Introduction

1Before 2013, Canadian government officials often cited Afghanistan as an example of successful whole-of-government engagement in fragile states. Official Ottawa lauded the Afghanistan taskforce system, with its extensive coordination of Canadian efforts from Ottawa down to the Provincial Reconstruction Team in Kandahar, as the gold standard for joined-up approaches (GOC 2012a). The peer review of Canada’s development cooperation, carried out by the Development Assistance Committee of the Organisation for Economic Co-operation and Development (OECD/ DAC), endorsed that view by praising Canada’s “effective whole-of-government approach to disaster response and fragile states, particularly in Afghanistan and Haiti” (OECD 2012, 9). It recommended that Canada’s “assistance to other partner countries would be more effective if it applied the relevant program considerations emerging from Afghanistan in those contexts” (OECD 2012, 12). Such considerations included ensuring high-level commitment from key departments, following the integrated approach down to the country level, injecting sufficient human and financial resources, and streamlining programming processes. Curiously, the DAC assessment focused almost entirely on “front-end” factors and did not refer to the considerable evidence of poor downstream results.

2Non-governmental organizations (NGOs) and scholars have questioned that rosy reading of Canada’s record in Afghanistan and other fragile and conflict-affected states (FCAS). Early on, the Canadian Council for International Co-operation suggested that joined-up approaches subordinated aid to security objectives and undermined poverty reduction in the poorest countries of the South (CCIC 2006). Banerjee (2008, 2010, 2013) grounded that critique in experience-based analyses of Canadian engagement in Afghanistan. Brown (2008) and Swiss (2012) suggested that the “securitization” of aid affected a wider swath of Canadian development policies. Together with Carment and Samy (2012), they also questioned the degree to which Canadian official development assistance (ODA) had been effective in high-profile FCAS – in terms of Canadian value-for-money or in terms of global aid effectiveness criteria rooted in the Paris Declaration.

3However, based on a systematic comparison of Canadian involvement in Afghanistan, Haiti, and Sudan, Baranyi and Paducel (2012) argued that the effectiveness and securitization of aid varied significantly – from high securitization and very low effectiveness in Afghanistan, to lower securitization, greater conflict-sensitivity, and slightly higher effectiveness in the other two high-priority FCAS. This chapter extends that line of analysis to other contexts that the Canadian International Development Agency (CIDA) did not categorize as fragile. Historically, CIDA distinguished between programming in FCAS such as Afghanistan, where “programming objectives align closely with Canadian foreign policy objectives associated with regional and global security and stability,” and in low-income countries (LICs), which “face pervasive poverty and limited institutional capacity, but have broadly stable governance and public security” (CIDA 2013, 17, 19).

4That distinction may have been justified in the immediate post– 9/11 era, when some CIDA officials sought to protect the agency’s space from further incursion by central agencies or other government departments. Is the distinction still useful today, given the creation of an integrated Department of Foreign Affairs, Trade and Development (DFATD)? What do the sudden shifts from stability to fragility in LICs such as Mali, and the relative success of development cooperation in LICs such as Bangladesh, suggest about the possibilities for more two-way learning across the fragility-stability spectrum? Finally, to what extent has Canadian ODA been linked to commercial objectives in those countries, as suggested by some critics?

5This chapter explores those issues by drawing lessons from five cases where Canadian involvement has been extensive but where state fragility has varied significantly. We call them “other” FCAS because, aside from the West Bank and Gaza, they have not historically been considered as such by the Canadian government, given that they have not recently suffered from the intensity of violent conflict or other forms of fragility found in Afghanistan, Haiti, and Sudan. However, as suggested by Table 1 and the fragility indices on which it is based, each of our five cases is affected in distinct ways by governance crises, weak state capacity, armed conflict, foreign military intervention, or vulnerability to natural disasters.

6We selected these cases based on two criteria: inclusion in the top twenty Canadian ODA recipients and the top fifty FCAS on standard fragility indices. We excluded Afghanistan, Haiti, and Sudan because they have been analyzed elsewhere (Baranyi and Paducel 2012). We included the West Bank and Gaza (also known as the Occupied Palestinian Territories or Palestine) because that case has not been adequately analyzed from a Canadian development and security angle, even though CIDA classified it as a fragile state. There are no cases in the Americas, since Haiti is the only country in the region that ranks in the top twenty partners for Canadian cooperation by volume of disbursements. Other African FCAS, such as the Democratic Republic of Congo and Somalia, also fit our selection criteria, but we excluded them to keep the sample size manageable for a focused comparison.

