The (In)Coherence of Canadian Refugee Education Policy with the United Nations’ Strategy

This study assesses the coherence of Canada’s educational policy regime with the United Nations High Commissioner for Refugees’ (UNHCR) Refugee Education 2030 strategy. We articulate a theoretical framework that combines theories about policy coherence, policy attributes, and policy tools, which informs a two-phase methodology. First, we conducted jurisdiction-based scoping reviews of policies in Canada’s 13 provinces and territories which have constitutional authority over education. This yielded a sample of 155 documents, which we then analyzed for its vertical coherence with Refugee Education 2030. Our analysis focused on five categories of need in the UNHCR strategy with respect to refugee students, namely access to education, accelerated education, language education, mental health and psychosocial support, and special education. The findings reveal there are policies across Canada that target responses to the five categories of need. Although some policies are exemplary in their coherence with Refugee Education 2030, Canada’s refugee Education Policy Analysis Archives Vol. 30 No. 39 2 education policy regime is characterized by many inconsistencies and significant gaps. Policymakers in Canada could use the specific findings to develop or revise policies to address shortcomings. Researchers and policymakers in other countries who find value in our approach could replicate the study’s method in their own jurisdictions, using the instruments provided in appendices to identify strengths and gaps.


The (In)Coherence of Canadian Refugee Education Policy with the United Nations' Strategy
Issues related to students with refugee protection (SwRP) and students seeking refugee protection (SsRP) 1 in educational systems in major resettlement countries like Canada have shifted in recent decades, with unprecedented influxes of asylum seekers and refugees arriving via established and non-traditional pathways (United Nations High Commissioner for Refugees [UNHCR], 2015;. Most traditional resettlement countries have well-resourced education systems relative to the countries of origin and transition countries of many people seeking and granted refugee protection (RP). But those systems can still pose barriers to the education of SwRP and SsRP (UNHCR, 2019a), despite their defined right to education under the 1951 Convention relating to the Status of Refugees and its 1967 Protocol (Office of the High Commissioner for Human Rights [OHCHR], 1967;UNHCR, 1951). 2 As part of these dynamics, UNHCR (2020) noted that, of the 26 million people with RP worldwide, 43% are school-aged (under the age of 18); yet, in comparison to global rates of primary and secondary enrolment of 91% and 84% respectively, the rates for children and youth with RP are only 63% and 24% (UNHCR, 2019a).
In response to what can be called a crisis in refugee education, UNHCR recently released Refugee Education 2030, a global strategy and framework to guide the development of country-specific education policies. 3 Refugee Education 2030 calls for governments to "establish dedicated policy 1 We thank one of the reviewers recommended we use the term "students from refugee backgrounds" to identify the population of concern. We agree that is an accepted term, but we chose to use the terms "students with refugee protection" and "students seeking refugee protection" because they reflect the circumstances and entitlements of individuals legally recognized as refugees and individuals not (yet) recognized as refugees. This is relevant to our study because it reflects the fundamental distinction in Canada's refugee system between individuals who seek refugee protection from outside Canada, those who come to Canada after being granted refugee status (i.e., resettled refugees), and those who make refugee protection claims from within Canada (i.e., refugee claimants) (Government of Canada, 2021a). 2 A convention refugee is a person who, "owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return" (UNHCR, 1951, Article 1(2)). The 1967 Protocol extended the scope of its applicability to individuals forcibly displaced worldwide (UNHCR, 2011). The Refugee Convention guarantees individuals with refugee status "the same treatment [as that] accorded to nationals with respect to elementary education" and "treatment as favourable as possible… with respect to education other than elementary education" (UNHCR, 1951, Article 22). Furthermore, the Convention on the Rights of the Child affirms the rights of all children, including refugee children, to free and compulsory primary education, to accessible secondary education, and to higher education on the basis of capacity (OHCHR, 1990, Article 28). 3 Refugee Education 2030 provides an action plan for the Global Compact on Refugees (UNHCR, 2019b), a nonbinding international agreement adopted on December 17, 2018 that identifies categories in need of support, one of which is education (notably in Articles 68 and 69) (UNHCR, 2018). regarding refugee-inclusive national education systems" (UNHCR, 2019b, p. 33) with education policies being "proactive and explicit on [the] inclusion of refugees" (UNHCR, 2019b, p. 33) to address "the particular learning needs of refugee students" (UNHCR, 2019b, p. 2). The strategy aims to have SwRP and SsRP achieving parity with their non-refugee peers in pre-primary, primary and secondary education, and to boost enrolment in higher education to 15% by 2030 (UNHCR, 2019b). A major concern for UNHCR (2019c) is that "national and regional education policies, plans and programmes…incorporate refugees" (p. 6) through the implementation of policies that are coherent within and across educational jurisdictions in each country. This is because incoherence and "policy gaps" have major implications SwRP and SsRP education pathways and subsequent life chances (UNHCR, 2019c, p. 37).
Previous studies have examined select refugee education policies in host countries neighboring on the countries of origin of SwRP and SsRP (e.g., Iran, Kenya, Lebanon, Rwanda, Uganda) with respect to access to national education systems of SwRP, with the latter often from one country of origin (e.g., the Syrian Arab Republic) Beltekin, 2016;Buckner et al., 2017;Dryden-Peterson et al., 2019;Hamadeh, 2019). Studies on traditional resettlement countries like Australia and Canada assert there is a lack of education policy addressing SwRP and support UNHCR's claim that the reliance on general education policies, in which students with refugee backgrounds are aggregated with other cohorts, can lead to inappropriate responses and systemic barriers to the learning and success of SwRP and SsRP (Brewer, 2018;Kanu, 2008;Matthews, 2019;Miller et al., 2018;Shakya et al., 2010;Wilkinson, 2002).
In Canada, studies have documented some of the unintended consequences that emerge when SwRP and SsRP are covered by general, rather than refugee-specific, education policies. They indicated that some SwRP were placed in grades or academic streams inappropriate for their ages, maturity levels, capabilities, and aspirations (Shakya et al., 2010;Wilkinson, 2002;Yau, 1996). Needs with respect to language learning, accelerated education, and special education were confounded in Canadian schools, resulting for example in SwRP with language learning needs being placed in academic streams below their academic proficiencies, and those with missed education being placed in special education programs despite having no diagnosed conditions for such placement (Kaprielian-Churchill, 1996;Wilkinson, 2002;Yau, 1996). MacNevin (2012) explained how SwRP who have missed education and therefore typically have had less formal literacy instruction are placed in language education programs that assume they have robust literacy skills that can be transferred to the learning of the new language. MacNevin showed this to be one unintended consequence of not providing refugee-specific policy instruments and/or appropriate funding to provide professional development to educators to teach beginner literacy skills to older students using appropriate materials.
These studies highlight that a lack of specific recognition and needed supports for SwRP and SsRP obstructs them from completing secondary school, being eligible for higher education, and using their education to pursue desired futures (Shakya et al., 2010;Wilkinson, 2002). 4 They also indicate it is not sufficient to simply include the mention of SwRP and SsRP in general policies, but that educational policies explicitly dedicated to SwRP and SsRP are needed. However, little is known about how educational policies in major resettlement countries have created or adapted their policy designs to address specific issues related to the increased numbers and mobility of people with and seeking RP, and to what extent refugee education policies in each jurisdiction correspond coherently to UNHCR's guiding framework that targets explicit inclusion of SwRP and SsRP.

