The Impact of Refugee Presence on Host Populations in Tanzania A Desk Review April 2018 The Impact of Refugee Presence on Host Populations in Tanzania A Desk Review April 2018 Helidah Ogude, The World Bank Copyright © World Bank, 2018 All rights reserved. This publication is a product of the staff of the World Bank Group and the United Nations. The findings, interpretations, and conclusions expressed in this report are entirely those of the author(s) and do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent, or those of the United Nations. The World Bank and the United Nations do not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank or the United Nations concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing therein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank or the United Nations, all of which are specifically reserved. Contents Acknowledgements ..................................................................................................................................................... vii Acronyms ..................................................................................................................................................................... viii Introduction .................................................................................................................................................................... 1 Background ..................................................................................................................................................................... 3 Scope ................................................................................................................................................................................. 6 Overview of the Impacts.............................................................................................................................................. 8 Pre-existing Livelihood Strategies in Host Communities...............................................................................8 Refugee Policymaking and its Impact on Host Communities ........................................................................8 Labor Market Outcomes: Jobs and Wages .........................................................................................................9 Labor Market Outcomes: A Gendered Perspective ....................................................................................... 10 Labor Market Outcomes: Age and Gender ...................................................................................................... 11 Labor Market Outcomes: Casual Labor ............................................................................................................ 12 The Local Economy and Food Prices................................................................................................................ 13 Food Security and Prices: Refugee and Humanitarian Impact ................................................................... 14 Local Infrastructure and Services ....................................................................................................................... 14 Infrastructure and services ............................................................................................................................. 14 Health outcomes ................................................................................................................................................ 15 Environmental Impacts......................................................................................................................................... 16 Security and Social Cohesion ............................................................................................................................... 17 Long-run Impacts on Welfare ............................................................................................................................. 18 Lessons Learned and Policy/ Practice Recommendations ............................................................................... 19 Table of Lessons Learned and Policy/Practice Recommendations................................................................ 20 Future Research........................................................................................................................................................... 24 Bibliography ................................................................................................................................................................. 26 v vi Acknowledgements This review was jointly prepared by the World Bank and UNHCR to enhance the evidence base on forced displacement given global commitments made at the New York Declaration on Refugees and Migrants in September 2016. The report was authored by Helidah Refiloe Ogude (Social Development Specialist, World Bank). Valuable guidance and comments were received from Joanna de Berry (Senior Social Development Specialist, World Bank) and Trine Lunde (Senior Development Economist, UNHCR). The review also benefited from being formally peer reviewed by Laura Bailey (Lead Social Development Specialist, World Bank), Theresa Beltramo (Senior Economist, UNHCR) and Beth Whitaker (Associate Professor, University of North Carolina at Charlotte, NC, USA). Funding for the review was provided by the UNHCR-World Bank ‘Just in Time’ trust fund for technical assistance. vii Acronyms CCM Chama Cha Mapinduzi Party CRRF Comprehensive Refugee Response Framework GoT Government of Tanzania IDP Internally Displaced Persons TCRS Tanganyika Christian Refugee Services UNHCR United Nations High Commissioner for Refugees viii Introduction development responses to situations of forced displacement. Therefore, an imperative exists to review what is known about previous experiences of refugee arrival and response and the impacts on host communities. Given the It is widely understood that host populations are depth and breadth of evidence that can be drawn affected by a sudden and large influx of refugees from the Tanzanian experience, it serves an (World Bank, 2016). Precisely how they are insightful case study from which policy lessons affected, however, remains under-researched and can be learned from and applied in a range of often ill-communicated. Several quantitative contemporary displacement contexts. studies have been carried out on the impact of forced displacement on host populations, mainly This desk review was conducted against this in Colombia, the Great Lakes and increasingly in backdrop of the new global commitment to the Middle East and Europe. However, until protecting refugees and better supporting the recently, this area of study has largely been countries and communities that host them. It is neglected by economists in particular (Ruiz et al, hoped that the evidence and analysis presented 2013 and Oxford Refugees Center, 2011). Only a here will inform policy responses for the various few studies rely on empirical data, and they are governments across the world faced with typically focused on short-term impacts significant refugee and Internally Displaced (Kreibaum, 2016 and Ruiz et al, 2013). Persons (IDP) populations, as well as the humanitarian and development actors involved Tanzania, however, is an exception in this in supporting them. Although critical analysis regard, partly because of its location (surrounded on its own cannot lead to sound and well- by countries periodically affected by conflict) and evidence policies, which rely on political will and its decades-long history in welcoming and available resources, it can however work to assisting large numbers of refugees. Unlike dispel myths that may otherwise be used to several other hosting countries, there exists a mobilize ill-formed practices and policies. considerable body of qualitative, mixed-methods Instead, this kind of analysis can redirect and empirical literature, mostly analyzing the attention toward people, places and processes impact of refugee inflows from Burundi (1993) that warrant attention and that may otherwise and Rwanda (1994) on host districts in be misunderstood or neglected (Landau & northwestern Tanzania. This literature covers a Achiume, 2017). range of impacts including on the labor market, environment, health and other areas. As such, this review will provide: • A brief history of refugee policy and The formulation of the forthcoming Global practice in Tanzania; Compact on Refugees and implementation of a Comprehensive Refugee Response Framework • An overview of the impacts/outcomes (CRRF)1 in countries such as Djibouti, Ethiopia, along different variables (e.g., jobs, Kenya, Rwanda, Somalia and Uganda all stress health, etc.); the inclusion of host communities in efforts to extend refugee protection and to bring 1 CRRF is a multi-stakeholder approach that aims at investments in the resilience of both refugees and linking humanitarian and development efforts during local communities. the early stages of an emergency while strengthening 1 • A list of lessons and policy/practice Since the primary focus of this review is an options that can be gleaned from an analysis of the impact of refugee presence on host analysis of the studies’ findings both in communities in Tanzania, the following terms of refugee impact and historical background section is not intended to humanitarian/development be comprehensive. Rather, it seeks to provide a impact/response; and summary overview of changes in Tanzania’s policy and practice toward refugees by • A brief taxonomy of areas for possible highlighting some of the most significant further research and understanding. documented socio-political and economic factors that led to these changes. Figure 1. Refugees and Asylum-seekers Hosted in Tanzania from 1973 to 2015 Source: UNHCR Online Population Statistics Database (accessed on March 1, 2018). Note: In 1972 approximately 160,000 Burundian refugees fled to Tanzania but UNHCR data only records 90,000. It was common for Burundians to cross the border into Tanzania. 2 Background Aside from the economic incentives for refugee accommodation at the time, external financial and technical resources are also considered to be key motivations for the establishment and sustainability of the settlements. The Tripartite Tanzania’s experience of accommodating Partnership Model—an agreement between the refugees dates back to the colonial era. Government of Tanzania, the UNHCR and the Thousands of Africans fled their countries and Tanganyika Christian Refugee Services predominantly settled in the “less populated and (TCRS)—institutionalized the TCRS assuming more fertile western borderlands” of Tanganyika responsibility for managing the settlements, (Buscher et al., 2009) (figure 1). Since while the UNHCR provided the financial and independence in 1962, Tanzania was considered technical support; “between 1963 and 1979, 13 one of the most hospitable countries in the world settlements were managed through such (Rutinwa, 1996 and Milner, 2013). Largely tripartite agreements, hosting an estimated credited to its first President, Julius Nyerere, it 182,000 refugees” (Milner, 2003). promoted an Open-Door Policy toward thousands of refugees fleeing wars, as well as However, the 1990s ushered in a significant shift liberation movements, including countries like in Tanzania’s posture toward refugees, from one Angola, Cape Verde, Comoro, Ethiopia, Kenya, of “self-sufficiency and local settlement” to one Mozambique, Namibia, Somalia, Seychelles, focused on repatriation (Milner, 2003). Factors South Africa, Uganda, Zaire and even people as that account for this change can be summarized far afield as Serbia (Chaulia, 2003 and Rutinwa, as follows: 1996). 