7In searching for answers to the questions noted above, we compared three aspects of Canadian whole-of-government engagement in the five contexts: (1) the extent to which Canada’s ODA has been effective in developmental terms; (2) the degree to which Canadian aid has been securitized, as opposed to conflict-sensitive; (3) and the extent to which aid been commercialized. Our understanding of development effectiveness is rooted in the Paris Declaration and subsequent international norms, which assume that ODA will be more effective if it is nationally “owned,” aligned with national priorities and systems, harmonized among donors, managed for development results, and based on the mutual accountability of partners. Despite their limitations, those standards are widely accepted, at least among Western donors, and have generated a significant body of comparable, longitudinal OECD/DAC data for most of our cases.

8To analyze the connections between security and development programming, we used the standards codified in the OECD/DAC Principles for Good International Engagement in Fragile States and Situations (OECD 2007). That text called for the recognition of links between political, security, and development goals, yet it urged OECD governments to emphasize the prevention of violent conflicts (i.e., “conflict-sensitivity”) through the coherent use of all policy instruments. It stressed the principles of taking context-specificities as the starting point for international engagement and the imperative of doing no harm – for example, when deciding in which sectors to use ODA, with which partners and to what ends. In the complementary OECD/DAC Guidelines for Security System Reform and Governance (OECD 2004), Western donors approved the use of ODA to support the development of state security institutions such as the police, but within clear parameters that excluded the use of ODA to support the repression of legitimate civilian opposition or directly strengthening the counterinsurgency capacities of Southern governments. The use of ODA to support such security-driven strategies is what we call the “securitization” of aid (Baranyi and Paducel 2012). Those standards enable us to situate Canadian ODA on a spectrum ranging from securitization to conflict-sensitive development.

9With regard to the commercialization of development cooperation, international norms are less clear. As such, we take the extent to which ODA is tied to Canada’s commercial goals and to specific trade or foreign investment promotion activities as a starting point to assess degrees of commercialization. This exploratory analysis is intended as a baseline for future studies on the consequences of Ottawa’s aim of forging a rapprochement between Canada’s aid efforts and its trade plus investment efforts.

10This chapter makes four main arguments. First, we argue that the development effectiveness of Canadian cooperation varies across countries, from being quite low in the West Bank and Gaza, as well as in Pakistan, to somewhat better in Ethiopia and Mali, to fairly high in Bangladesh. Second, we suggest that the degree to which development cooperation is securitized, rather than conflict-sensitive, also varies: from high securitization in the West Bank and Gaza, to less securitization in Ethiopia, Mali, and Pakistan, to little securitization and considerable conflict-sensitivity in Bangladesh. Third, we argue that the commercialization of ODA is not a major concern in any of these cases. In Bangladesh, we suggest that there is complementarity between ODA and the trade relations. Finally, we build on those observations to conclude that it is time to rethink CIDA’s traditional distinction between fragile states and low-income countries, and to foster more two-way learning across that spectrum, instead of assuming that the Afghanistan model is the appropriate template for all whole-of-government interventions in difficult contexts.

11Our research is based mostly on publicly accessible primary documents and secondary literature. Given space limitations, it is not possible to provide deep background, though key characteristics of each case are noted. We do not analyze important aspects such as the institutional processes that undermined Canadian programming in FCAS because they are addressed by Carment and Samy (this volume).

Canadian Development Cooperation in FCAS

Size, Modalities, and Priorities

12CIDA programming in this sample has reflected major trends in Canadian cooperation since 2001, notably its budget increases until 2011 and its concentration in a limited number of “countries of focus.”1 Beyond those general trends, ODA levels and modalities were influenced by events in each context, as suggested by the data in Table 2.