Purpose of this Study
Our purpose was to analyze the vertical coherence (Carbone, 2009) of current policies in Canada towards the primary and secondary education of SwRP and SsRP with Refugee Education 2030. Specifically, we sought to determine the extent to which policies aim to realize "Expected Result 1" of "Strategic Objective 2" of Refugee Education 2030 while using the "Enabling Activities" and "Strategic Approaches" advised by the UNHCR (2019b, p. 25). Table 1 summarizes this key section of the UNHCR strategy. We selected Refugee Education 2030 as the key reference document for this policy coherence study for two reasons. First, UNHCR's holds a central position in refugee education due to its longstanding global mandate to protect the rights of refugees (UNHCR, 2013), including the right to education. The latter is recognized as a fundamental right in the 1951 Convention Relating to the Status of Refugees (UNHCR, 1951) that is one of the core policy instruments of international refugee law. As 5 In Refugee Education 2030, a hyperlink is attached to the phrase "social and emotional learning", which goes to background paper of the Inter-Agency Network for Education in Emergencies (INEE) titled Psychosocial support and social and emotional learning for children and youth in emergency settings (INEE, 2016). Social and emotional learning is defined there as "a process of acquiring core competencies to recognize and manage emotions, set and achieve goals, appreciate the perspectives of others, establish and maintain positive relationships, make responsible decisions, and handle interpersonal situations constructively" (p. 10). This learning aims to foster the development of five interrelated sets of competencies: self-awareness, self-management social awareness, relationship skills, and responsible decision-making (p. 10).
UNHCR's exercise of its mandate presupposes commitments from and cooperation with states, other international agencies, and non-governmental organizations (UNHCR, 2013), it has been establishing global refugee education policy, coordinating with governments around national refugee education policy development and implementation, and collaborating with organizations involved in refugee education since the 1960s (Dryden-Peterson, 2016a). Recent years have seen increased involvement of numerous international and multilateral organizations in refugee education. The positioning of these actors with respect to refugee education is ongoing, and UNHCR continues to feature centrally. Collaboration and coordination with UNHCR is a common feature of the diverse initiatives undertaken by the numerous other global institutions involved in refugee education policy, finance, and praxis (i.e., UNICEF, UNESCO, the World Bank, Education Cannot Wait [ECW], the Organisation for Economic Co-operation and Development [OECD], and the Global Partnership for Education) (see ECW, 2019;Hanafi et al., 2021;OECD, 2020;UNHCR & UNICEF, 2020;UNESCO, 2020;UNHCR & GPE, 2016). Second, from a normative standpoint, the refugee education policies of UN member states should be coherent with Refugee Education 2030 because this strategy is inscribed within the 2030 Agenda for Sustainable Development adopted by all 193 UN member states (UN, 2015;UNHCR, 2019b), including Canada (Government of Canada, 2021a). 6 We selected Expected Result 1 of Strategic Objective 2 of Refugee Education 2030 as the specific reference point because it articulates the core goal of education, which is for children and youth to learn and succeed. 7 Our study is thus concerned with the vertical policy coherence between UNHCR's international refugee education as a key policy reference framework and the assessment of refugee education policy designs enacted in Canada as a UN member state. We selected Canada because 20% of all refugees resettled globally between 2010 and 2020 were received by Canada (UNHCR, 2020). For constitutional reasons, Canada does not have a centralized, national system of education or a federal department of education. Rather, Canada's 10 provinces and three territories have the responsibility for education, and each has its own ministry that establishes policies for primary and secondary education (Council of Ministers of Education, Canada, 2020). As each of Canada's 13 educational jurisdictions has its own education policy set, we define vertical coherence with Refugee Education 2030 as the coherence of sets of Canadian refugee education policies that if properly designed, can potentially achieve larger goals of the Enabling Activities of Expected Result 1 of Strategic Objective 2 and the Strategic Approaches of Refugee Education 2030.

Theoretical Framework
We begin by situating this study within the fields of educational policy analysis and political science. Policy coherence has been studied in the field of educational policy analysis, where at least three perspectives exist. Some researchers have viewed coherence as a problem of policy design by top government officials-mainly district central offices and state and federal agencies-with such studies typically looking at the external, top-down alignment into schools (e.g., Fuhrman,1993). Other researchers have taken a bottom-up approach to finding solutions within schools and 6 Sustainable Development Goal 4 (SDG4) of the 2030 Agenda is for free primary and secondary education for all (UNGA, 2015). By virtue of the Incheon Declaration for the Implementation of SDG4 (UNESCO, 2016), refugee children and youth are included among the "all boys and girls", "all youth", and "all men and women" targeted by SDG4 (UNGA, 2015, p. 17). Canada is legally obligated to provide education to refugee children and youth: It acceded to both the 1951 Refugee Convention and its Additional Protocol in 1967 and ratified the Convention on the Rights of the Child in 1991 (Government of Canada, 2017). 7 The Enabling Activities of Refugee Education 2030 align with all the education-related aims of the Global Compact on Refugees (UNHCR, 2019b). engaging school leaders in setting their own goals and improvement strategies that fit local contexts (e.g., Honig & Hatch, 2004). Still others have focused on the internal coherence of curriculum policy documents as a technical problem of aligning standards, curricula, and assessments (e.g., Bateman et al., 2007).
In the present study, we chose to integrate the literature from political science to look at policy coherence as a "process where policy makers design a set of policies in a way that, if properly implemented, they can potentially achieve a larger goal" (Cejudo & Michel, 2017, p.750). This view encourages a focus on how each educational policy is part of an existing set that constitutes a whole. Thus, using specific criteria and a systematic methodology outlined below, we analyze and assess the overall coherence of sets of provincial and territorial policies in Canada that seek to support, coordinate and achieve the full integration of SwRP and SsRP using the global strategy Refugee Education 2030 as the key policy reference. We chose the latter based on some authors' claims of the absence of overarching policy frameworks for refugee education in traditional resettlement countries resulting in fragmented education policies and activities (Brewer, 2018;Christie & Sidhu, 2006;Matthews, 2008;Miller et al., 2018). This choice allowed us to compare and assess the connectedness and associations of ideas promoted by the international governing body with existing sets of policy designs in Canada.
Our theoretical framework is summarized in Figure 1. It combines concepts about policy coherence to assess the vertical coherence 8 of the of policy sets in Canadian provinces and territories 9 relative to Expected Result 1 of Strategic Objective 2 of Refugee Education 2030, considering the related Enabling Activities and Strategic Approaches and five categories of need inherent in the UNHCR strategy. Specifically, we draw on Cejudo and Michel's (2017) policy coherence framework, which conceives the coherence of policies within a domain (e.g., refugee education) as a function of 1) coherence among policies' objectives; 2) coherence among policies' targeted populations; and 3) coherence among policies' instruments. Cejude and Michel define these three elements as 1) "the consistency between the individual objectives of the policies that coexist within the same policy domain" (p. 755), 2) "the sum of all the people targeted includes the entire policy domain's target population" (p. 755), 3) "the way [policies] are designed, to solve the same public problem with different tools" (p.755). Our definition of policy coherence incorporates these three elements.
We also integrate Desimone's (2002) and Porter's (1994) work on policy attributes. These authors observed that successful policy implementation depends on the interactions of interrelated attributes of policies, of which we concentrate on consistency and specificity. Consistency designates the extent to which various policies contradict or reinforce each other (Desimone, 2002;Porter, 1994), with the latter scenario allowing for greater coherence. Specificity refers to "how extensive and detailed a policy is" (Desimone, 2002, p. 438). 8 Vertical policy coherence is one of four types identified by Carbone (2009). The others are donor-recipient, multilateral, and horizontal coherence. Although "vertical" typically refers to coherence of policies at different levels of government within a country (Steiner-Khamsi, 2012), we extend the definition to the level of global governance, with the UN at its pinnacle.

Theoretical Frameworks Defining Policy Coherence and Attributes across Five Categories of Need Inherent in Refugee Education 2030
Consequently, the first element, coherence among objectives, refers to the consistency between the specific objectives of individual policies coexisting in each jurisdiction with the overall objectives of the policy, with the latter referenced to Refugee Education 2030. 10 The second element, coherence among targeted populations, is understood as the consistency of the specific populations targeted by all the policies coexisting in each jurisdiction, referenced to the populations targeted by Refugee Education 2030: those who benefit, those who implement, and those who partner. 11 The third element, coherence among instruments, is defined as the consistency of the use of capacity building and learning instruments in the individual policies coexisting in each jurisdiction, again referenced to Refugee Education 2030. Policy instruments are techniques governments use to prompt targeted implementation agents to act in ways consistent with a policy objective, and policies must use different yet complementary instruments to motivate and enable heterogeneous individuals in different situations to take a range of actions to address the same objective (Schneider & Ingram,10 In each Canadian educational jurisdiction, policies that had objectives of responding to one category of need of refugee students were aggregated. Coherence among objectives is therefore primarily concerned with the consistency of the aggregation of objectives of the policies with the relevant area of need. 11 In Refugee Education 2030, the targeted benefiting populations are refugee and refugee claimant children and youth of primary and secondary school age (UNHCR, 2019b, p. 7). The targeted implementation agents are government agents, such as teachers, administrators, and support personnel (UNHCR, 2019b, p. 33). The targeted partners include organizational partners (UNHCR, 2019b, p. 33) and partners from "a whole of society" perspective (UNHCR, 2019b, p. 15). Organizational partners may include intergovernmental, international non-governmental, civil society, private sector, and/or academic organizations (UNHCR, 2019b, p. 33-37). We interpreted partners from "a whole of society" perspective (UNHCR, 2019b, p. 15) as being educational stakeholders such as students, parents/guardians, families, and community members. ). 12 In this study, policy instruments are defined as the capacity building (i.e., provision of information, training, education, resources) and learning (i.e., open-ended derivation of approaches and activities, evaluation) instruments because they are at the heart of the strategic objective of Refugee Education 2030 that serves as the touchstone for this study (i.e., Strategic Objective 2). They encourage targeted implementation agents to devise approaches and undertake activities towards an objective, and to evaluate progress towards the objective. 13