1. changing international trends toward Critically, Nyerere’s Open-Door Policy was refugee policy and protection standards; guided by domestic economic incentives and the 2. the threat of a regional conflict and political formations of the time. Nyerere’s vision domestic insecurity; for Tanzania’s development was guided by the principles of ujamaa na kujitegema (socialism and 3. local party politics and the concurrent self-reliance), principles that were nationally change of the GoT’s foreign policy; adopted in 1967 through the Arusha Declaration 4. declining international financial (Milner, 2013 and Coulson, 1982). A significant support; element of the Declaration necessitated the 5. the unprecedented magnitude of refugee establishment of “farming collectives to presence and the accompanying stresses encourage self-reliance” (Milner, 2003). Within it placed on host populations; and the context of the Declaration, several scholars have argued that refugees were seen as a critical 6. the perceived failure of the Open-Door means to the economic development of especially Policy (Milner, 2003, Rutinwa, 1996 and peripheral regions in Tanzania. For instance, Whitaker, 2002b). Daley (1992) argues that refugee settlements were instrumental in the development of remote regions of the country both in terms of subsistence crops and export-earning crops that provided the Tanzanian government with invaluable foreign currency (Daley, 1992 and Milner, 2003). 3 Figure 2. Burundian Refugees in Tanzania Whitaker (2002b) asserts that to understand the shift in Tanzania’s policies in the 1990s, one should situate these changes in broader 600,000 international developments that tended toward 500,000 “more restrictive refugee policies and declining 400,000 protection standards.” For instance, according to 300,000 a 1997 UNHCR publication, in 1996 more than 200,000 20 countries ejected refugees from their 100,000 countries (UNHCR, 1997). Whitaker (2002b) 0 argues that the easing of relations between the 2007 1991 1993 1995 1997 1999 2001 2003 2005 2009 2011 2013 2015 3-Apr-17 Soviet Union and the United States ushered in a change in the nature of conflicts and more Returnees Refugees importantly, how asylum-seekers were perceived in this changing geopolitical context. Wars perpetuated by the Cold War were now replaced Source: UNHCR Online Population Statistics with conflicts that were entangled with issues of Database (accessed on March 6, 2017), UNHCR “identity [and] nationalism” (Whitaker, 2002b). Burundi Situation Regional Update #32 As a result, “refugees were no longer perceived December 2016, UNHCR Burundi Operational as victim of broader geopolitical conflicts but Portal (accessed on May 5, 2017) rather actors in the conflicts. As support for Western allies declined, host countries in the Note: The naturalization of about 162,000 developing world increasingly viewed refugees Burundian refugees accounts for the sharp as a source of instability and an economic decrease in the number of Burundian refugees in burden” (Whitaker, 2002b). Relatedly, there was 2009. a narrowing of durable solutions after the Cold Figure 3. DRC Refugees in Tanzania War. Whereas integration and third-party resettlement were common place during the 200,000 Cold War, these options were less favorable in a climate where refugee flows increased 150,000 significantly and thus the motives for their 100,000 movements were met with more suspicion (Whitaker, 2002b). 50,000 Further, Rutinwa (1996) proposes that the 0 influence of international precedents influenced 1992 1991 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 17-Oct-16 the Tanzanian Government’s move toward restrictive policies. Notably, the Haitian refugees as well as the Cuban exodus—the so called “boat Returnees Refugees people”—were cited by senior Tanzanian officials at the time as a justification for their Source: UNHCR Online Population Statistics change in policy. They noted that it was unfair Database (accessed on March 6, 2017), UNHCR to expect poorer countries to uphold their DRC Regional Refugee Response Information international obligations, while major powers— Sharing Portal (accessed on May 6, 2017) such as the United States—failed to when “their own national rights and interests were at stake” (Rutinwa, 1996). 4 Significantly, the nature of refugee populations Notably, although there was significant and the association of them with threats to international assistance at the beginning of the domestic and regional insecurity were Rwandan influx, over time, this assistance instrumental in justifying a change in policy. diminished. In part, much of the humanitarian The perception and reality that some among attention shifted to Bosnia and elsewhere, and as refugee flows from Rwanda and Burundi were such, by 1996 the UNHCR was struggling to armed, heightened domestic insecurity and fund the refugee response in the Great Lakes threatened tensions between countries of origin Region (Whitaker, 2002b and Milner, 2003). and host countries (Whitaker, 2002b). In 1995, Some scholars claim, in fact, that “failures of the the Minister of Foreign Affairs noted: international community to give adequate assistance to Tanzania was the main reason for The presence of the refugees is a source of the closure of the border” with Burundi, and that tension in the relations between Tanzania this border closure signified a shift in the GoT’s and Burundi and to a certain extent policy toward refugees (Rutinwa, 1996). Rwanda, arising from the suspicion that the refugees are regrouping and training in Finally, the unprecedented magnitude of warfare for attacking the countries of origin refugees arriving in Tanzania in the 1990s and (Rwegasira, 1995). some of the detrimental impacts it had on the environment, infrastructure and services, as well Within the context of growing regional as a belief by the government that the Open- insecurity, as well as the view by the Tanzanian Door Policy had not achieved its intended government that the security situation in objective—to provide interim relief, while more Rwanda had improved, the decision to repatriate long-term and permanent solutions are Rwandan refugees was made; the GoT believed determined in the country of origin – also this decision would protect it from being drawn facilitated a move toward more restrictive into a regional conflict (Whitaker, 2002b). policies (Rutinwa, 1996). The Foreign Minister in his Arusha speech echoed these sentiments: An often-underestimated factor in under- standing refugee policy changes at the time was Experience has proved that such measures as the change to a multi-party system of granting of permanent refugee status, government in 1992 and the impending elections permanent settlement are not a formula for a in October of 1995 (Milner, 2003). In their permanent solution to the refugee crisis. The campaigns, parties claimed they had the solution indeed lies in the countries of origin solutions to the refugee crisis, while Mkapa rather than in the countries of asylum which claimed he “could prevent the spread of violence are burdened with obligations [from] the from Rwanda and Burundi” (Milner, 2003). The refugees (Rwegasira, 1995). post-election era saw the instituting of a new foreign policy regime which moved away from Ultimately, the interplay of these multiple Pan-Africanism and in “support [of] liberation factors resulted in the 1998 Refugees Act, which movements” (Milner, 2003) toward “an active Kamanga (2005) argues had two objectives: policy of maintaining good relations with all 1. to “signal disengagement from the neighbors”, despite their ideologies or actions, as Open-Door Policy of the Nyerere long as they did not harm Tanzania’s interests administration” and (Rutinwa, 1996). 2. to “assure the populace” that the GoT was “determined to address the problem 5 of seemingly endless refugee influxes” higher concentration of refugees in these regions (Kamanga, 2005). (Maystadt and Verwimp, 2014): Furthermore, restrictions after the passing of the 1. proximity—there were higher Act deepened. For instance, refugees were concentrations of refugees in the prohibited from travelling more than 4kms from western part of Kagera compared to the the camps. Significantly, the 2003 National eastern part given proximity to Burundi Refugee Policy institutionalized many of these and Rwanda; restrictions, including controls on freedom of 2. government practice—authorities movement and economic activity, while rejecting decided to situate refugees closer to the citizenship as a viable durable solution, and border; and asserting voluntary repatriation as the “best solution for the refugee problem” (GoT, 2003). 3. natural topographic boundaries such as mountains prevented the mobility of Scope refugees. Most studies of impacts on host communities in Tanzania focus on the large exodus of people fleeing Burundi and Rwanda in the 1990s. These studies were therefore conducted within the context of an encampment policy with restrictions on refugee movements and with repatriation as the preferred durable solution as noted in the Background. Over one million people sought refuge in western Tanzania during this period, and in some regions, refugees outnumbered natives five to one (Whitaker, 2002). The outflow of refugees from Rwanda was concentrated mainly during the 1994–1996 period, while in the case of Burundi, which experienced a longer conflict, there was a more gradual but steady outflow of refugees throughout the 1990s. Kagera is the northern most region that borders Burundi and Rwanda in western Tanzania, while Kigoma lies south of Kagera, bordering only Burundi. These regions were the main locations in which refugees settled in the 1990s and as such have been the focal point of most of the studies covering impacts on Tanzanian host communities. Several reasons account for the 6 Kagera provides an especially useful case as concurrently, since these variables are closely refugees were not evenly spread across the related and therefore cannot be neatly region. A series of mountains separated east and disentangled. In fact, the review is intentionally west Kagera, and there are also a series of natural framed in a manner that enables policy makers reserves (mostly inhabited) that reinforce the to recognize the interdependences of variables geographical barrier between the two areas that affect hosts. Needless to say, hosts are (Ruiz & Vargas-Silva, 2015). These geographical affected by numerous variables simultaneously characteristics, government policies and and make subsequent livelihood choices on that proximity, resulted in a ‘natural experiment’ in basis. which an area (i.e., west) was much more affected by the refugee inflow in comparison to the other area (i.e., east). The following section captures the main findings of several studies that examined the presence of refugees in these regions. It covers the impact that refugees, and