Table 2. Canadian ODA to Sample Cases, 2007–2013 (in C$ millions)

13Ethiopia has consistently been the largest recipient of ODA in our sample – and was the largest CIDA country program in the world in 2011–12. However, in 2006, Canada cut bilateral aid to Ethiopia, following the government’s alleged fraud and violence during the 2005 national elections, redirecting ODA through multi-donor trust funds and NGOs (OECD 2012, 110). Mali received less but still considerable ODA during the 2001–11 decade, most of it channelled through general budget support. After the military coup in April 2012, Ottawa also suspended government-to-government cooperation to Mali.

14Pakistan and Bangladesh have also received significant levels of development assistance. In Pakistan, Canada channelled most of its ODA through multilateral and civil society organizations. In Bangladesh, those modalities have also been important, though considerable cooperation has also been channelled through government programs. The West Bank and Gaza received the lowest level of ODA in our sample, mostly as humanitarian aid channelled through multilateral agencies and Canadian NGOs.

15According to CIDA country strategy documents, between 2007 and 2009, priorities in these countries ranged from private sector development and democratic governance to basic education, natural resource management, and gender equality (CIDA 2009abc). Country priorities became somewhat more focused in 2009, in line with CIDA’s revised aid effectiveness strategy.

Effectiveness

16The country reports of the 2011 Paris Declaration (PD) Survey suggest that the effectiveness of Canadian ODA has varied across the five cases, at least on certain indicators (OECD 2011abcd).2 It has been most aligned with government priorities and systems in Bangladesh. In Ethiopia and Mali, Canadian cooperation has aligned with government priorities, but it largely disengaged from government systems after the 2005 governance crisis in Ethiopia and the 2012 coup in Mali. Canada–Pakistan cooperation is the least aligned with government systems.

17Although CIDA has reported on development results in all cases since 2007, it is difficult to match those reports with host country results data. This seems partly attributable to weak development evaluation systems in those countries. It also reflects the problematic performance of CIDA and its host country partners with regard to the PD principle of mutual accountability. That pattern seems to apply across the board, even in relatively enabling contexts such as Bangladesh.

18CIDA also suggests that its aid has been fairly effective in the West Bank and Gaza, where Canadian assistance priorities have been aligned with priorities codified in key documents, including the 2008–2011 Palestinian Reform and Development Plan (GOC 2013). Canada can also credibly claim that it coordinates with Western donors and with the Palestinian Authority in the West Bank via key sector working groups, such as the security and justice working groups. However, Ottawa’s claims about development results should be treated with caution, given the compelling questions that have been posed about the effectiveness of aid in the context of protracted occupation (Ibrahim and Beaudet 2012; Abu-Zahra and Kay 2012).

19The evolution of CIDA’s engagement in Ethiopia illustrates the challenges of remaining engaged in LICs when they drift towards authoritarian governance. The last evaluation of CIDA’s Ethiopia program suggested that after the 2005 electoral crisis, CIDA managed to balance support for the poverty reduction agenda of the government with a consistent message on human rights and democratic governance (CIDA 2010, x). However, the report acknowledged that CIDA missed an historic opportunity to significantly scale up its support to Ethiopian civil society and thereby foster the social base for longer term democratization (CIDA 2010, 16).

20The Ethiopian experience is not unique. The difficulties of adapting cooperation with LICs when apparently good governance degenerates into major crises raise profound questions about simple distinctions between LICs and FCAS. As we will demonstrate in the next section, the cases of Ethiopia and Mali also underscore the importance of building conflict sensitivity into programming in such contexts, despite their image of stability.

From Securitization to Conflict Sensitivity

21As explained in the introduction, the distinction between securitization and conflict sensitivity enables us to situate our cases on a spectrum ranging from the Palestinian Territories, where Canadian ODA has become quite securitized, through Pakistan, Mali, and Ethiopia, where traces of securitization are present, to cases such as Bangladesh, where development-security linkages appear more conducive to conflict prevention.