Methodology
This study sought to answer: To what extent are Canadian provincial and territorial education policies coherent with the Enabling Activities for Expected Result 1 of Strategic Objective 2 and the Strategic Approaches in Refugee Education 2030? This question built on findings from a prior study that triangulated the five categories of need indicated in the UNHCR document with empirical studies (see Table 2) and confirmed the presence of these needs in the Canadian context (Schutte, 2020). That prior study also highlighted unique characteristics based on the distinction between children and youth seeking refugee protection (RP) in Canada and those granted RP before arriving, and on the population distribution in Canada of children and youth seeking RP and with RP (see Appendix 1), including that most reside in five provinces (AB, BC, MB, ON, QC; Schutte, 2020). 14 The present study proceeded in two steps: first by conducting jurisdiction-based scoping reviews of policies in Canada's 13 provinces and territories; then by analyzing the sample of policies for their vertical coherence with Refugee Education 2030, focusing on policies addressing the five categories of needs.

Step 1: Jurisdiction-Based Scoping Reviews of Policies
To identify relevant policy documents, we conducted jurisdiction-based scoping reviews of policies (Hare et al., 2016) based on the PRISMA-ScR checklist and explanation (Tricco et al., 2018) using the websites of provincial and territorial governments. Based on our review of existing scholarship on policy coherence, a systematic approach using PRISMA-ScR had not previously been used. We needed to create and follow our own strategy (see Appendix 2). We performed English and French language searches in March 2020. We also identified additional documents through external websites linked to government webpages and from an appendix of a grey literature review (Ratković 12 Schneider and Ingram (1990) identified five types of policy instruments: authority (i.e., permission, prohibition, or mandatory action), incentives (i.e., inducements, charges, sanctions, force), symbolic and hortatory gestures (i.e., pronouncements, rationales, labeling), capacity building (i.e., information, training, education, resources), and learning (i.e., open-ended derivation of approaches and activities, evaluation). 13 There is a balance to be struck in the application of the combined concepts in our theoretical framework in addressing five categories of need and focusing on the independence or intersectionality of the individual policies comprising each jurisdiction's set of policies. For example, policies with respect to special education may need to consider the proficiency in the language of instruction of SwRP and SsRP who have special needs-that is, some SwRP and SsRP with special needs may be proficient in the language of instruction while others may need to learn the language and may thus be further disadvantaged. 14 This prior study was based on publicly available statistics about refugee claimants 14 and refugees in Canada from January 2015 to March 2020. Datasets on refugee claimants were sourced from Open Government and from the website of the Immigration and Refugee Board of Canada [IRB] (IRB, 2020a). These datasets were used in combination with datasets published by global governance institutions, including UN agencies and the World Bank. Where statistical data were not available for a category of needs, empirical studies were sourced and analyzed (see Schutte, 2020Schutte, , p. 10). et al., 2017. We also assessed for inclusion the Education Act or Schools Act of each province and territory. Our scoping review revealed that Canada's refugee education policies comprised 155 policy documents as of March 2020. Of these documents, 117 were in English and 38 were in French, and they came from 11 jurisdictions, with the remaining 2 jurisdictions (NU, YK) lacking any documents (see Appendix 3). As a result of this latter review, our subsequent analyses of the documents often refer to 11 rather than 13 jurisdictions because of the absence of policies in two territories.

Data Charting
The content of each of the 155 eligible policy documents was charted using a data abstraction instrument (DAT; see Appendix 5). The first section of the DAT was used to record the characteristics of the policy document (i.e., jurisdiction, date of publication, date of effect, type of policy document) and to summarize its contents. The second section was used to record data about the policy objectives that pertained to each of the five categories of need (i.e., access to education, accelerated education, language education, mental health and psychosocial support, and special education), the targeted populations, and the instrument(s) associated with each objective. Each objective and its corresponding targeted populations and instrument(s) were grouped as one policy data item unit, as conceptualized in Figure 3.
The final section of the DAT was used to record considerations relating to the intersectionality of objectives because a single policy document could include multiple objectives pertaining to different categories of needs. We used NVIVO TM data analysis software to code all the policy data items in each DAT and produced a master codebook (see Appendix 6) that could be disaggregated into codebooks by jurisdiction as required. We then extracted all policy data item units in each jurisdiction and sorted them in a series of policy analysis matrices that addressed each of the elements of policy coherence -i.e., matrix 1 brought together data on policy objectives, matrices 2, 3 and 4 on targeted populations, and matrix 5 on instruments, again conceptualized in Figure 3. Each series of matrices addressed one of the five categories of need, such that each jurisdiction's data was sorted into five completed series.

Analysis of the Policy Coherence Among Objectives
We used matrix 1 to organize data on objectives from each jurisdiction's refugee education policies based on each of the five categories of need, and then analyzed for coherence in term of consistency between specific objectives of individual policies that coexist within the same category of need and the overall objective related to that category of need drawn from the relevant Enabling Activity from Refugee Education 2030. An objective was categorized in the DAT as "high-level" if it was a main objective in a policy document or as "low-level" if it was secondary. 16 We assigned "not applicable" when documents did not have any objectives pertaining to a category of need. When "not applicable" was assigned to a category of need for all documents in a jurisdiction, we determined that to signify a "policy gap" (i.e., an absence of policy objectives for addressing that need). When documents from a jurisdiction had high-and/or low-level objectives with respect to a category of need, we analyzed these to assess their consistency (as the measure of coherence) or lack thereof (as the measure of incoherence) with the overall objective related to that category of need. We used a policy coherence rubric to make these assessments (see Appendix 7).

Analysis of the Policy Coherence Among Populations
We used three matrices to analyze coherence among the three populations identified in Refugee Education 2030: those who benefit, those who implement, and those who partner (see Figure  3). 16 For example, a language education policy with a subsection on promoting the mental health of language learners would have objectives pertaining to language education as high-level objectives, while objectives pertaining to mental health and psychosocial support were deemed low-level. Refugee Education 2030 describes the targeted benefiting population to be children of primary and secondary school age with or seeking RP "regardless of legal status, gender or disability" (UNHCR, 2019b, p. 43). With this definition in mind, we used matrix 2 to record and analyze data on the benefiting populations articulated in individual policies. To facilitate analyses of the specificity by which SwSR and SsRP are targeted as beneficiaries, we incorporated three categories into the matrix "refugee-focused", "refugee-relevant", and "refugee-mentioning". 17 Refugee-focused policies are those in which SwRP and/or SsRP are the targeted benefiting population. 18 Refugee-relevant policies are those in which such students are not the primary benefiting population but are named as a cohort of a broader targeted population. Refugee-mentioning policies are those in which the term "refugee" appears, but not in the context of explicitly identifying SwRP and SsRP to be the targeted population.
When inputting a benefiting population in the matrix, specified characteristics were recorded. Here, age, gender, and ability were used to analyze consistency with Refugee Education 2030, and other specified characteristics (e.g., countries of origin, refugee pathway, immigration category, etc.) were used to analyze the inclusivity of policies with respect to all SwRP and SsRP or their exclusivity to specific cohorts. We used matrix 3 to record and analyze data about targeted implementation agents. Here we categorized implementing populations as "government agents" and as "not government agents". Matrix 4 was used to record and analyze data with respect to partnering populations. We sourced two categories of targeted partners from Refugee Education 2030: "organizational partners" and "partners from a whole of society approach". The former was to include intergovernmental organizations, international non-governmental organizations, civil society organizations, private sector organizations and foundations, and academic networks (UNHCR, 2019b, p. 34-37) and the latter included individual stakeholders from Canadian civil society, such as student, parents, guardians, families, and communities (UNHCR, 2019b, p. 14). Once these data were organized in matrices, we used a policy coherence rubric (see Appendix 7) to assess the coherence, incoherence, and gaps in policies with respect to targeted populations in each jurisdiction.