in some cases the humanitarian community, had on local communities in Tanzania. It starts by covering a brief overview of some mediating factors that influenced impacts, including pre-existing livelihood strategies in various refugee-hosting districts and immediate policy responses to the refugee influx, such as camp locations. The section then covers areas of research that have been more comprehensively analyzed: 1. labor market outcomes, with some studies placing more emphasis on the gendered dynamics, as well as the distinct impact on causal labor; 2. the local economy and food prices; 3. food security and prices in terms of the humanitarian impact; 4. local infrastructure and services; 5. environmental impacts; 6. security and social cohesion; and 7. long-run welfare impacts. Notably the variables covered here are not mutually exclusive in the way they affect hosts; they mediate impacts simultaneously. As such, there are cases where related impacts, such as the local economy and food prices, are discussed 7 Overview of the Impacts Due to poverty levels in Kibondo, few residents had enough “up-front capital to start businesses and many could not afford to hire refugee labor”, and as such, low levels of production meant that hosts could not take advantage of the new Before refugees arrive in any given context, markets that the refugee presence brought there are pre-existing factors that will mediate (Whitaker, 2002). the livelihood strategies that hosts pursue and their resultant short-term livelihood outcomes There was also a marked difference in district when refugees arrive (World Bank, 2016). These officials’ response to opportunities presented by factors, in addition to government policies, both aid agencies. In Kibondo, despite encouragement at a national and sub-national level, as well as by agencies for district officials to submit NGO and aid agency policies mediate impacts on proposals for grants for inclusive projects hosts. This paper begins with a brief review of between refugees and hosts, little was done. In these factors and their impacts in Tanzania Kasulu, however, “donor encouragement led to before addressing literature that covers labor regular government coordination meetings and market outcomes. project proposals by the district council” (Whitaker, 2002). Pre-existing Livelihood Strategies in Host Communities Whitaker (2002) further notes that the host population in Ngara was similar to that of Whitaker (2002) explores how different pre- Kibondo. Low levels of education, limited exiting factors within host populations business and trade experience meant that locals influenced the impact of the refugee situation. could not take full advantage of the refugee She notes how differences in poverty, education presence, despite there being a large population and capacity for agricultural production led to of refugees in Ngara. Further, locals of Ngara varied capacity for hosts to take advantage of experienced competition from Tanzanians from refugee presence in Karagwe and Kibondo other parts of the country; “people flocked there districts. from all over Tanzania to open businesses and exploit trading opportunities” (Whitaker, 2002). In the latter district, she highlights how poverty In 2006 when Rwandan refugees were limited the degree to which host communities repatriated, the same Tanzanians who had come could benefit. Kibondo’s remoteness, its to Ngara, moved to districts in Kigoma where communities’ dependence on subsistence Burundian refugees were still present. agriculture, as well as few “local business Ultimately, she notes that existing socio- ventures and limited economic opportunities”, economic conditions in the various districts resulted in less favorable livelihood outcomes influenced the extent to which hosts could compared to Karagwe. In contrast, Karagwe benefit from the refugee presence. district communities—where over 95% of residents sold coffee as a cash crop (Ndege et al., Refugee Policymaking and its Impact 1995)—benefitted from refugee presence on Host Communities because of their capacity to produce a surplus of food, and because of their “higher levels of In terms of how policy-making and practice education” (Whitaker, 2002). affected host communities, Whitaker (2002) asserts that government policies were particularly important in three ways: 8 1. camp locations; and Rwanda—they explore the implications of this shock for labor market outcomes of 2. restrictions on refugee-host interactions Tanzanians. Their results are consistent with after 1996; and immigration literature which shows that the 3. limits imposed on refugees to engage in arrival of refugees influences the economic agricultural production. activities in which locals engage. The latter, she argues, led more locals to be Ruiz and Vargas-Silva (2015) find that “greater exposed to “crop theft and banditry” since exposure to refugee presence resulted in refugees had less options to engage in Tanzanians having a higher likelihood of agricultural production. Camp locations had working in household shambas (farming) or varying effects on livelihood outcomes, caring for household livestock.” In other words, depending on host community proximity to the there was a positive correlation between the camps; nearby villagers were likely to benefit presence of refugees and an increase in farming more than those further afield, if they had the and livestock activities among host populations. capacity to do so. For example, the capacity to They found this result to be consistent across the produce more crops that refugees prefer and did different measures of the intensity of the refugee not have access to through aid. She also notes presence. that the “government’s effort to enforce tighter controls on refugee-host interaction after 1996 Their results also suggest that the influx of influenced the opportunities available to refugees did not affect the likelihood of having Tanzanians”. Lastly, she argues that on the self-employment as the main economic activity. NGO side, “decisions to implement development This result may seem surprising given projects in some locations and not in others— significant anecdotal evidence that Tanzanians often based on the recommendations of were opening numerous shops and starting government officials—also shaped host different businesses to service the needs of experiences” (Whitaker, 2002). refugees and employees of international organizations (Whitaker, 2002). However, this She concludes by asserting that overall, the finding is consistent with the idea that policies of both the Tanzanian government as Tanzanian farmers were rather expanding their well as aid agencies implemented soon after the farms in response to increased demand for their arrival of refugees, “had a significant impact on crops and in some cases with assistance from the extent to which host communities could cheaper refugee labor (Maystadt and Verwimp, benefit from the refugee response” (Whitaker, 2014). Ruiz and Vargas-Silva (2015) further 2002). argue that another possible explanation, which corresponds to previous literature, is that much Labor Market Outcomes: Jobs and of the new small business activity was initiated Wages by Tanzanians moving from other regions of the In their study, Ruiz and Vargas-Silva (2015) country to Kagera (Maystadt and Verwimp, exploit the ‘natural experiment’ of the 1990s in 2014). Kagera. Using panel data (pre- and post-refugee inflow) and making use of the Kagera Health and Furthermore, Ruiz and Vargas-Silva (2015) Development Survey (KHDS)—a longitudinal found that greater exposure to the refugee data set which contains information about population resulted in Tanzanians having a Kagera residents before and after the refugee lower likelihood of working outside the inflow or ‘shock’ of the early 1990s from Burundi household as employees. They were particularly 9 less likely to be agricultural employees, deterioration experienced by those involved in suggesting that refugees may have taken particularly small business could be explained as available agricultural job opportunities, hired by a selection effect resulting from the reported Tanzanian farmers. ` entry of larger-scale entrepreneurs from other regions, as opposed to the presence of the Similarly, Maystadt and Verwimp (2014) traced refugees themselves. how households between 1991 and 2004 were affected by the refugee inflows originating from Whitaker’s (2002) study assesses, not only the Burundi (1993) and Rwanda (1994) and argue refugee impact but also the humanitarian impact that refugees provided cheap labor in sectors on jobs and wages. She finds that the refugee such as agriculture, construction, housekeeping relief operation resulted in an increase in and catering. Their results show that local hosts employment opportunities for hosts and higher did not necessarily suffer from the refugee wages but also some negative consequences. presence. The combination of increased demand NGOs hired Tanzanians at all skill levels from for local produce and the availability of cheap “guards, drivers and maids to field staff, labor resulted in an expansion of agricultural administrators and accountants” (Whitaker production. In some villages close to refugee 2002). Waters’ (1996) research found that camps agricultural production doubled (World salaries linked to the relief operation were two to Food Program and UNHCR, 1998), three times the salaries of comparable positions in other parts of Tanzania. However, although Critically, constraints that were present prior to many locals in the refugee-hosting regions the arrival of refugees, such as “labor shortages benefitted from these new jobs and inflated and lack of markets,” affected agricultural salaries, other institutions such as government production (Whitaker, 2002). Specifically, the departments were negatively affected. For significant number of refugees expanded both instance, numerous employees from the public the number of laborers that hosts could hire and sector, including from “hospitals, schools, and the size of the local market. According to government departments left their positions” for Whitaker (2002), hosts promptly took advantage what were ostensibly better job opportunities in of refugee presence for farm expansion and to relief-related sectors (Whitaker, 2002). increase agricultural production. For instance, Whitaker’s (2002) study indicates that in Ngara, “in Karagwe district […] famers on average “more than 50 percent of health center staff and doubled the size of their cultivated lands and 35 percent of dispensary workers left their doubled their production of bananas and beans government posts to work with relief agencies.” between 1993 and 1996” (Whitaker, 2002). Further, employees with fixed salaries suffered greatly from the general rise in the cost of living; However, the economic benefits appear to have “the salaries of civil servants, bank employees, been unevenly distributed among the refugee- and parastatal staff did not cover nearly as many hosting population. Casual laborers (meant here expenses as they did” before refugee arrival and onwards as locals that work for self- (Whitaker, 1999). employed farmers) were likely to suffer the most from an increase in competition on the labor Labor Market Outcomes: A Gendered markets and the surging prices of