22Canada’s engagement in the West Bank and Gaza has been transformed by the Harper government. From 1948 to 2006, Ottawa maintained an even-handed approach to the Israeli-Palestinian conflict, supporting peacekeeping through the United Nations and maintaining a small ODA program focused on relief, social development, and dialogue on Palestinian refugees’ rights (Bell et al. 2007; Funk 2007). After the violent fragmentation of the Palestinian territories in 2006, Ottawa doubled its assistance and developed new projects aiming to “establish a safe environment in the West Bank conducive to Israel’s security priorities” (DFATD 2012, 1). Those projects included counterterrorism capacity building for the Palestinian Authority, and support for Joint Operations Centres with the Israeli Defence Force, linked to the deployment of Canadian security personnel embedded in the US and European Union security assistance missions. They also included using ODA to support justice sector projects such as strengthening the Palestinian Authority Prosecution Services and building new courthouses. Ottawa has justified those projects in terms that echo OECD/DAC norms on security-development linkages. Yet critics seem correct in viewing Canadian and broader Western assistance since 2006 as driven largely by the aim of modernizing the Palestinian Authority in the West Bank, while keeping it subservient to Israeli security interests and isolating Hamas (which Ottawa lists as a terrorist organization) in the Gaza Strip (Brynen 2007; Ibrahim and Beaudet 2012).

23Canadian cooperation with Pakistan, Mali, and Ethiopia differs significantly from that approach. In Pakistan, post–9/11 concerns about Islamic terrorism also affected the rationale of Canadian programming, particularly in education and in the conflict-affected province of Balochistan, which borders on Afghanistan (Lee 2013). However, Ottawa did not establish a security assistance program and does not appear to have aligned other elements of the CIDA program on security objectives. CIDA programming does not appear to have been particularly conflict-sensitive either, notably in Balochistan, and instead focused on socio-economic development “to support Canadian interests in Afghanistan” (CIDA 2009b, 3).

24There is more historical evidence of conflict-sensitivity in Mali, where CIDA programming was framed by the goals of democratic development and peacebuilding in the early 1990s. That sensitivity dwindled after 9/11, with the abandonment of substantial CIDA support to decentralization in the North (McGregor 2013). With the rise of Al-Qaeda in the Islamic Maghreb and the kidnapping of two former Canadian diplomats, Canadian policy took on a harder edge. That shift was reflected in Canadian counter-terror assistance to the Malian Army in 2011–12 (Pugliese 2011, 2012). As mentioned above, Canada cut government-to-government assistance after the coup, but the retrenchment of ODA programming to southern Mali suggests that Ottawa is not yet addressing the factors underlying the renewed North–South conflict (McGregor 2013).

25In Ethiopia, Canada displayed more consistent conflict-sensitivity when it cut government-to-government cooperation after the ruling party and security forces committed grave violations of human rights during the 2005 elections (OECD 2012, 110). Despite Ethiopia’s importance for the West in the Horn of Africa, Canada did not establish a security assistance program and does not appear to have overtly aligned its ODA with Western regional security goals. Nonetheless, the spectacular growth of the Ethiopia program as the largest Canadian ODA partnership in 2011–12, despite the authoritarian record of the regime in Addis Ababa, suggests that geopolitics can influence development cooperation more subtly.

26In Bangladesh, there has been no drift towards the securitization of aid and more progress towards conflict-sensitive policy coherence. The Bangladeshi Army has long benefited from Canada’s Military Training Assistance Program. Most of that training seems to be linked to the country’s participation in UN peacekeeping operations. Indeed, Ottawa frames its relations with Bangladesh as part of a broader partnership that includes converging positions on regional and global security issues, notably on international peacekeeping (GOC 2014).

27The distance between Canada’s support for counter-terror operations in Palestine or even Mali, and its approach in Bangladesh, highlights both the risks of securitization in certain FCAS and the potential of conflict-sensitive policy coherence in others. It also validates the idea of situating Canada’s involvement in different FCAS along a spectrum of development-security relationships that runs from complete securitization at one end to consistent conflict-sensitivity at the other, a point to which we will return in the last section of this chapter.

