns="http://www.w3.org/1998/Math/MathML">M}_{{csd}}^{{ind}}$ existence direction approaches indices, minimize personal bias full transparency, follow advice OECD Handbook constructing 2008) weigh major strategy without thorough theoretical justification, approach while being transparent. When sample-size versus detail trade-off. 5,162 unique (of 5,093 2011), compute fewer. Initially, income real median earnings, government inclusion decreased sample size 29%. Income Labour Dependency Indices were correlated. Intuitively, correlation could due aging dependents needing transfers (thus raising score), rising decreasing such, clear omit underlying heavily. Conveniently, responsible largest values total, 4,301 calculate sub-indices. section, interpret lay guidelines usage inference. 5 severity Low decline: far evidence, weaker stable. Low-Moderate Moderate moderate Moderate-High High circumstances, “relatively” refers appropriate region (provincial/territorial national). ranks, unfavourably declining faster persistently tend fall right tail (disparity 0.6 above). favourably respect concentrated left 0.4 below). Appropriate should adhered detailed interpretation. Foremost, internally heterogenous, rates decline/growth. only representative aggregate, necessarily smaller geographies suggest users treat similarly similar interventions. circumstances similarly, (1 5) wide bins conditions. performing poorly scored example put them category. Further, interpretation left-hand inverse right-hand tail. 4 observed 2 inconsistent quality issues easily identified data. regard, evidence behind,” (communities positively well-off). With caveats explicitly stated, flag exhibiting Statistics rurality Score, available. assigns 1, center. several distinct advantages make useful like Metropolitan Influenced Zones (MIZ) Consolidation Subdivisions. level, allows straightforward merger Disparity. Areas (CMAs). MIZ defines integration CMAs, meaning municipalities integrated CMA described 2018). considers access services centers, CMAs. al. (2017) comprehensively discuss After normalizing gives robust, continuous (and of) centres. includes Census, enumerated. 861 divided groups: updated (69 CSDs), sufficient (792 CSDs). Analysis category mentioned section. maps below provinces territories (Figure 1a) 1b). majority lives close border United States Canada's south (due larger population) exhibit Descriptive statistics Figure 2a 2d summarizes 5), (0 1) adjusted scores, rare. c) 2b d), roughly centered 0.5 3 respectively. 3) suggesting diversity, share. Table descriptive summary provincial/territorial tends worth noting b shows intra-provincial slightly standard deviation integer. explanations result. first provinces/territories, economically secure second explanation expected. Score (either Ordinary Least Squares regression produced remoteness. summarized Section interaction Sections D 1. (rural) (urban) threshold separated. column six, 90th (0.5809) defining urban/rural 0.5809 “urban,” equal defined “rural.” row, respective classification, given greater neither nor linear Moreover, “remote” small. reaches percentile, 75th percentile. remote (90th percentile) result result, challenges notion homogenous On average, marginal linear. another look grouped Integer created: 2, 3, mean calculated reported. bin had bin. positive score. sense, more-remote (on average) provincial interesting table discrepancy 1/2/3 4/5 leads determinant ${{Remoteness}}_{{csd}}$ CSD, i s p r t y c o e $Disparity{\ }Scor{e}_{{csd}}$ raw, reveal weakness relationsh