several goods. Perspective In contrast, non-agricultural workers and self- employed farmers were in a better position to Gendered consequences of hosting refugees are benefit from such a refugee inflow. Maystadt and significant for several reasons. For instance, the Verwimp (2014) further argue that welfare arrival of refugees in rural areas often leads to 10 greater demand for resources such as firewood In some cases, however, Whitaker (1999) found and water. In rural Tanzania, household chores, that firewood became a source of income for both including fetching firewood for cooking and host and refugee women as they would sell the drinking water, are typically the responsibility of firewood by the bundle. Further, some local women. Conducting these chores on a regular women employed refugees for low wages to do basis in turn limits their access to income- household tasks such as fetching firewood and generating activities (Leavens and Anderson water (Ruiz and Vargas-Silva, 2017). This freed 2011 and Whitaker, 1999). up time for local women to engage in income- generating activities. Ruiz and Vargas-Silva It is within this context that Ruiz and Vargas- (2017) argue that where local women employed Silva (2017) in a separate and more recent paper, refugees to do household tasks, this increased expand on their previous study to examine their ability to engage in income-generating whether the presence of refugees alters the intra- activities and, for some local women, may have household allocation of tasks across genders in resulted in greater independence and “control the hosting population. The focus of their study over household spending decisions.” is on the impact of the refugee shock on three different groups of tasks: farming, outside However, the results differ by (pre-shock) employment and household chores (specifically literacy and mathematical skills. For women fetching water and collecting firewood). Their who could read and perform simple written overall findings show that the presence of mathematical actions, the refugee shock resulted refugees’ results in differing impacts on time in a “higher likelihood of engaging in outside allocation and tasks for men and women, which employment” (Ruiz and Vargas-Silva, 2017). In is further influenced by skill levels and age. contrast, they find that when illiterate women were exposed to the refugee shock, they were They find that overall the increased exposure to more likely to participate in farming and refugees led to women being less likely to household chores. Thus, their study indicates a engage in employment outside the household differentiated impact on women with different and more likely to engage in household chores literacy and mathematical skills. The presence of relative to men. This was ascribed to the refugees, therefore, potentially benefitted those increased time spent by women collecting with the skills because they were “more likely to firewood because of deforestation associated take advantage of the additional supply of cheap with the arrival of refugees, as well as the labor represented by refugees” Ruiz and Vargas- additional competition for natural resources Silva’s (2017). such as wood and water (Ruiz and Vargas-Silva, 2017 and Jacobsen, 1997). Berry (2008) found Labor Market Outcomes: Age and that increases in tree felling was also related “to Gender wind- and water-induced soil erosion” as well as the depletion of water resources. This is The results of Ruiz and Vargas-Silva’s (2017) especially critical to note since the communities study are substantially different across age sampled by Ruiz and Vargas-Silva (2017) were cohorts too. Firstly, for those women that were highly dependent on natural resources, not only 30 years of age or younger, refugee presence for cooking but for drinking water too. Further, resulted in them being less likely to engage in women were often only capable of doing farming outside employment and more likely to engage or fetching fire on any given day—not both in “farming and fetching water/collecting (Whitaker, 1999). firewood than men”. In contrast, for those over 30, refugee presence did not have much of a 11 gender-specific effect. This finding is in line with with women in Ruiz and Vargas-Silva’s (2017) what Ruiz and Vargas-Silva (2017) hypothesize study. in their paper, in part based on other literature: “younger individuals, who are more flexible in A study by Whitaker (2002) explores data the labor market, would be more likely to make captured between 1996 and 1998 in a total of 15 adjustments for the presence of refugees than villages and 5 districts, including in Ngara, older ones.” In justifying this assertion, they cite Kibondo, Kasulu and Kigoma on the impacts on literature on the labor market impacts of casual labor and access to basic resources. immigration in high-income countries, which Whitaker (2002) notes that refugees represented intimates that younger local workers are more a source of cheap labor for Tanzanian villagers. likely to compete with migrants in the labor Her findings show that given a significant drop market and more so then their older in acceptable wages—in some areas up to 50%— counterparts (Angrist and Kugler 2003). local farmers generally hired refugees to do agricultural work but also to build houses, tend In terms of the impact of the refugee shock by livestock and fetch water or firewood. household activities of those who were children Furthermore, wages varied depending on the (7 to 14 years of age), “higher household farms distance from the camps and the type of exposure to the refugee shock was associated work. In camp areas, where there was a larger with girls dedicating more time to outside supply of causal labor, local workers earned employment and collecting firewood/fetching significantly less than in other areas (Whitaker, water than boys” (Ruiz and Vargas-Silva, 2017). 1999). In fact, nearly three quarters of the time, These results suggest that hosting refugees refugees were paid with food instead of money increase girl’s participation in household chores (Kibreab, 1985). However, wages were higher and might lead to worse future outcomes. during the agricultural season when labor Interestingly, the refugee shock had no impact demand was higher, nevertheless, some on school attendance, which suggests that the concerned Tanzanians still hired refugees in the increase in time dedicated to other activities did low season even if their labor was not required not come at the expense of school attendance. It (Whitaker, 2002 and Maruku Agricultural is unknown, however, if girls’ outcomes at school Research Institute, 1997). were affected, as this issue was not explored in the study. Critically, many of the locals who were casual workers before the arrival of refugees changed to Labor Market Outcomes: Casual other activities, including self-employment in Labor the post-shock period (Ruiz and Vargas-Silva, 2015). These local casual workers were the most Tanzania also has a long history of casual labor likely to have competed with refugees for jobs. (Mbilinyi, 1986). Casual workers are typically Ruiz and Vargas-Silva (2015) note that these hired daily to do basic jobs with low degrees of results are, in general, “consistent with the responsibility for which they receive relatively evidence from the ‘voluntary’ migration low payment. Refugees were generally willing literature in developed countries which suggests and crucially able to engage in casual labor for that natives adjust to immigration flows by even lower payment (because of support they changing economic activities.” Given the dearth received from the relief operation in terms of of evidence on this type of adjustment of local non-food assistance, food rations, etc.). This population in low-income countries, Ruiz and could have led to a substitution of casual local Vargas-Silva (2015) emphasize that this finding workers for casual refugee workers, as was seen is an important contribution to the literature. 12 Overall, both the sharp decline in day labor Further, the boom in the local market was not wages as well as the rise in food prices resulted limited to refugee-hosting areas; “entre-preneurs in the welfare of local casual laborers becoming and aid agencies conducted considerable precarious (FAO, 1995). business at supply centers in Bukoba, Mwanza, Kigoma and Dar es Salaam” (Whitaker, 1999). The Local Economy and Food Prices Critically, prior to the influx of refugees, Ndege et al (1995) found that local markets were The arrival of the displaced usually results in the insufficient for their harvest and as such, they growth of local economies (de Montclos, M.-A. often traded across the border in Rwanda and P. and Kagwanja, P. M., 2000). As consumers Burundi. However, in the wake of the refugee who in some cases receive remittances, their crisis, trade increased at a village level; the purchasing power benefits local business owners arrival of refugees effectively moved markets and traders. However, increased demand for closer to villagers. Both the instability in certain products can result in increased prices Rwanda and Burundi, as well as the sudden which adversely affects local consumers (World arrival of refugees, had negative consequences Bank, 2016). Informal trade has also been noted for border traders and communities from border as a source of social tension when the displaced trading towns. The new population centers were and hosts trade the same goods or in the same now refugee camps, which were typically 20-40 location. Further, depending on the nature of km away from the border (Whitaker, 1999). This food-aid provision, local producers and meant that towns that were once economic consumers can be negatively or positively centers “were negatively affected by this abrupt affected (Mabiso et al., 2014). collapse of local [border] markets” (Whitaker, 1999). Whitaker (1999) found that the presence of refugees, as well as the resources associated with Mabiso et al. (2014) contend that the large-scale their arrival, changed economic opportunities for arrival of refugees can improve market efficiency hosts. The arrival of refugees resulted in the and trade dynamism, in part because of road growth of some local markets and with that came investments made by international a sharp increase in the participation in business organizations—“given the strong link between activities of both hosts and refugees. Critically, road accessibility and economic development” Mabiso et al. (2014) note that “land availability (Mabiso et al., 2014 and Jacobsen, 2002). This in […] northwestern Tanzania facilitated the market efficiency and trade dynamism was also expansion of agricultural production.” Local observed in Tanzania. Refugees also sold non- farmers traded and sold a variety of products to food items such as blankets and plastic sheets. both expatriate and refugee communities, Such trading activities were easy to observe and including “sweet potatoes, cassava, pineapples, even institutionalized by the aid community and palm oil, vegetables, bananas, and local brew” the Ministry of Home Affairs (MHA) through (Whitaker, 1999). Refugees themselves took the creation of a “common market” around each advantage of the food items they would be refugee camp, where refugees and local people provided by relief agencies. Some agencies were invited to trade. For example, Maystadt claimed that refugees traded up to 75% of their and Verwimp (1999) note that the “common food, including “vegetables, soy beans, flour, market of Lukole—opened between 1994 and plastic tarps, soap, and even farming hoes” 2003 close to one of the largest Rwandan refugee (Whitaker, 1999). camps—is estimated to have been the biggest market of the Kagera region, after the one in the capital town (Bukoba).” 