Commercialization

28What is the relationship between Canadian trade and development cooperation in these cases? As suggested in Table 3, there is no simple correlation between volumes of Canadian trade and development cooperation with these countries. Canada trades little with Ethiopia, its largest ODA partner. It trades much more with Bangladesh and Pakistan, but very little with Mali, and apparently little with the Palestinian Territories.3

29Due to lack of data on Canadian foreign direct investment (FDI) for our sample, it is also difficult to establish clear links between Canadian FDI and ODA. However, anecdotal evidence suggests that Canadian investments are significant in some contexts. For instance, in Ethiopia, thirteen Canadian companies have signed contracts for the exploration of potash and precious and base metals, with a registered capital of C$6.5 million (GOC 2012b). Canadian mining firms are active in Mali (Cousineau and Mackrael 2013). Yet there does not appear to be a direct relationship between the content of Canadian development cooperation and business involvement in these countries. For example, in Ethiopia, despite Canadian firms’ growing interests in mining, ODA remains focused on food security, health, and strengthening democratic institutions.

30Canada’s growing trade relations with Bangladesh illustrate the indirect synergies that can be nurtured between trade and ODA. Indeed, Bangladesh is one of the countries that has benefited from the reduction of Canadian tariff barriers on garment manufactures and other imports from least developed countries in 2003 (Weston 2003). Ottawa’s decision to provide assistance to improve health and safety practices in the Bangladeshi garment industry, in the wake of an industrial catastrophe in Dhaka (DFATD 2013d), suggests that the linkages between trade and ODA might grow tighter. That could be positive, but it seems wise for scholars to monitor that scenario in Bangladesh, as well as in FCAS such as Ethiopia and Mali, to understand if Ottawa’s renewed drive to promote Canadian commerce might undermine conflict-sensitive and sustainable development.

Conclusions and Options

31Our study of Canadian development engagement in a range of fragile states other than the high-profile cases of Afghanistan, Haiti, and Sudan, enables us to extend Baranyi and Paducel’s (2012) analysis to other contexts. It suggests that the effectiveness and securitization of aid in lower-profile cases has varied significantly – from questionable effectiveness and high securitization in West Bank and Gaza, to lower securitization yet diminishing effectiveness in Ethiopia, Mali, and Pakistan, to more conflict-sensitivity and effectiveness in Bangladesh. The shifts from stability to fragility in LICs such as Ethiopia and Mali, and the relative success of development cooperation in LICs such as Bangladesh, suggest that it would be useful to draw (context-appropriate) lessons from the latter, rather than assuming that the “Afghanistan model” should be applied to all FCAS. The cases of Ethiopia and Mali certainly highlight the importance of building conflict sensitivity into programming in such contexts, despite their historic image of stability. The 2013 evaluation of the Mali country program pointed towards the need for “additional resources and tools to more thoroughly analyse risk scenarios, and make relevant plans and adaptation measures” (DFATD 2013c). It also noted that this approach would be useful “for the Mali Program as well as other CIDA Country Programs operating in countries in conflict” (DFATD 2013c).

32The differences between Canada’s counter-terror capacity building in Palestine and Mali, and Ottawa’s approach in Bangladesh, clearly underscore the risks of securitization in certain FCAS and the potential of conflict-sensitive policy coherence in others. Together, these cases suggest that it is useful to look beyond the one-way application of considerations lessons from whole-of-government involvement in Afghanistan, to more two-way learning across a wider range of cases on the fragility-stability spectrum.

33Some of these lessons appear to have been drawn in official Ottawa. In 2012, Canada scaled down its aid program in Afghanistan, in line with the reduction of Western forces on the ground (GOC 2012). In January 2013, it announced a freeze on new Canadian aid to Haiti, pending a review of the apparently slim development results on the ground (Fantino 2013). Around the same time, Minister of Foreign Affairs John Baird declared that Canada would not be dragged into “another Afghanistan” in Mali (Murray 2013). Development programming in Pakistan and Sudan, as well as the West Bank and Gaza, were also placed under review. As a result of those reviews, in July 2014 Pakistan was moved from being a country of focus to becoming a “development partner” with a more modest ODA program (DFATD 2014). So despite earlier claims about the success of Canadian whole-of-government programming in Afghanistan, the “Afghan model” seems unlikely to become the template for Canadian engagement in other FCAS.