Analysis of the Policy Coherence Among Instruments
A fifth matrix and rubric (see Appendix 7) were used to analyze the coherence of instruments addressing each category of need with the Strategic Approaches of Collaborative Learning and Capacity Development and Innovation, Evidence, and Growth in Refugee Education 2030.

Synt esis o t e o eren e o anada's Re ugee Edu ation o i y Sets
The data organized in the preceding matrices for all jurisdictions were compiled into a single synthesis matrix (see Table 4). This provided for an assessment of the overall coherence of the refugee education policy sets of all provinces and territories based on the aggregation of data for each jurisdiction with respect to the five categories of need, the three main elements of policy coherence (i.e., objectives, populations, instruments), and the coherence level of the combined policies (i.e., their coherence, incoherence, and absences). Note. Org. = organizational partners; WoS = partners from a whole of society perspective.
We deemed policies addressing a category of need to be coherent with Refugee Education 2030 for each element of policy coherence when they met all the criteria in the theoretical framework for that element. Policies were assessed as incoherent when: a) they did not fulfill the criteria in the theoretical framework for an element of policy coherence (e.g., the targeted benefiting population of the policies explicitly includes SwRP but not SsRP); b) they contradicted criteria in the theoretical framework for an element of policy coherence (e.g., the targeted benefiting population excludes SsRP); or c) they obstructed coherence with criteria in the theoretical framework with respect to a different element of policy coherence (e.g., capacity instruments, if implemented, would prevent certain SwRP and SsRP from being included in the targeted benefiting population). Where no relevant policies were identified, the coherence level was classified as absent, denoting a clear policy gap.
Our use of this synthesis matrix occurred in two steps. We inputted the official English alpha codes of the provinces and territories to indicate that we assessed their policies to be coherent, incoherent, or absent with respect to each element of policy coherence. Once we had classified all 13 jurisdictions, we replaced the alpha codes by the sum of the number of jurisdictions they represented. This then indicated the number of jurisdictions with coherent, incoherent, or absent refugee education policies. Consequently, the sum for each element of coherence related to each category of need is 13.

Findings
The Canadian refugee education policy sets address the five of the categories of needs of SwRP and SsRP derived from Refugee Education 2030, but the vertical coherence of policies with UNHCR varies across the 11 educational jurisdictions in Canada that have policies as per our sample (see Table 5). In what follows, we describe the extent to which Canadian educational jurisdictions respond to the five categories of needs through policies coherent with Refugee Education 2030.

Access to Education
Refugee Education 2030 states SwRP and SsRP are to "receive any supports required to enable their access to the education system, including assistive technology and accessible learning materials" (UNHCR, 2019b, p. 43). 19 Such policy objectives are absent in four of Canada's educational jurisdictions (NB, NS, NU, YK), marking a significant policy gap. Nine jurisdictions have refugee education policies with objectives of enabling access to public primary and secondary education systems (AB, BC, MB, NL, NT, ON, PE, QC, SK). All nine have objectives of providing accessible learning materials, but only three also have objectives of enabling access through the provision of assistive technologies (AB, ON, SK). We thus deemed these three to be coherent with UNHCR objectives in this category of need and determined the other six jurisdictions to be incoherent because they did not mention or prescribe the use of assistive technologies.
Of the nine jurisdictions with objectives of enabling access, we deemed the targeted benefiting populations of six to be coherent with those of Refugee Education 2030 (AB, BC, MB, NT, ON, SK; see Table 4). Policies in ON seemed exemplary in their coherence with the targeted population of Refugee Education 2030 because they addressed the important clause about access "regardless of legal status, gender, or disability" (UNHCR, 2019b, p. 43). n addition to guaranteeing access to children who themselves or whose legal guardian(s) have or are seeking refugee protection, in ON, access is guaranteed to minors who themselves or whose legal guardian(s) are in Canada unlawfully.
We deemed the policies about access in three jurisdictions (NL, PE, QC) to be incoherent with Refugee Education 2030 with respect to targeted benefiting populations because they do not fulfill all the criteria: NL guarantees access only to government-assisted refugees as permanent residents; PE guarantees access only to SwRP, not SsRP; and QC guarantees access only to minors whose legal guardians have RS, not to minors who themselves have or are seeking RS. 20 Eight of the nine jurisdictions with refugee education policies pertaining to education access have targeted implementation agents that we deemed coherent with Refugee Education 2030 (AB, BC, MB, NL, NT, ON, QC, SK). The targeted implementation agent in the ninth jurisdiction (PE) was deemed incoherent because parents and guardians are targeted for policy implementation, but government agents are not. Seven of these nine jurisdictions were assessed to be coherent in terms of targeted partners (BC, MB, NL, ON, PE, QC, SK), while we considered two incoherent because their policies targeted stakeholders but not organizational partners (AB, NT).
Of the nine jurisdictions with access policies, six have capacity instruments in this category of need that we judged to be coherent with the Strategic Approach of Collaborative Learning and Capacity Development in Refugee Education 2030 (AB, BC, MB, NL, ON, SK). We deemed as incoherent the capacity instruments in policies of two jurisdictions because they did not include instruments for building the capacity of educational professionals (PE, QC), and noted that capacity instruments are absent in the access policies of one jurisdiction (NT). In terms of learning instruments, we concluded that only one of the nine jurisdictions (MB) with access policies was coherent with Refugee Education 2030, and three were incoherent (ON, QC, SK) because they did not require evaluations of progress towards objectives and that would be needed to support the evidenced-based innovation sought by the Strategic Approach of Innovation, Evidence, and Growth in Refugee Education 2030. Learning instruments were absent in all the other jurisdictions.

Accelerated Education
Refugee Education 2030 states SwRP and SsRP are to be "supported to make up for missed schooling" (UNHCR, 2019b, p. 43). 21 Nine of Canada's 13 educational jurisdictions have objectives of responding to SwRP and SsRP with interrupted formal education. Of these nine, we considered the objectives of five to be coherent and four incoherent with Refugee Education 2030 in this category of need. Of the five jurisdictions with coherent objectives, two have the objective of responding to the needs of students with missed schooling through the provision of regular educational programming with additional supports (AB, BC) and three aim to do so through the provision of accelerated education programs (AEPs; MB, NL, ON). The four jurisdictions with objectives deemed incoherent with Refugee Education 2030 have policies that aim to provide regular educational programming to students with no, limited, or interrupted formal education and do not have objectives for accelerating learning or making adaptations or accommodations specific to those with missed formal schooling (QC, SK, PE, NS).
The targeted benefiting populations of this category of need were assessed as coherent with Refugee Education 2030 only in two jurisdictions (BC, MB). Those in the seven other jurisdictions with accelerated education policies were deemed incoherent because they do not include SsRP. Of these seven, three jurisdictions (N , ON, QC) use the term "refugee camp" to implicitly target SwRP with limited or interrupted formal education. This was an additional reason for determining these policies to be incoherent with Refugee Education 2030 because it excludes many SwRP and SsRP who have not been in camps. We deemed policies in one of these jurisdictions (NL) to be additionally incoherent because it exclusively targets a specific group of SwRP (i.e., those resettled through the governmentsponsorship pathway). All nine of the jurisdictions with refugee education policies pertaining to accelerated education have government actors as targeted implementation agents, which is coherent with Refugee Education 2030. We deemed three of Canada's nine jurisdictions with policies in this category of need to be coherent with Refugee Education 2030's Strategic Approach of Partnership because they target both organizations and educational stakeholders as partners (MB, NL, ON). The remaining six jurisdictions were assessed as incoherent because they do not target organizational partners (AB, BC, NS, PE, QC, SK).
Eight of the nine jurisdictions with refugee education policies addressing accelerated education have capacity instruments for responding to missed education (AB, BC, MB, NL, ON, NS, PE, QC). Seven of these were deemed coherent with Refugee Education 2030 because they had one or more capacity instruments pertaining to the provision of supports to students with missed schooling. One jurisdiction (QC) has a capacity instrument that could be used to identify students with missed schooling but not to support such students in catching up. As a result, this jurisdiction's policies were deemed incoherent in this category. Only three jurisdictions have learning instruments in policies related to accelerated education. Of these, we deemed the learning instruments in only one jurisdiction (MB) to be coherent with Refugee Education 2030 because that province's policy requires that AEPs be developed with the input of all key stakeholders and that organizational partners be involved in program development and implementation. The learning instruments in the other two other jurisdictions (ON, BC) were deemed incoherent because they do not encourage evaluations of progress towards objectives, a component of Refugee Education 2030.