13 Food Security and Prices: Refugee etc.) and their staff (local and international) and Humanitarian Impact induced significant increases in demand from people with much higher purchasing power. The sudden population increase affected food security in local villages, particularly at the Alix-Garcia and Saah (2010) found large beginning of the influx in 1994 (Whitaker, increases in the prices of non-aid food items (e.g., 2002). Despite receiving rations, refugees plantains, legumes, milk and beans) and more remained dependent on local crops and livestock. modest prices effects for aid-related food items. Out of a need to diversify their diets, refugees Thus, they found that food aid ameliorated looked to food from local farmers, whose produce (although did not offset completely) the impact included “vegetables, cassava, cooking bananas, on prices of the population increase which and sweet potatoes”, while in contrast their resulted from the presence of refugees (World rations mainly consisted of “beans, maize, Bank, 2016). Their examination of household cooking oil, and salt” (Whitaker, 2002). To assets suggested positive wealth effects of access these foods, refugees traded with, and refugee camps on nearby rural households and purchased from, local farmers, and in some cases, negative wealth effects on households in urban stole from them. areas. This finding is consistent with a scenario where producer households benefit from higher Due to increased demand in certain local crops, prices for agricultural goods. Overall, Mabiso et particular food prices sharply increased. al. (2014) note that in cases such as Tanzania, Bananas, for instance, were popular for both Kenya and Uganda prices are significantly refugees and hosts, as such, prices for bananas affected by the arrival of refugees—however the skyrocketed (Whitaker 2002). In response to “general equilibrium and long-term effects […] these market forces, many Tanzanian farmers depend on the extent of food aid inflows and the sold high proportions of their own food stocks ability of households to adjust their production (FAO, 1995). Various relief organizations tried and consumption decisions to changes in prices.” different strategies to avoid creating scarcity of supply. The WFP, for instance, “purchased Local Infrastructure and Services beans and maize for refugee rations from other regions of Tanzania and neighboring countries” Infrastructure and services (Whitaker, 2002). Although this resulted in temporary relief in western Tanzania, it Studies related to the impact of inflows on nevertheless resulted in the prices of these goods infrastructure and services in Tanzania have to sharply decline, as refugees sold their rations generally found that border area schools were in order to purchase other items. “Tanzanian damaged in the early weeks of the influx farmers who produced surplus beans and maize (refugees slept in classrooms and burned desks were thus unable to sell them for any profit at for firewood); local health facilities and referral all” (Whitaker, 2002). hospitals became overstretched; and the criminal justice system was overburdened (Whitaker Similarly, Maystadt and Verwimp (2014) also 1999). found that the price of some goods sharply increased, threatening the food security of some Whitaker (1999) notes that social services were households. The increase of prices resulted from initially insufficient to meet demand but an increasing demand from aid workers but also ultimately improved after the construction of from the refugees themselves. The arrival of infrastructure in the camps and the international organizations (UNHCR, NGOs, implementation of development projects in host 14 communities. Specifically, donors responded diseases (15-20%), an increase in mortality for positively, and “throughout western Tanzania, children under five (7%); childhood exposure to more than 50 primary schools and 20 refugee crisis reduced height in early childhood dispensaries were rehabilitated, 4 district by 1.8 cm (1.2%), schooling by 7.1% and literacy hospitals expanded, 120 water systems were by 8.6%. improved or installed, a community center was constructed, and several teacher resource Baez (2010) also found important human capital centers were built” (Whitaker, 1999). Host consequences for younger cohorts affected by communities proximate to the camps also the presence of refugees. He hypothesizes that benefitted from access to the health facilities poor childhood health can disrupt human capital within the camps. Women in particular found accumulation and affect labor market outcomes access to free health care—something they had in adulthood. That is, decrease the likelihood of to partially pay for before the arrival of doing jobs which require more human capital. refugees—as a significant gain due to the refugee Baez (2010) also hypothesized that this impact presence. was due to an increase in the prevalence of infectious diseases and vector-borne illnesses or Further, donors invested in road and in the competition for various resources (labor, transportation infrastructure. For example, food, land and wood) caused by the arrival of Whitaker (1999) highlights that in Kagera refugees. One caveat in the Baez’s (2010) study, “more than $15 million went toward the however, is that at the time he did not observe rehabilitation of main and feeder roads, airstrips, the final adult height of the children studied, and telecommunications infrastructure.” only their height in puberty or just before onset Further, donors pledged to improve the main of puberty. Recent studies in human biology, Kibondo-Kasulu-Kigoma road in late 1998. however, show that puberty offers an opportune These kinds of investments enabled travel for window for recovering height growth losses host communities, and as traffic increased, local experienced in early childhood (Mabiso et al. businesspeople established bus services 2014). (Whitaker, 1999). It is within the context of these human biology Health outcomes studies that Mabiso et al. (2014) extended Baez’s (2010) study to several years after refugees More detailed studies related to services have repatriated, and they did not find evidence of focused on health and education outcomes of long-lasting health impacts. Their findings hosts. Baez (2010) argues that the health suggest that the children whose growth, as outcomes of host children are adversely affected measured by height, was hindered due to the by hosting displaced populations. Using the arrival of refugees could catch up with the 1992 and 1996 Tanzania Demographic and control group during puberty. In this same vein, Health Survey and the 1991-1994 and 2004 Maystadt and Verwimp (2014) note that Kagera Health and Development Survey, Baez refugees had a limited impact on health (2010) notes a 15-20% increase in the incidence outcomes, especially given the qualitative of infectious diseases, a 7% increase in the evidence reporting health services to have mortality of children under the age of 5 and a improved following the refugee inflows. reduction in height in early childhood by 1.2%. The study also found adverse impacts over a year Generally, researchers have found that the host after the shock: worsening of children’s government’s ability to manage a sudden anthropometrics, an increase in the incidence of increase in demand for health services 15 determines the impact on healthcare provision. Whitaker (1999) also explains that refugees in This is in turn dependent on its “financial and Tanzania used more firewood per person than administrative capacity” (Mabiso, et al, 2014). the locals; “refugees used an average 2.8 kg of For instance, “during the Rwandan refugee wood per person per day, whereas locals used 1.7 crisis, the Tanzanian government was much kg” (UNHCR, 1999). Two reasons for this better equipped to handle the shock to its difference was that refugees were less likely to healthcare system than was its Zairean (now put out fires between meals because of a lack of DRC) counterpart.” (Mabiso, et al, 2014). matches and that they depended more on dried food, which took longer to cook than the fresh Environmental Impacts crops consumed by locals. UNHCR (2002) estimates that at the peak of the refugee crisis in The initial arrival phase of refugee influxes is Kagera, the camps consumed about 1,200 tons of often accompanied by severe environmental firewood each day and that by 1996, 225 km2 had impacts when the displaced move into, and been completely deforested and 470 km2 through, an area to secure their immediate needs partially deforested. UNHCR and other (World Bank, 2016). Negative impacts are most organizations did establish tree-planting felt in terms of land, water, natural resources and programs later on in order to combat slum growth. As the emergency period passes deforestation and soil erosion (Renner, 2007). and refugees become settled, the nature of the environmental impact changes but can still be Other scholars, however, have provided an significant. Environmental impacts are closely alternative interpretation for how refugees associated with the type of refugee settlements affected the environment and the discourses and particularly the concentration of people in related to this. Bonne-Moreau (2012) studied large camps/ settlements. Mtabila Camp in Kasulu district during the mid- 1990s. In his paper, he contextualizes a shift in As explained by Berry (2008) in her study of discourse and policy on the environmental environmental degradation and its impact on impact of refugees in the run-up to the refugee-host relationships in Tanzania, the presidential elections of 1995 and the general presence of refugees meant that it was necessary adoption of “anti-refugee policies and actions” of to “travel much greater distances to find the time. Bonne-Moreau (2012) notes that firewood and wood for construction than was “concerns of the environment—deforestation necessary 10 years [prior].” The environmental and land degradation in particular—were put impacts of refugees indirectly affected the food forward alongside human security issues by the security of the host community through Tanzanian government as a way to justify deforestation, soil erosion and land degradation, encampment policy interventions in a widely unsustainable water extraction and water accepted narrative.” He goes on to note that state pollution, which had both short-run and long- policies played a central role in human- run effects (Whitaker 2002; Martin, 2005). environment interactions. In Kasulu district in Although deforestation was a problem prior to particular, “environmental discourses [were] the arrival of refugees, the presence of refugees used to supplement, or perhaps, complement, in Tanzania accelerated deforestation rates and [existing] political and security discourses in depletion of soil nutrient availability for refugee contexts in