34Nonetheless, our comparative analysis suggests that while Canada may continue to scale down assistance to FCAS where results fall dramatically short of Western expectations, it will continue to provide assistance to most countries. In cases with relatively stable governance such as Bangladesh, government-to-government cooperation and support to civil society may remain the preferred modalities, partly because they have demonstrated fairly satisfactory results. Despite its concerns about democratic governance and poor development results in Ethiopia, Mali, and Pakistan, Ottawa will stay involved because of their regional importance. In the West Bank and Gaza, where policies and their outcomes are more profoundly contested at the national and international levels, Ottawa will probably be more cautious and channel relatively low levels of funding through international and civil society organizations. Humanitarian assistance is likely to endure as a major component of cooperation programs there and in other FCAS, with predictably negative consequences for national ownership and development sustainability.

35The chapter also suggests that development-security linkages will vary considerably in different FCAS. The degree of securitization that emerged in Afghanistan is unlikely to be replicated in other contexts. Nonetheless, a significant security assistance portfolio linked to justice sector programming will probably endure in the West Bank and Gaza, due to its connections with Western and Israeli geopolitical interests. Ottawa may be tempted to take a similar approach in Mali, given the latter’s increasing regional importance, though Prime Minister Harper’s reaction to calls for Canadian military intervention in early 2013 suggests that caution may also prevail there. At the other end of the spectrum, there should be opportunities to cultivate more positive development-security linkages in Bangladesh, where a multi-dimensional development partnership fits well with conflict-sensitive capacity building for UN-mandated peacekeeping.

36Our preliminary exploration of links between Canada’s commercial and ODA relations suggests that there is only modest cause for concern about the commercialization of ODA in these contexts. In Bangladesh, commercial ties may keep expanding in ways that complement ODA-based cooperation. In countries such as Ethiopia, Mali, and Pakistan, where commercial activity is more risky due to state fragility or armed conflict, Ottawa will likely use ODA to nurture an environment for future investment and trade. Given the announcement in 2014 of new ODA programs in Benin, Burkina Faso, the Democratic Republic of the Congo, Mongolia, and Myanmar, where Canadian commercial interests could converge with problematic governance to generate high risks for many stakeholders, and given the government’s aim of fostering synergies between FDI and ODA, this dimension of our study clearly requires further examination.

37Finally, our analysis calls into question CIDA’s traditional distinction between FCAS and LICs. As we have seen in the cases of Ethiopia, Mali, and Pakistan, the distinction between FCAS and states with “broadly stable governance and public security” is sometimes based more on wishful thinking about stability rather than on a rigorous understanding of conditions in these societies. Viewing such partners as moving along a spectrum of fragility over time may enable Canada to adapt its cooperation more wisely in fluid contexts. It may also enable DFATD to draw on relevant lessons from less fragile states such as Bangladesh to inform policy in more fragile contexts, rather than trying to apply a problematic “Afghanistan model.”

38There is considerable convergence between the lessons one can draw from these contexts and the learning codified in the New Deal for Engagement in Fragile States, a document signed by OECD donors and nineteen self-designated FCAS in 2011 (IDPS 2013). The New Deal’s core principles – letting FCAS lead their own development processes; prioritizing legitimate politics, peoples’ security, and employment; and remaining engaged in partnerships over the long run – echo the norms that have governed cooperation in contexts such as Bangladesh. The New Deal and its follow-up processes embody other lessons learned in FCAS over the past decade, including the need to help partner countries conduct their own fragility assessments and the importance of using common assessments to orient external assistance. DFATD might be in a better position to practice such conflict-sensitive programming, as envisaged by Carment and Samy (this volume). The merger might even enable the department to provide coherent advice about the risks of unduly commercializing ODA in contexts of fragile governance, particularly in the extractive sector.

39While that advice may not always be heeded by the Harper government, it could be picked up in the future. The next government in Ottawa could certainly rethink Canada’s narrowly partisan approach in the West Bank and Gaza, by bringing Ottawa back into line with international norms, including the New Deal. Scholars could contribute to that scenario through nuanced analyses of how the development dimension of Canada’s whole-of-government engagement evolves in different FCAS. In that connection, it is essential to complement research on (decreasingly) high-profile cases such as Afghanistan and Haiti with more grounded research on Canadian involvement in “other” FCAS, such as those examined herein. It also seems important to extend that analysis in particular to the Democratic Republic of the Congo and Myanmar, since their “new deals” have attracted Canadian attention across the ODA-commerce spectrum – despite enduring governance and human rights concerns.