Language Education
Refugee Education 2030 states that SwRP and SsRP are to be "provided with adequate language training where necessary" (UNHCR, 2019b, p. 43). 22 Eight of Canada's 13 educational jurisdictions (AB, BC, MB, NL, ON, PE, QC, SK) have refugee education policy documents with objectives of providing language training in at least one of the official languages of instruction of the jurisdiction, making this element of their policies coherent with the UNHCR direction.
Of the eight jurisdictions with language education policies for SwRP and SsRP, we assessed three to be coherent with respect to the targeted benefiting populations identified in Refugee Education 2030 (AB, BC, MB). The targeted benefiting populations in the other five were deemed incoherent because only SwRP, and not SsRP, are targeted (QC) and they are targeted in exclusionary (i.e., only SwRP resettled through the government-sponsorship pathway in NL) and/or implicit ways (ON, PE, SK). Regarding the latter, for example, SwRP and SsRP are implicitly targeted in policies that refer to students who have spent time in refugee camps; however, those policies do not explicitly target SwRP or SsRP. 23 While we assessed as coherent with Refugee Education 2030 the targeted implementation agents of refugee education language policies in all eight jurisdictions with such policies because they include government agents (AB, BC, MB, NL, ON, PE, QC, SK), the targeted partners of four were assessed as coherent for targeting both organizations and educational stakeholders (MB, NL, ON, SK). The targeted partners of four other jurisdictions were assessed as incoherent. Policies in one jurisdiction (NB) do not contain any references to partnerships. Those in another (BC) do not address partnerships with organizations. Three jurisdictions (AB, QC, PE) have absences in partnerships from a whole of society perspective (missing references to parents, guardians, and/or families).
Six of the eight jurisdictions with policies addressing language education contain capacity instruments for policy implementers to provide training to SwRP and SsRP in the language of instruction. These seven were therefore deemed coherent with Refugee Education 2030 (BC, MB, NL, ON, PE, SK). Of the two jurisdictions deemed to have incoherent policies, one (QC) has instruments building the capacity of administrators to apply for funding for language programs but not the capacity of teachers to teach the language of instruction to SwRP and SsRP, while the other (AB) does not include any capacity instruments related to language learning. Six of the nine jurisdictions with policies for language education have learning instruments that were deemed coherent with UNHCR policy. Learning instruments are absent from the category of language education in one jurisdiction (AB). Learning instruments were deemed incoherent in another (QC) because they do not promote evaluations of progress towards policy objectives.

Mental Health and Psychosocial Support
Refugee Education 2030 states that SwRP and SsRP are to be "provided with conditions that foster social and emotional learning and, where needed, receive mental health and psychosocial support" (UNHCR, 2019b, p. 43). 24 Based on the clauses of this statement, we disaggregated findings into two subcategories: social and emotional learning and mental health support. Another subcategory emerged out of our analysis of findings pertaining social and emotional learning: Community sensitization to refugeehood. We defined this emergent subcategory as the promotion of awareness of refugee experiences and issues within the school community, particularly among students. Our presentation of findings proceeds here via these three subcategories.

Social and Emotional learning
Eleven of Canada's 13 educational jurisdictions have refugee education policies with objectives of fostering social and emotional learning (AB, BC, MB, NB, NL, NS, NT, ON, PE, QC, 23 See three language education policy documents in ON (i.e., ON05, ON07, ON14) in which the word "refugee" appears exclusively in the term "refugee camp". An example from ON05 is "Many newcomer students have arrived in Canada with their families as part of a voluntary, planned immigration process. However, some students have arrived from countries in chaos, have spent time in refugee camps, or have experienced personal trauma caused by natural disaster, political upheaval, or family disruption" (p. 9, italics added). SK). Three of these 11 jurisdictions with such policies have a targeted benefiting population that we deemed coherent with Refugee Education 2030 (AB, MB, QC). We assessed six as incoherent because their policies include SwRP but not SsRP (NL, NS, NT, ON, PE, SK) and/or because their policies focus exclusively on certain SwRP and SsRP (i.e., those who fled from armed conflict in NT; SwRP resettled through the government-sponsorship pathway in NL). SwRP and SsRP are absent from the targeted benefiting populations in the remaining two jurisdictions that have policies with objectives for social and emotional learning (BC, NB). The targeted implementation agents of policies in this category of need were deemed coherent with Refugee Education 2030 in all 11 jurisdictions because government agents are identified. Ten of these jurisdictions have partnering populations that we assessed as coherent with Refugee Education 2030. The policies referencing partnership in one jurisdiction (AB) were deemed incoherent because they are borrowed from a different jurisdiction (MB) and the list of educational stakeholders has not been adapted to the provincial context. All 11 of these jurisdictions also have capacity instruments that we considered coherent with Refugee Education 2030 because they target the capacity of educational professionals to support students' social and emotional learning. Ten of the 11 jurisdictions have learning instruments that are coherent. The learning instruments of one jurisdiction (QC) were deemed incoherent because no policies pertained to gathering evidence of social and emotional learning, an important feature of this instrument in the UNHCR document.

Community Sensitization to Refugeehood
Nine of the 11 jurisdictions with objectives of fostering social and emotional learning aim to do so, in part, through sensitization to refugeehood (AB, BC, MB, NB, NS, NT, ON, PE, SK). The objectives of two jurisdictions (NB, NS) were deemed incoherent with Refugee Education 2030 because no policies aim to sensitize students and school staff to contemporary refugee experiences. These policies do not fulfill the criteria because they do not aim to help students who have not sought or had RP provide psychosocial support to their peers with or seeking RP.
Three of the nine jurisdictions with policy objectives of sensitizing students to refugeehood through classroom instruction about refugee-related topics include SwRP and/or SsRP as a targeted benefiting population of such instruction (AB, MB, NT). The targeted benefiting populations of two of these jurisdictions (AB, MB) were assessed as coherent with the targeted benefiting population of Refugee Education 2030 because policies target SwRP and SsRP. 25 The targeted benefiting population of policies in the third jurisdiction (NT) was deemed incoherent because it is exclusionary (i.e., limited to those who fled armed conflict). SwRP and SsRP are absent from the targeted benefiting populations of the other six jurisdictions (BC, NB, NS, ON, PE, SK). The targeted implementation agents of policies towards community sensitization were assessed as coherent with Refugee Education 2030 in all nine jurisdictions with such policies because they include government agents. Eight of these jurisdictions also have targeted partners deemed coherent for including organizations and educational stakeholders. Policies pertaining to partnership in one jurisdiction (AB) were deemed incoherent because the sole policy pertaining to partnerships is borrowed from another jurisdiction (MB) and the policy is specific to educational stakeholders in that other jurisdiction.
Five of the nine jurisdictions with policies about community sensitization to refugeehood have capacity instruments that we assessed as coherent with Refugee Education 2030 because they build the capacity of government agents (AB, BC, MB, PE, SK). Such instruments are absent from the other four jurisdictions (NB, NS, NT, ON). While seven of the nine jurisdictions with policies addressing this subcategory of need have learning instruments that we considered coherent because those instruments recommend or require assessments of student learning about refugeehood (AB, MB, NB, NT, ON, PE, SK), there are no learning instruments in policies about community sensitization in the two remaining jurisdictions (BC, NS).