order to justify continued agricultural crops, causing additional soil demands for funding associated with the erosion and thereby affecting the host’s refugee’s operation in Tanzania.” Bonne-Moreau agricultural production and food security (Berry, (2012) does not disagree that refugees did in fact 2008). contribute to a pre-existing context of 16 environmental degradation, but he argues that throughout all districts, had a total population of their depiction as ‘exceptional resource 1,739,183 of whom about 1,355,000 were degraders’ was part of a larger narrative that Tanzanians and 384,183 were refugees. This failed to account for the fact that it was difficult meant refugees constituted 22% of the total to distinguish between potential refugee impacts population” (Rutinwa, 2003). The ratio of and agricultural and land use practices of local refugees in the population is parallel to the ratio communities in the borderland (Berry, 2008 and of crimes committed, simply indicating that Bonne-Moreau 2012). He thus concludes that “as refugees did not necessarily have a greater the [refugee] situation became protracted and propensity to commit crimes than their insecurity decreased, claims of environmental Tanzanian counterparts in the region (Rutinwa, concern gained momentum” and were thus 2003). instrumentalized to justify the further containment of refugees in camps. In another study, the criminal justice system was found to be overburdened, with 75% of inmates Security and Social Cohesion being refugees (Lwehabura et al., 1995). This high rate of refugee inmates may have been Little research has been conducted on the impact influenced by the refugee policy of the time, of refugees on social cohesion and security in which restricted movements outside camps, and Tanzania. One study, however, found that crime as such, refugees were jailed for being outside rates rose sharply, especially for murder, robbery camps without the requisite documents. and possession of illegal firearms (Lwehabura et Interestingly, Whitaker (2002) found that al., 1995). However, despite officials attributing refugees were not necessarily blamed by increased crime to refugees, Whitaker (2002) villagers for increases in crimes, instead they found that crimes were also committed by “thought of it as an inevitability with such a Tanzanians, and in some cases with Tanzanians; drastic increase in population density.” “refugees and locals would sometimes cooperate to rob local communities, and armed banditry In terms of social cohesion, Whitaker (2002) was a problem prior to the arrival of the found that Tanzanians established extensive refugees” (Whitaker, 2002). Similarly, Rutinwa relationships with refugees, including attending (2003) found that government officials social functions and intermarriage. acknowledged within their own reports that Nevertheless, some negative perceptions and increases in crime were not fully attributed to social relations were persistent. She notes in her the influx of refugees. Refugees had certainly research that “camps were perceived as places of been party to crimes, but the extent to which this drunkenness, prostitution and sexual affected local communities was in some cases promiscuity. The elderly perceived a breakdown minimal (Rutinwa, 2003). of the traditional social structure” (Whitaker, 2002). Rutinwa (2003) studies the rate of criminal cases linked to refugees between the years 2000-2001. Interestingly, Whitaker (2002) finds that the He finds that the rates of criminal cases linked to varied backgrounds of the refugee populations refugees were significantly higher in Kigoma. impacted not only the nature of interactions Nevertheless, Rutinwa (2003) shows that when between them and their hosts but also the considering the population ratio of refugees to opportunities that their hosts had. She notes that locals, refugees did not commit more crimes than “Rwandan refugees in Karagwe were primarily the general population. “As of 2002, Kigoma, the farmers like their local hosts, while in the camps only region where refugees [were] spread in Ngara included ‘wanjanja wajanja’ (con artists) 17 who came from towns and cities in Rwanda.” In contrary, the effects became more positive Ngara, villagers there asserted that Burundian between 2004 and 2010. They further refugees who arrived in 1993 “caused fewer investigate the possible channels of problems than subsequent Rwandan refugees, in transmission, underscoring the importance of a part because they were ‘peaceful farmers just like decrease in transport costs (due to increased [them]’” (Whitaker, 2002). road provision) as a key driver of this persistent change in welfare. They interpret these findings In Kasulu, hosts claimed that Burundian as the ability of a “temporary shock to induce a refugees from cities, had a worse impact after persistent shift in the equilibrium through their arrival in 1996. Further, farmers from the subsequent investments”, specifically, they argue rural district of Kigoma criticized Congolese that one of the channels through which refugee refugees for their attitudes; they “refused to presence had a positive impact was through the work on Tanzanian farms and demanded food, investment in road infrastructure by the places to stay, and other assistance when passing UNHCR and WFP (Maystadt and Duranton, through local villages” Whitaker (2002). 2014). They refer to Whitaker’s (1999) study in Although Whitaker (2002) admits that the which she notes, “in Kagera, more than 15 research findings were based on anecdotal million went toward the rehabilitation of main evidence from villager’s impressions, these and feeder roads, airstrips and findings nonetheless suggest that refugee’s telecommunications infrastructure”, thus nationality and previous locality of residence making “internal transportation for host (city or rural dwellers and perhaps the communities cheaper and easier” (Whitaker, accompanying preferred occupations), 1999). They note that improvements in influenced hosts’ attitudes toward them and transportation is especially important in areas refugee opportunities. where remoteness is a critical factor in hindering a community’s ability to fight poverty (De Long-run Impacts on Welfare Weerdt, 2006). A very limited number of studies have been Furthermore, the expansion of transportation conducted in many displacement contexts that infrastructure, they find, also had an impact on address long-run impacts of any form. The need prices of goods, which in turn had positive for long-term impact studies continues to be a welfare outcomes. “The welfare-improving significant research gap. impact of road accessibility in high-refugee areas is further corroborated by the decreasing effect Maystadt and Duranton (2014), however, on goods prices” (Maystadt and Duranton, exploit a 1991–2010 Tanzanian household panel 2014). Particularly in remote rural areas like to assess the effects of the temporary refugee Kagera, improvements in road infrastructure has inflows originating from Burundi (1993) and a decreasing effect on the prices of traded goods Rwanda (1994). Their study is of significance (Casaburi et al. 2013). because compared to other studies on impacts, theirs does not focus on short-run impacts but Their study also notes other reasons that may rather on consequences of hosting refugees account for the persistence of positive, long-run almost ten years after they have left. They find effects of the refugee inflow. Critically, the that the refugee presence had a persistent and refugee camps in particular, and the economic positive impact on the welfare of the local opportunities that arose from them, attracted population. Interestingly, they found that the economic migrants from other parts of Tanzania. positive effects did not fade over time. On the This inflow of economic migrants followed the 18 inflow of humanitarian aid. Many of these to which the refugees repatriated and the trade economic migrants subsequently stayed after the that continued with them and locals thereafter. refugees left. In this context, Maystadt and Through interviews of Red Cross officers, they Duranton (2014) highlight several reasons that find that “refugees repatriated just beyond the could explain the persistent positive impact of border and continued to trade with the local refugees, including (de Montclos and Kagwanja population”. This anecdotal evidence is 2000; Buscher and Vlassenroot 2009; Alix- consistent with research conducted in other Garcia, Bartlett, and Saah 2013): contexts in which economic exchanges continue between hosting communities and displaced 1. due to increase in population, labor people (after their return) (Burchardi and markets were more efficient because of Hassan, 2013). However, despite the above labor pooling; anecdotal evidence, their empirical evidence 2. more investments were made to further shows no strong impact of the refugee inflows on expand and maintain existing transport trade flows with neighboring countries. infrastructure; and 3. local authorities noted that there was Lessons Learned and Policy/ increased tax revenue given a surge in activity around refugee camps—some of Practice Recommendations which may have been invested in growth-enhancing sectors such as education or health services. Maystadt The evidence presented in this review, has and Duranton (2014) further important practice and policy implications, not substantiate their point by mentioning only for those who work in the refugee-hosting how “anecdotal evidence in other regions of Tanzania, but in other displacement countries suggests that refugee inflows contexts that are faced with similar challenges. may strengthen the urbanization Opportunities abound for humanitarian, process in the regions of destination”. development and government authorities—in myriad refugee situations—to work more closely Maystadt and Duranton (2015) go on to observe together to develop well-substantiated practices that the provision of local public goods could and policies that support the preparedness and have also improved through subtler channels. resilience of host communities, as well as the Improved management skills and institutional inclusion and self-reliance of refugees. efficiency were reported by local authorities, due to dealing extensively with international The following are some lessons that can be organizations. They argue that this engagement learned from studies that have assessed host with international organizations could have population impact and that could inform future enhanced local authority efficiencies with non- responses. They are accompanied by related governmental organizations long after the initial policy or practice recommendations. arrival of refugees. Interestingly, Maystadt and Duranton (2015) argue that another possible channel of transmission might have to do with the location 19 Table of Lessons Learned and Policy/Practice Recommendations Lesson Policy/Practice L1 Lesson (L1): The developmental needs and P1 Policy/Practice (P1): To the extent possible, challenges of a host country are development actors should pragmatically, instrumental in framing how the host identify how refugee’s socio-economic government may understand the value of integration can contribute to, and align with, refugees. This was evident from the previous the host country’s