Mental Health and Psychosocial Support
Seven of Canada's 13 educational jurisdictions have refugee education policy documents with objectives of providing mental health and psychosocial support, and we assessed all of them to be coherent with Refugee Education 2030. At a more granular level, these seven jurisdictions have objectives of providing acculturation supports (AB, BC, MB, NL, ON, QC, SK), while five of them have objectives for providing school-and classroom-based trauma-related supports (AB, BC, MB, NL, ON) and five have aims of identifying and referring students requiring greater support to specialized mental health services (BC, MB, NL, ON, SK).
The targeted benefiting populations of policies pertaining to the provision of mental health and psychosocial supports of four of the seven jurisdictions with such refugee education policies were assessed as coherent with Refugee Education 2030 (AB, BC, MB, QC). Policies in these jurisdictions include SwRP and SsRP requiring mental health supports. The targeted benefiting populations of the remaining three jurisdictions we considered incoherent because SwRP are included but SsRP are not (NL, ON, SK). The targeted implementing and partnering populations in all seven jurisdictions with policies addressing mental health and psychosocial supports were assessed as coherent with Refugee Education 2030 because government agents are targeted as implementers and educational stakeholders and organizations as partners.
The capacity instruments in four of the seven jurisdictions with policies addressing this subcategory of need were assessed as coherent because they aim to support government agents in providing mental health and psychosocial supports (AB, BC, MB, SK). The capacity instruments in another two jurisdictions (NL, ON) were deemed incoherent because their use could result disparities in the targeted benefiting population, specifically among SwRP and SsRP who experience trauma-and stressor-related mental health concerns. 26 The learning instruments with respect to mental health in six of the jurisdictions were determined to be incoherent with UNHCR policy because they do not recommend or require formal evaluations of progress towards the desired objectives. Learning instruments in this category of need were absent from the seventh jurisdiction (BC).

Special Education
Refugee Education 2030 states that SwRP and SsRP should be "taught by teachers who have been adequately prepared to include refugee children and learners with diverse learning 26 Capacity instruments for identifying students with mental health needs in NL and ON are based on the fourth revision of the American Psychiatric Association's (APA) Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) and not the current version of the DSM, the fifth revision (DSM-V; APA, 2013). Significant changes were made to the diagnostic criteria from the DSM-IV to the DSM-V that are relevant to the identification of, and provision of services for, mental disorders of high prevalence in children and youth with refugee experiences "the relocation of PTSD from the category of anxiety disorders to a new diagnostic category called 'trauma-and stressor-related disorders'", "the explication and tightening of the definitions of trauma and exposure to it", and the addition of criteria for diagnosing trauma-and stressor-related disorders in children six years of age or younger that reflect their levels of development (Pai et al., 2017, p. 1). requirements, including children and youth with disabilities" (UNHCR, 2019b, p. 43). 27 In Canada in the last two decades, there has been a movement away from labelling certain students as "disabled" or "with disabilities" and then placing students thus labelled into separate classrooms or schools. Instead, the language of diverse needs and differentiated responses to those needs in an integrated classroom setting is becoming more commonplace. This trend was reflected in nine of Canada's 13 educational jurisdictions having policies with objectives of including SwRP and SsRP with diverse learning requirements and/or differentiating education in response to their diverse learning requirements (BC, MB, NL, NS, NT, ON, PE, QC, SK). 28 Such policies were absent in four jurisdictions (AB, NB, NU, YK).
The targeted benefiting populations of three jurisdictions with such policies were assessed as coherent with Refugee Education 2030 because they included SwRP and SsRP with diverse learning requirements (BC, MB, QC). The policies of the other six jurisdictions with policies were deemed incoherent because they exclusively targeted SwRP, but not SsRP (NL, NS, NT, ON, PE, SK). While all nine of these jurisdictions have at least one policy targeting SwRP and/or SsRP with diverse learning requirements, most of these policies have a targeted benefiting population that includes SwRP and/or SsRP as one subgroup and students with diverse learning requirements as another subgroup, and do not make the potential intersectionality of these two subgroups explicit. Only two jurisdictions (ON, MB) have policies that target SwRP and/or SsRP with disabilities; however, the objectives of these policies pertain to the provision of mental health and psychosocial supports rather than special education. There are therefore no policies in Canada that pertain explicitly to the provision of special education services to SwRP and SsRP with disabilities.
The targeted implementation agents of the special education policies in all nine jurisdictions with such policies were assessed as coherent with Refugee Education 2030 because they include government agents. Only one jurisdiction (ON) has policies that target partnerships for the express purpose of responding to SwRP and SsRP with diverse learning requirements, including disabilities; yet we deemed these policies about partnership incoherent with Refugee Education 2030 because they were limited to responding to mental health needs during the Syrian refugee resettlement initiative of 2015 and 2016.
All nine of the jurisdictions with policies in this category of need had capacity and learning instruments deemed coherent with Refugee Education 2030 because the former promote strategies and resources for differentiating teaching, learning, and assessment, and the latter encourage evidencebased responses.

Balancing the Independence and Intersectionality of Objectives
As mentioned at the outset of this findings section, two Canadian jurisdictions have no specific policies for refugee education. The 11 others that do have policies all include considerations relating to the co-occurrence or intersectionality of two or more needs (AB, BC, MB, NB, NL, NS, NT, ON, PE, QC, SK). Nonetheless, there are gaps in the targeted benefiting populations of special education policies and language education policies due to the overlap of these categories. First, Canada's refugee education policies about special education are mostly from policy documents with overall objectives of responding to language education needs. This means that these policies target only those students with both special education and language learning needs. SwRP and SsRP who have special education needs but who do not have language learning needs (i.e., who have the required proficiency level in the language(s) of instruction) are excluded from the targeted benefiting populations of special education policies. Second, policies about responding to missed education in SwRP and SsRP target only those who have both accelerated education and language learning needs. A fissure exists in the targeted benefiting population of accelerated education supports because SwRP and SsRP whose education has been limited or interrupted but who are proficient in the language(s) of instruction are excluded. These fissures in the targeted benefiting populations make such policies incoherent with Refugee Education 2030.

Discussion
We discuss the findings in two ways. First, we use three indicators in assessing policy coherence relative to each category of need. In doing so, we also include select implications of the distribution of populations of children and youth with or seeking RP across Canada's 13 provinces and territories (see this population distribution in Appendix 1). This helps to show where potential priorities for policy review or development exist. Second, we illuminate how and why the combination of two theorical frameworks allowed for the identification of possible sources of policy (in)coherence in refugee education.

Three Policy Coherence Indicators in Relation to the Five Categories of Need
ased on Cejudo and Michel's (2017) three elements (coherence among policies' objectives, targeted populations, and instruments), we scored the policy sets of each Canadian jurisdiction with respect to five categories of need (access to education, accelerated education, language education, mental health and psychosocial support, special education) using three policy indicators (coherent, incoherent, absent). This allowed us to assess the extent to which each set of policies was coherent with respect to Refugee Education 2030 and each category of need.

Access to Education
Of the education policies addressing the five categories of needs of SwRP and SsRP across Canada, those addressing access to education are among the most coherent with Refugee Education 2030. This suggests that when the policies are implemented, their design could complement each other in addressing the complex issues of social adaptation of SsRP and SwRP into the schooling systems (Cejudo & Michel, 2017). Six Canadian educational jurisdictions have refugee education access policies that are coherent across all elements (i.e., objectives, instruments, targeted populations), and these six are home to approximately 60% of SwRP and 56% of SsRP. 29 In the seven other jurisdictions, when policy gaps were observed, they were not due to the explicit exclusion of SwRP or SsRP from education access. 30 Instead, those gaps were due to overly specific targeted benefiting populations of SwRP and SsRP (e.g., references only to refugees from the Syrian Arab Republic) and absences of capacity and learning instruments to support the policy objectives. Although general education policies, including laws, requiring that all children in Canadian jurisdictions attend primary and secondary school would extend to SwRP and SsRP, Refugee Education 2030 calls for explicit policy on the inclusion of SwRP and SsRP (UNHCR, 2019) and previous studies have revealed the importance of having policies featuring specific recognition and needed supports for SwRP and SsRP (Shakya et al., 2010;Wilkinson, 2002). Here, our analysis suggests the 29 Population estimates throughout this section are based on resettlement and refugee claim patterns between January 2015 and March 2020 (see Appendix 1). 30 Two jurisdictions did place temporal limits on free education access for SsRP based on time elapsed since the submission of their refugee claim (AB, SK). development of policies explicitly guaranteeing access to education for SwRP and SsRP is needed in seven jurisdictions, especially in QC which was the destination of 18% of SwRP and 44% of SsRP arriving in Canada between 2015-2020.

Accelerated Education
Policies regarding missed education are incoherent with Refugee Education 2030. This means that every policy, by pursuing the objective, provided some assistance to refugee education in relation to accelerated education, but also created inconsistencies and gaps in the targeted population and instruments in supporting students' needs that missed schooling (Cejudo & Michel, 2017). Only one (MB) of the jurisdictions has policy for accelerated education that is coherent with UNHCR direction, but only 7% of SwRP and 1% of SsRP reside in that jurisdiction. Each of the eight other jurisdictions with policies addressing missed education had incoherencies and gaps across the targeted populations and/or provision of instruments, and these eight jurisdictions were home to an estimated 90% of SwRP and 98% of SsRP reside. Policy incoherencies in this area of need suggest policies do appear to provide supports to all SwRP and SsRP with missed education and/or to provide instruments that sought to build the capacity of educational professionals to address needs resulting from missed education, or to promote the testing and scaling up of evidence-based innovative approaches and practices. Of particular concern are the potential implications of the incoherencies of policies governing AEPs in Ontario, which suggest such programs may not be provided with a policy context conducive to meeting the accelerated learning needs of SwRP and SsRP in that jurisdiction. This situation could pose barriers to SwRP and SsRP with missed schooling from learning the curriculum content and developing the skills that will enable them to succeed in age-appropriate courses, and may pose barriers to those who are "aging out" of the public education system to pursuing higher education and desired futures. Because Ontario is home to an estimated 41% and 46% respectively of Canada's SwRP and SsRP, our analysis suggests a review of policies regarding AEPs is warranted in that province.

Language Education
Canadian refugee education policies about language education are largely coherent with Refugee Education 2030 in the eight jurisdictions that address this category of need in their policy sets. This means the policies pursued specific objectives and provided instruments but featured inconsistencies and gaps in targeted population (Cejudo & Michel, 2017). Where incoherence exists in the existing policy sets, it is a product of problems in specifying targeted benefiting populations, such as omissions and exclusions of certain cohorts of SsRP and SwRP. For instance, we observed policies mentioning "refugee camp" rather than explicitly addressing SsRP and SsRP. This lack of explicit specification could prompt school actors to overlook the experiences and needs of many SsRP and SwRP whose journeys did not include time spent in camps. It deserves noting that 60% of persons with refugee status live in out-of-camp accommodation, a statistic that has been stable since 2014 (UNHCR, 2019d). Our analysis suggests that some existing Canadian refugee education policies addressing language learning could use revision to explicitly target all SsRP and SwRP. Because approximately 99% of SsRP and 94% of SwRP live in these jurisdictions, the absence of policies in five other jurisdictions does not currently represent a critical policy gap unless resettlement patterns shift in the future.

Mental Health and Psychosocial Support
Supporting mental health and psychosocial needs of SwRP and SsRP is one of the most challenging categories related to the education of children and youth with refugee backgrounds. The ways in which all subcategories (e.g. social-emotional learning, awareness of refugee experiences, mental and psychosocial support) are incorporated in policy designs is crucial: They provide the necessary conditions to ensure the social adaptation of SwRP and SsRP, while ideally complementing each other in addressing needs. Mental health and psychosocial support was addressed in many Canadian jurisdictions. Eleven had policies towards social and emotional learning, nine addressing sensitization to refugeehood, and seven attending to mental health support). 31 However, no jurisdiction had a policy set that was coherent with Refugee Education 2030 across all sub-categories of need and all elements of policy coherence. Thus, policy sets in this category of need were deemed incoherent. One example of incoherence included two jurisdictions (NB, NS) where no policies in their sets aimed to sensitize students and school staff to contemporary refugee experiences. Another example is the omission of certain cohorts of SwRP and/or SsRP as target populations. This is seen, for instance, in 11 jurisdictions where SwRP and SsRP were not explicitly identified as agents in teaching and learning experiences about the refugeehood-related issues that affect many of their lives. Our analysis suggests that many existing Canadian refugee education policy sets addressing mental health and psychosocial support could benefit from greater consistency across subcategories of need and in explicitly including SwRP and SsRP as target populations.

Special Education
Canadian refugee education policies addressing special education are incoherent with Refugee Education 2030. This is because not all elements of policy coherence are reflected in their designs to complement each other in providing guidance, through their objectives, to targeted populations using instruments that adequately prepare educators to include SwRP and SsRP with diverse learning requirements, "including [those] with disabilities" (UNHCR, 2019b, p. 43). 32 Two areas of incoherence stood out. First, we found no policies in Canada that pertain expressly to the provision of special education supports to SwRP and SsRP who have disabilities. 33 Because the incidence of disability is higher in displaced populations than the global prevalence of 15% (UNHCR, 2019d), this policy gap could have adverse implications on the learning, wellbeing, and success of more than 15% of SsRP and SwRP in Canada. Second, of the nine jurisdictions with refugee education policies addressing diverse learning requirements and/or differentiating instruction in response to diverse learning needs, six omitted SsRP from the targeted benefiting population. These six jurisdictions are where approximately 46% of Canada's SsRP reside. Our analysis suggests that many existing 31 All 11 jurisdictions with objectives addressing socio-emotional learning also have capacity instruments that we considered coherent with Refugee Education 2030 because they target the capacity of educational professionals to support students' social and emotional learning. Nine of those 11 jurisdictions included objectives that seek to foster social and emotional learning, in part, through sensitization to refugees (AB, BC, M , N , NS, NT, ON, PE, SK). Seven of Canada's educational jurisdictions have refugee education policy documents with objectives of providing mental health and psychosocial support, which we assessed to be coherent with Refugee Education 2030. These seven have objectives of providing acculturation supports (AB, BC, MB, NL, ON, QC, SK), five of them have objectives for providing school-and classroom-based traumarelated supports (AB, BC, MB, NL, ON) and five have aims for identifying and referring students requiring greater support to specialized mental health services (BC, MB, NL, ON, SK). 32 This statement is from Enabling Activity 1 for realizing Expected Result 1 of Strategic Objective 2 in Refugee Education 2030. 33 n Canada, there has been a movement away from labeling certain students as "disabled" or "with disabilities" and then placing students thus labelled into separate classrooms or schools. Instead, the focus has shifted to addressing diverse needs using differentiated instructional responses in an integrated classroom setting. This shift in focus and terminology likely influenced our findings because Refugee Education 2030 emphasizes "disability" as a specific area of need within the broader notion of diverse learning needs.
Canadian refugee education policies addressing special education need to explicitly address SwRP and SsRP with diverse learning needs, especially those with disabilities.

Education Policy Coherence and Policy Attributes
The findings reveal how the combination of policy coherence among different education policies' objectives, targeted populations, instruments and attributes of specificity and consistency, as articulated in the theoretical framework, allowed for the identification of possible sources of policy incoherence in refugee education. For example, the gap in most jurisdictions of education policies specifically targeting SsRP as a targeted benefiting population wasa key element of policy incoherence across all five categories of need. This could have negative consequences for SsRP because policies do not draw the attention of school actors to this unique and important population of learners or guide their actions to support the complex issues of social adaptation of SsRP. The gap of SsRP as a benefiting population in most Canadian refugee education policies is problematic because children seeking RP represent 51% of the total number of children seeking and granted RP in Canada between 2015-2020. In the two provinces in which nearly 75% of SWSRS reside (ON, QC), there are more SsRP than SwRP; however, few of their refugee education policies have one or more policies targeting SsRS.
When including specificity as an attribute in our comparison and assessment of the sets of policies in refugee education in relation to the targeted benefiting populations, we observed that being vague (as in the case of SsRP) or too specific are both problematic in terms of potentially benefiting all groups of students. With respect to the latter, we noted that some policies were very specific in definitions or prescriptions, resulting in omissions or exclusions. For example, some policies targeted only SwRP or SsRP from specific countries of origin (e.g., the Syrian Arab Republic), granted refugee protection on specific grounds (e.g., armed conflict), with specific experiences in their transition countries (e.g., living in a refugee camp), or entering Canada through specific pathways (e.g., resettlement through government sponsorship). In other cases, some policies used "euphemistic" labels such as "newcomers". Although this terminology may be an attempt to avoid pejorative connotations that are sometimes attributed to the label "refugee", its lack of specificity could be a potential source of ambiguity in policy application.
Similarly, by including consistency as an attribute in our comparison and assessment of the sets of policies in refugee education, we also observed that policy designs could be too consistent about a specific targeted benefiting population. Some policy sets targeted highly specific subgroups so consistently that there was insufficient variability for the aggregated targeted benefiting population to encompass all SwRP and SsRP. Conversely, we also observed a lack of consistency in the terms used to identify populations of students and deemed this to be a source of incoherence that could affect implementation.