developmental objectives. Open-Door Policy which was framed in a context where refugees were understood as critical to the development of peripheral regions in Tanzania. L2 The sustainability of development responses P2 Development and humanitarian actors, to protracted forced displacement situations should to the extent possible, devise are highly contingent on the political will of responses in collaboration with host the government, sustained technical governments, and ensure their local political assistance and financial resources. The backing. In the best cases, the responses Tripartite Partnership Model between the should involve local actors (such as NGOs) UNHCR, GoT and TCRS is an example of and the use of government systems to ensure this. the sustainability of the interventions. L3 Government policies targeted toward P3 Government policies devised in response to refugees, can inadvertently have negative the arrival of refugees, should carefully implications for host communities. For consider the implications these may have for instance, limitations on refugee engagement hosts, both positive and negative. Policies in agricultural production left some refugees should be devised in a manner that benefits with little choice but to engage in village both groups, with consideration for short crop theft. terms costs and long-term benefits. Some of these include: P3.a: If a camp policy exists, smaller camps that are relatively near to villages, allow for hosts to take social and economic advantage of their presence, with less negative effects, particularly in terms of the environment; P3.b: Common markets and freedom of movement can enable mutually beneficial trade and labor exchanges between hosts and refugees; P3.c: More deliberate and closer collaboration between local officials as well as local and international agencies, can 20 mitigate against duplication of projects and funding wastage which allows for hosts to benefit more from development projects; and P3.d: The provision of refugees with small plots of land for agricultural production can enhance self-reliance and food security, and possibly reduce likelihood of petty crimes (like crop theft). L4 The most vulnerable in host communities P4 Protect the most vulnerable with an are disproportionately affected by refugee emphasis on labor market outcomes, gender presence, those tend to include women, the and youth since impacts have distributional elderly, the disabled, the already poor and effects. unskilled. L4a Consequences of hosting refugees are not P4a Programs and interventions should not only gender-neutral. Evidence showed that host be tailored to consider the gendered impacts women were less likely to engage in outside on the labor market and household tasks but employment and more likely to engage in also the differentiated impacts between household chores (i.e., water fetching and higher skilled and lower skilled women. firewood collection) relative to men. Further, literate women were more likely to engage in outside employment in response to the shock, while illiterate women were likely to engage in farming and collecting firewood/fetching water. L4 Research indicates that poorer households P4b Interventions should differentiate between b will likely benefit from more public goods “social and economic distributional effects” (e.g., hospitals/health centers) and services, (Mabiso et al.). To protect vulnerable groups however, they will likely not farewell in (e.g., poor households, casual workers) from terms of market-related economic the likely initial negative economic effects, opportunities that arise from an increased actors should identify the groups who may number of refugees (Mabiso et al.). This require support, evaluate their differentiated effect in terms of social and vulnerabilities and their coping strategies, and economic impacts, will likely result in develop responses accordingly. Targeted wealthier households being rewarded social safety nets can potentially have economically (e.g., new businesses or jobs in ameliorating impacts. Further, impact the humanitarian sector), while less off evaluations should be conducted to assess households will find themselves in the effectiveness of the responses. precarious conditions (e.g., casual farm workers competing with the cheap labor of refugees). 21 L5 Refugees also occupy other roles and P5 Humanitarian and development actors identities, such as students, consumers, should be as concerned with refugee policies businesswomen etc. Therefore, their and practices as they are with policies and livelihoods are certainly shaped by practices that pertain to housing, health, protection policies and practices, but are not education etc., particularly in contexts of exclusively framed by protection concerns. non-encampment. It is also these policy areas that most intersect with the lives of host communities (Landau and Achuime, 2017). L6 Humanitarian and development actors can P6 Humanitarian and development deplete human capital from public organizations should try to mitigate the institutions due to inflated salaries offered by depletion of labor from public institutions. these organizations. This could be done, for instance, through labor and skills exchange agreements between public and humanitarian agencies to protect from the depletion of skills, and ensure skills transfers to public institutions that will be of use after the humanitarian organizations have left. L7 Labor and goods markets as adaptation P7a Investments and responses should be geared mechanisms are vital in refugee situations. toward building the skills and capacity of These markets can provide mechanisms for local producers to respond to increased positive outcomes on food security and in demand in food. some cases, negative outcomes for some Program design should consider pre- subgroups of the host community. These existing socio-economic conditions and factors “depend on preexisting conditions P7b policies that constrain and enable host such as infrastructure, labor skill levels, land adaptation mechanisms—such as road availability, and agricultural potential, but infrastructure and encampment policies. likely also on refugee policy (for example, refugee work regulation, refugees’ access to land, restrictions on trade and refugee mobility, and so on)” (Mabiso et al., 2014). L8 The “timing, source and type of aid, as well P8 Humanitarian actors, to the extent possible, as where and to whom it is targeted, have should “substitute food aid for cash transfers significant implications for the food security or vouchers to both refugees and poor and resilience of the host community, households in the host community” (Mabiso especially if food and other goods can be et al., 2014). However, assessments of their purchased at affordable prices from local impact should be periodically conducted to markets” (Barrett and Maxwell, 2005 in ensure positive impacts. Mabiso et al., 2014). 22 L9 Environmental degradation is a critical P9a Food security of host communities could be mechanism through which refugees impact protected by 1) no encampment policies 2) the food security of host communities, in using a settlement structure, 3) in the case of both the short and the long run (Mabiso et encampment policies, careful selection of al., 2014). Further, environmental camp location and 4) environmental degradation can be used as a political tool to programs such as reforestation and soil justify restrictive policies on refugees, that conservation (Mabiso et al., 2014). are in fact contradictory with P9b Development and humanitarian actors environmentally sustainable solutions in should carefully consider how they frame protracted situations. environmental challenges in the face of a refugee shock as this may have implications for wider protection needs of refugees. L10 Interdependencies between refugees and P10 Humanitarian and development actors their hosts, last beyond the period after should work together early in the event of refugees have left the host areas Therefore, refugee arrival. Humanitarian short-term households and communities may require assistance should then pave the way for time and support to adjust to not only development interventions that support population shocks associated with the hosts in the gradual or sudden departure of sudden arrival of refugees but also with the refugees. gradual or sudden departure of refugees Further, socio-economic networks made (Mabiso et al., 2014). during the refugee situation, should be fostered by development actors after refugees are repatriated and if back to their country of origin, possibly through regional infrastructure and trade enabling responses. L11 Improvements in road infrastructure has a P11 Local governments and development actors decreasing effect on the prices of traded should prioritize road provision and goods (Casaburi et al. 2013). Improvements maintenance to support the reduction in the in transportation is especially important in cost of traded goods and transport costs. areas where remoteness is a critical factor that hinders community’s ability to escape poverty (De Weerdt, 2006). Road infrastructure has been shown to be a key driver of persistent positive changes in hosts welfare (Maystadt and Duranton, 2014). 23 Future Research Factors Areas for further research 1. Pre-existing Pre-existing We know that levels of education, access to capital, skills, livelihood socio-economic poverty levels etc., affect how various impacts are mediated, strategies in conditions within including jobs, wages, food security, the ability to reap the host host communities benefits of a growing local economy. communities ▪ In what precise and varied ways do these pre-exiting conditions mediate outcomes? ▪ What can be done by the various actors, early in an emergency, to ameliorate how these pre-existing conditions affect outcomes? The role of sub- There is little literature that looks at the role that local national authorities play in addressing forced displacement and the government impact this may have on host community and refugee/IDP livelihood outcomes, including the pre-existing capacity of local administers. ▪ In what ways to do the pre-existing capacities of local government actors in refugee-hosting regions affect outcomes for hosts? 2. Labor market Gender and age Recent literature on the gender-specific impacts of outcomes immigration is focused on high-income countries (Barone and Mocetti 2011; Cortes and Tessada 2011; Furtado 2015). This limits our understanding of the potential consequences of hosting refugees, since most refugees worldwide are in neighboring developing countries (UNHCR 2016). In this same vein, although it’s known that young girls school attendance is likely to be affected by refugee presence, little is known about the impact this has on girl’s educational outcomes. ▪ What are the differences and similarities in gender- specific labor market impacts in high and low incomes countries? ▪ What are the impacts on host women in the labor market across the informal vs. formal sector? 