Conclusion
Unprecedented influxes of persons with and seeking RP are occurring through established and non-traditional resettlement pathways globally. This calls upon policymakers in resettlement countries to enact policies in education to support, coordinate and ensure the full integration of SwRP and SsRP. It is crucial that the set of educational policies in a given jurisdiction respond coherently to the needs of children and youth with or seeking RP. Incoherence or gaps can mean the specific needs of these populations are not adequately identified and addressed, and that implementing agents do not have specific and consistent guidance and instruments. Unintended consequences can thus ensue for SwRP nd SsRP. This study thus paid close attention to the quality of policy design and its various elements (objectives, target populations, instruments) to assess the extent to which the set of policies in each province and territory in Canada contribute to a "coherent" vision for refugee education (Cohen, 1995;Clune, 1993). Our use of Refugee Education 2030 as a reference point allowed for an assessment of all sets of educational policies that coexist in Canada and a comparative examination of how they help coordinate, reinforce, and improve the understanding and actions of school actors (and other implementing agents) in addressing the social adaptation and needs of SwRP and SsRP (Cejudo & Michel, 2017). Consequently, we highlight three implications of this study for refugee education policy researchers and policymakers in and beyond Canada.
First, following UNHCR (2019b), inclusion in national education systems is the recommended policy option for refugee and refugee claimant children and youth, and Refugee Education 2030 aims to ensure that "the particular learning needs of refugee students…are addressed" in such systems so they achieve parity with their non-refugee peers by 2030 (UNHCR, 2019b, p. 2). To be coherent with this strategy, policies in Canada's major educational jurisdictions should respond to the needs of children and youth with and seeking RP. Our study shows there are sets of education policies in effect across Canada that target responses to the five major categories of need. Numerous of these are exemplary in their coherence with Refugee Education 2030, but some are characterized by significant incoherencies and gaps. Canadian policymakers could use the findings from this study to develop or revise policies to address these shortcomings, and we made some suggestions in the regard in the discussion. Researchers and policymakers who find value in the approach offered here could replicate the study's method in their own targeted jurisdictions to identify incoherencies and gaps.
Second, this study points to the importance of policy coherence using Refugee Education 2030 as the key reference against which to assess and consider the importance of specificity in refugee education policies. The findings showed the importance of the specificity of policy objectives and targeted benefiting populations. For example, we operationalized specificity through the explicit identification of children and youth with refugee backgrounds no matter what their legal protections or resettlement pathways, as indicated by the 1951 Convention relating to the Status of Refugees (UNHCR, 1951). By doing so, we were able to identify policies that were inconsistent not only with Refugee Education 2030 but also with international convention because they did not include children and youth seeking RP. Following Desimone (2005), policies that explicitly specify the targeted benefiting population send an important signal, in this case to school actors, by narrowing the room for individual interpretation and thereby increasing fidelity in policy implementation and respect for international convention.
Third, for constitutional, theoretical, and methodological reasons outlined earlier, our findings rely on provincial and territorial education policies. This brings two considerations. First, each of Canada's constitutionally warranted educational jurisdictions has its own particularities. The main national mechanism for coordinating policy responses among them is a federally sponsored council (www.cmec.ca) that convenes ministers of education and pursues initiatives of mutual concern to federal, provincial and territorial governments (Wallner, 2017). This intergovernmental approach relies on the "soft" power of cooperation and influence. Historically, it has been relatively effective in building a pan-Canadian convergence in terms of educational standards (Wallner, 2018). But no extant research exists about its influence on policy coherence. Future research could examine efforts at creating policy coherence from within this intergovernmental perspective to provide further insights about the state and dynamics of education policy in Canada. Second, most of the jurisdictions in the present study are comprised of local school districts. It is likely that districts in large urban settings (e.g., Toronto, Montreal) that serve concentrated and diverse student populations have developed policy programs and instruments that would resonate with Refugee Education 2030, and from which researchers and policymakers could extract important lessons. A bottom-up perspective in future studies and policymaking should thus be considered as a complement to a focus on vertical coherence. Note. Values representing between 1 and 4 individuals are shown as "--" to prevent individuals from being identified. I/D = insufficient data for calculation. • Not a policy document (e.g., news release, report, student assessment tool, event poster, immigration welcome guide) Status

Appendix 2: Search Strategy for Jurisdiction-based Scoping Review of Policies
For policy guides: • Current For other policy documents: • Most recent or up-to-date version containing the term "refugee" and or "réfugié" For policy guides: • Expired For other policy documents: • Not the most recent or up-to-date version containing the term "refugee" and or "réfugié" Note. a i.e., The use of " mmigration, Refugees and Citizenship Canada" [ RCC] or " mmigration and Refugee Protection Act" [ RPA] for purposes other than defining targeted populations and/or topics of instruction. The use of the literary expression "to take seek refuge" to describe psychological dissociation or seeking shelter in a secure area during a school lockdown.

Policy objectives
The objectives of the refugee education policy are consistent with an Enabling Activity: They aim to respond to the category of need derived from an Enabling Activity of Expected Result 1 of Strategic Objective 2 of Refugee Education 2030 and address every detail provided in the Enabling Activity The objectives of the refugee education policy aim to respond to the category of need derived from an Enabling Activity of Expected Result 1 of Strategic Objective 2 of Refugee Education 2030; however, they do not address every detail provided in the Enabling Activity and/or one or more of the objectives of the policies comprising the refugee education policy contradicts the goal of the Enabling Activity and/or one or more of the objectives of the policies comprising the refugee education policy obstructs coherence at another element of policy coherence.
No policy objectives of responding to the category of need derived from the Enabling Activity of Expected Result 1 of Strategic Objective 2 of Refugee Education 2030 were identified in the refugee education policy set, should the jurisdiction have such a policy set. This does not necessarily mean that there are no such policy objectives in the jurisdiction's entire education policy set.

Policy instruments
The policies include both capacity instruments that are fully coherent with the Strategic Approach of Collaborative Learning and Capacity Development and learning instruments that are fully coherent with the Strategic Approach of Innovation, Evidence, and Growth.
The policies do not include capacity instruments and/or learning instruments; or, including both capacity and learning instruments, the learning instruments do not encourage or require evaluation of progress towards the objective(s) and/or the capacity instruments provide erroneous information that could decrease implementation capacity and/or the instruments obstruct coherence at another element of policy coherence.
The policies included neither capacity instruments nor learning instruments. Policies comprising refugee education do not explicitly or implicitly include any part of the targeted benefiting population of Refugee Education 2030.

Coherence among implementing populations
Policies comprising refugee education are to be implemented by all of the government agents relevant to the response in the given context.
Policies comprising refugee education are to be implemented by some but not all of the government agents relevant to the response in the given context and/or policies comprising refugee education contradict descriptors of the targeted implementing population in Refugee Education 2030 and/or policies comprising refugee education are to be implemented by agents that would obstruct coherence at another element of policy coherence.
Policies comprising refugee education do not have an explicit targeted implementing population (e.g., imperative sentences without a subject implied elsewhere in the policy).

Coherence among partnering populations
The policies target one or more partner(s) from both of the two categories of partners sourced from Refugee Education 2030: Organizational partners (e.g., intergovernmental organizations, international non-governmental organizations, domestic civil society organizations, private sector organizations and foundations, civil society organizations, and academic networks) and partners from a whole of society approach (e.g., educational stakeholders -students, parents, guardians, families, and communities) The policies in refugee education fail to target one or more partner from one or both of the two categories of partners sourced from Refugee Education 2030: Organizational partners and partners from a whole of society approach and/or policies in refugee education prohibit or prevent one or both of the two categories of partners from Refugee Education 2030organizational partners and educational stakeholdersfrom being partners and/or the targeted partnering populations of the refugee education policy obstruct coherence in another element of policy coherence.
There are no policies in refugee education pertaining to partnership.