24 ▪ How are girl’s educational attainment and learning outcomes affected by refugee presence? Households and While there is a growing interest in estimating the economic skills impacts of hosting refugees (Azevedo et al. 2016; Balkan and Tumen 2016; Del Carpio and Wagner 2015; Ruiz and Vargas- Silva 2015 and 2017, 2016; Tumen 2016; among others), we know little about the consequences of refugee inflows on different household members and across skills levels. ▪ How are people within the same household and with varying skills affected by refugee presence? ▪ Some evidence shows that refugee presence affects the division of household chores (Ruiz and Vargas-Silva, 2017). However, little is known about how the change in household chores affects social relations within a household? ▪ How do female-headed and male-headed households differ in their ability to benefit from refugee presence? What differing challenges may these households endure in the face of a refugee shock? 3. Food security Cash-based The effects of different types and combinations of and resilience interventions humanitarian aid on food security and resilience in host communities are not well understood. Little is known about the relative efficiency of cash-based interventions in refugee- contexts despite its growing use in responses. ▪ How should humanitarian actors consider issues such as program design, the magnitude of the transfers and the frequency of the transfers required when choosing cash- based interventions? 4. Longitudinal Environment and Generally, more quantitative and longitudinal studies need to studies health be conducted in refugee settings, particularly related to environmental and health outcomes. 25 Bibliography Alix-García, J. (2007). “The Effects of Refugee Inflows on Host Country Populations: Evidence from Tanzania.” University of Montana. Alix-García, J. and D. Saah (2010). "The Effect of Refugee Inflows on Host Communities: Evidence from Tanzania." The World Bank Economic Review 24(1): 148-170. Angrist, J. D. and Kugler, A.D. (2003) "Protective or Counter-Productive? Labour Market Institutions and the Effect of Immigration on EU Natives," Economic Journal, v113(488, Jun), F302-331. Antman, F. (2014). “Spousal Employment and Intrahousehold Bargaining Power.” Applied Economics Letters, 21(8): 560–63. Baez, J. E. (2010). "Civil Wars beyond their Borders: The Human Capital and Health Consequences of Hosting Refugees." Journal of Development Economics. Barrett, C.B., and Maxwell, D. (2005). “Food aid after Fifty Years: Recasting Its Role”. New York: Routledge. Berry, L. (2008). “The Impact of Environmental Degradation on Refugee–Host Relations: A Case Study from Tanzania.” New Issues in Refugee Research, Research Paper 151. Bonne-Moreau, M. (2012). “Cutting the Heart of Tanzania”? Refugee Policies, Livelihoods and Natural Resources in Kasulu District, Tanzania: a Political Ecology of Mtabila Camp.” University of London. Burchardi, K. and Hassan, T. (2013). “The Economic Impact of Social Ties: Evidence from German Reunification.” Quarterly Journal of Economics, pages 1219–1271. Buscher, K. and Vlassenroot, K. (2009). “Humanitarian Presence and Urban Development: New Opportunities and Contrasts in Goma, DRC.” Disasters, 34(2):256–273. Casaburi, L., Glennerster, R., and Suri, T. (2013). “Rural Roads and Intermediated Trade: Regression Discontinuity: Evidence from Sierra Leone.” Unpublished. Chaulia, S. S. (2003). “The politics of refugee hosting in Tanzania: From open door to unsustainability, insecurity and receding receptivity.” Journal of Refugee Studies, 16(2), 147-166. Clark, L. (1987). “Key issues in post-emergency refugee assistance in Eastern and Southern Africa.” Washington, D.C., Refugee Policy Group. Coulson, Andrew (1982). “Tanzania: A Political Economy.” Oxford: Clarendon Press Daley, P. (1992) “The politics of the refugee crisis in Tanzania', in H, Campbell and H, Stein (eds) de Montclos, M.-A. P. and Kagwanja, P. M. (2000). “Refugee Camps or Cities? The Socio-economic Dynamics of the Dadaab and Kakuma Camps in Northern Kenya.” Journal of Refugee Studies, 13(2):205– 222. De Weerdt et al (2006) “Moving Out of Poverty in Tanzania’s Kagera Region.” World Bank. Feldman, D. (1971). "Appraisal of the economic viability of four refugee settlements in Southern Tanzania." Typescript. 26 Ellis, S., M. Blackden, J. Cutura, F. MacCulloch, and H. Seebens (2007). “Gender and Economic Growth in Tanzania: Creating Opportunities for Women.” The World Bank. Food and Agriculture Organization of the United Nations (FAO). (1995). “Tanzania: Refugee Affected Areas Development Project: Identification/Formulation Mission Report (Final).” Fujita, M. and Thisse, J. (2002). “Economics of Agglomeration, Cities, Industrial Location and Regional Growth.” Cambridge, MA, US: Cambridge University Press. Gentry, D. J. (2009). “Refugee impacts on host communities in Western Tanzania: A GIS and mixed- method geographic analysis.” Geography and Planning, Appalachian State University. Green, R. H. and UNICEF (1994). “That they may be whole again: Off-setting the refugee influx burden on Ngara and Karagwe districts.” Executive summary: Report for UNICEF, UNICEF. Government of the United Republic of Tanzania (2003), Ministry of Home Affairs “The National Refugee Policy”, Dar es Salaam, September 7. Huisman, H. (2003). "Refugee Aid and its Impact on the Host Country ‘s Regional Economy. The Case of Ngara District, NW Tanzania." Die Erde 134(1): 65-80. Jacobsen, K. (1997). "Refugees' environmental impact: the effect of patterns of settlement." Journal of Refugee Studies 10(1): 19-36. Jacobsen, K. (2002). “Can Refugees Benefit the State? Refugee Resources and African State-building.” Journal of Modern African Studies 40 (4): 577-596 Kamanga, K. (1995) “The Genocide Convention in Its Application to Tanzania: Rusticated Legal Obligations? paper presented at the Arusha Refugee Crisis Workshop. Ketel, H. and UNHCR Programme and Technical Support Section. (1994). “Environmental assessment report of the Rwandese refugee camps and the affected local communities in Kagera Region, Tanzania.” Dar Es Salaam, UNHCR/FAO Environmental Assessment Mission. Ketel, H. and UNHCR/FAO (1994). “Tanzania: Environmental Assessment Report of the Rwandese Refugee Camps and the Affected Local Communities in Kagera Region, 2-30 June 1994.” Dar es Salaam, UNHCR/FAO Environmental Assessment Mission. Kibreab, G. (1985). “African Refugees: Reflections on African Refugee Problem. Trenton: African Word Press. Kreibaum, M. (2016). “Their Suffering, Our Burden? How Congolese Refugees Affect the Ugandan Population.” World Development Vol. 78, pp. 262-287, 2016 Landau, L. B. and Achiume, E.T. (2017). “Misreading Mobility? Bureaucratic Politics and Blindness in UN Migration Reports”. Development and Change 48(5). International Institute of Social Studies Landau, L. B. (2003). "Beyond the Losers: Transforming Governmental Practice in Refugee-Affected Tanzania." Journal of Refugee Studies 16: 20-43. Landau, L. B. (2004). "Challenge without transformation: refugees, aid and trade in Western Tanzania." Journal of Modern African Affairs 42, 31-59. Leavens, K., and C.L. Anderson (2011). “Gender and Agriculture in Tanzania.” Evans School Policy and Research Working Paper. 27 Lwehabura, J.M.K, et al. (1995) “A Short Brief on the Refugee Problem and Security along the Tanzania- Rwanda-Burundi Border.” Office of the Prime Minister and First Vice President. Dar es Salaam: United Republic of Tanzania. Mabiso, A., Maystadt, J.-F., Hirvonen, K., and Vandercasteelen, J. (2014). “Refugees and Food Security in Host Communities: A Review of Impacts and Policy Options to Enhance Resilience.” 2020 Conference paper 2. Washington, DC: International food Policy Research Institute. Martin, Adrian. 2005. “Environmental Conflict Between Refugee Host Communities.” Journal of Peace Research, 42(3):329-346. Maruku Agricultural Research Institute and Ngara District Agriculture and Livestock Development Office (1997) “Diagnostic Survey of Ngara District”, Working Paper No 21., February. Maystadt, J.-F. and P. Verwimp. (2014). “Winners and Losers among a Refugee-Hosting Population.” International Association for Research and Teaching, ECORE, Belgium. Maystadt, J.-F. and Duranton, G. (2014) “The Development Push of Refugees : Evidence from Tanzania. ” Economics Working Paper Series, The Department of Economics Lancaster University Management School Mbilinyi, M. (1986) “Agribusiness and casual labor in Tanzania. African Economic History.” 15: 107–141. Milner, J. (2003). “Two steps forward, one step back: understanding the shifting politics of refugee policy in Tanzania”. New Issues in Refugee Research, Research Paper no. 255 Musoke, I. K. (1995). “The impact of the refugee crisis in the Kagera region of Tanzania: local community responses and the Government's reaction.” International Workshop on the Refugee Crisis in the Great Lakes Region. Arusha. Mwakasege, C. (1995). “Study report: impact of refugees on host communities, the case of Kasulu, Ngara and Karagwe Districts in Tanzania.” International Workshop on the Refugee Crisis in the Great Lakes Region. Arusha. Ndege, L.J. et al. (1995) “Tanzania/Netherlands Farming Systems Research Project, Lake Zone: Diagnostic Survey of Karagwe District, Karagwe District Rural Development Programme, October.” Ongpim, P. A. (2009). "Refugees: asset or burden?" Forced Migration Review 33. Oxford Refugee Studies Centre (2011). “Study on Impacts and Costs of Forced Displacement – State of the Art Literature Review”. Renner, M. (2007). “UN Refugee Agency Turns to Tree Planting”. World Watch 20(5): 7. Ruiz, I. and Vargas-Silva, C. (2013). “The Economics of Forced Migration”, The Journal of Development Studies. Vol. 49, No. 6, 772–784, http://dx.doi.org/10.1080/00220388.2013.777707 Ruiz, I. and Vargas-Silva, C. (2015) “The labour market consequences of hosting refugees.” Journal of Economic Geography (2015) pp. 1–28 Ruiz, I. and Vargas-Silva, C. (2017). “The impact of hosting refugees on the intra-household allocation of tasks: A gender perspective.” WIDER Working Paper 2017/66, United Nations University Rutinwa, B. (1996). “The Tanzanian’s Government Response to the Rwandan Emergency” Journal of Refugee Studies, Vol 9. No. 3 28 Rutinwa, B. and K. Kamanga (2003). “The Impact of the Presence of Refugees in Northwestern Tanzania.” The Center for Study of Forced Migration, University of Dar es Salaam. Rwegasira, J., MP (1995) “Guest of Honour's Speech, delivered at the International Workshop on the Refugee Crisis in the Great Lakes Region”, Arusha, Tanzania 16-19th August 1995 Turner, S. (2004). “New Opportunities: Angry Young Men in a Tanzanian Refugee Camp.” Refugees and the Transformation of Societies: Agency, Policies, Ethics, and Politics. P. Essed, G. Frerks and J. Schrijvers. Oxford and New York, Berghahn. United Nations High Commissioner for Refugees (UNHCR). (1997). “The State of the World’s Refugees: A Humanitarian Agenda.” New York: Oxford University Press. Waters, T. (1996) “Emergency Assistance and Development, or what has a Cash Bath done for Western Tanzania’s Wahangaza.” unpublished manuscript. Whitaker, B.E. (2002). “Refugees in Western Tanzania: The Distribution of Burdens and Benefits among local hosts.” Journal of Refugee Studies. Vol 15. No. 4, 2002. Whitaker, B.E. (2002b). “Changing priorities in refugee protection: the Rwandan repatriation from Tanzania”, New Issues in Refugee Research, Working Paper no. 53 Whitaker, B.E. (1999). “Changing Opportunities: Refugees and Host Communities in Western Tanzania.” New Issues in Refugee Research, Working Paper no.11. 29