JOBS INTERVENTIONS FOR REFUGEES AND INTERNALLY DISPLACED PERSONS

This Jobs Solutions Note is intended to identify practical solutions for development practitioners and policymakers to design and implement policies and programs for the labor market integration of refugees and IDPs. The Note (a) identifies the specific obstacles that refugees and IDPs face when integrating the labor market, (b) highlights interventions designed to address those barriers, focusing on the World Bank, and (c) summarizes existing knowledge on what works. It builds on a comprehensive literature review on Jobs Interventions for Refugees and IDPs, and is part of a broader work on Supporting Jobs in Fragility, Conflict, and Violence Situations. The Note uses a broad definition of jobs as any income source, formal or informal, including employment, self-employment and entrepreneurship. 1 and middle-income countries, especially related to IDPs, demand-side interventions, and longer-term outcomes from interventions.

¬ The forcibly displaced face significant and multiple obstacles: legal restrictions, loss of assets, physical and mental health issues, skills mismatches, lack of social networks and market information, excessive labor supply, and discrimination.
¬ Interventions to help the forcibly displaced access labor markets must address these multiple constraints simultaneously. How important the different obstacles are in practice depends on the individual characteristics of those forcibly displaced as well as their country of origin and destination.
¬ Thorough assessment of labor market demand and supply characteristics is especially important to design effective interventions in the context of forced displacement. Assessment needs to include the legal context and perceptions and aspirations of the forcibly displaced.
¬ Making up for assets lost during displacement through cash injections seems particularly important.
¬ Healthcare and psychosocial support can help refugees and IDPs overcome trauma, improve their outlook and integrate into labor markets.
¬ Relaxing legal restrictions related to if, when, and how the forcibly displaced have the right to work, move freely and obtain permanent residency status significantly improves labor market outcomes.
¬ In terms of job matching, intensive coaching and individualized assistance seem to help, but it is challenging to implement them cost effectively.
¬ Interventions to build social networks and overcome spatial mismatches seem promising and require further testing and evaluation.
¬ Rigorous evidence on effects of jobs interventions for the forcibly displaced is scarce, and mainly comes from high-income countries. More rigorous research is needed in low and middle-income countries, especially related to IDPs, demand-side interventions, and longer-term outcomes from interventions.

MOTIVATION: WHAT IS THE PROBLEM? The labor market outcomes of refugees and IDPs lag behind locals and economic migrants
Jobs are key for the nearly 80 million forcibly displaced worldwide to become self-reliant, re-build their lives, and integrate into the host communities and contribute to the economy.
Having a job allows individuals to obtain an income and meet their needs and the needs of dependents, offering a pathway out of poverty. It also gives a sense of purpose and allows one to feel he or she is a contributing part of a community. This is also true for those forcibly displaced by conflict, violence, persecution, and human rights violations. Of the 79.5 million forcibly displaced people worldwide, 45.7 million are internally displaced in their own countries, 26 million are refugees (including 5.6 million Palestinians), 4.2 million are asylum seekers, and 3.6 million are Venezuelans displaced abroad. About 85 percent of all the forcibly displaced worldwide live in low and middle-income countries. Many are in protracted displacement situations lasting more than five years.
Rigorous studies on high-income host countries in Europe and the US consistently show that refugees have lower employment rates and income than locals and take longer to integrate into the labor market than economic migrants. The gap remains, even after accounting for different individual characteristics between refugees, locals and economic migrants, such as mental and physical health, lower level of education and language skills, gender and age. Female refugees have particularly low employment rates. Refugee employment rates and wages also vary by country of origin and destination. Over time, however, refugees can manage to narrow the employment gap. 2 Statistics from representative cross-sectional refugee surveys in low and middle-income countries show a similar pattern of lower employment rates and wages, and higher likelihood to work in the informal sector. In Uganda, 54 percent of refugees report aid as their main source of income. In Kampala, 74 percent depend on remittances as their main source of income. Even if the legal and policy framework is very conducive in Uganda, 3 of 4 refugees are unemployed, twice the rate of hosts. 3 In Ethiopia, aid is the major source of livelihood for over 80 percent of refugees after displacement, compared to less than 10 percent before displacement, and only 22 percent of working-age refugees are employed. 4 Refugee women have even lower labor force participation and employment rates than refugee men in the Kalobei refugee camp in Kenya. 5 Large employment and wage gaps also exist between refugees and hosts in Jordan, Lebanon, and Turkey, and these gaps are even larger for refugee women. Refugees who do have work are more likely to be working informally compared to host country counterparts, for example in Jordan and Lebanon. 6 Evidence on IDPs also shows negative association between forced displacement and employment rates and wages. Using propensity score matching (PSM), a study finds that IDPs in Colombia who are working receive significantly less wages, compared to those who moved due to reasons other than violence (10-29 percent for male, and 18-37 percent for female workers, depending on the matching method used). 7 In Georgia, compared to local residents IDPs are less likely to be in the labor force and more likely to be unemployed and receive lower wages, sometimes even after 20 years of forced displacement. 8

Worse labor market outcomes reflect specific obstacles refugees and IDPs face
Laws and regulations limit the right to work for a majority of refugees and some IDPs. The 1951 Refugee Convention grants refugees the right to work. Nevertheless, some countries completely legally ban refugees from working, be it as an employee or starting a business. Other countries restrict refugees' right to work, for example by limiting sectors and geographical areas in which refugees can work, or the number of refugees a firm can employ, or by requiring refugees to have a local partner when setting up a business. 9 Refugees also often must first apply for a work permit. Even if refugees are legally allowed to work, the processes to obtain work permits and register a business are in practice often confusing, cumbersome, and costly. As a result, many refugees are not formally allowed to work. As citizens, IDPs are in principle not confronted with these legal challenges, but restrictions exist. 10 Besides the right to work itself, a complex set of additional laws and regulations influence labor market access. How, when, and if refugees can access protection, and how temporary or permanent their refugee status, is a key determinant of their ability to obtain right to work and integrate into the labor market. 11 Refugees are often not allowed to move freely and choose where to settle in the host country, and placed in geographically remote areas with few economic opportunities. A study collecting data on IDPs in and outside of camps in Nigeria and Somalia found that IDP camps are also further away from markets and services. 12 In many countries, refugees also face restrictions on their ability to own property and access bank accounts or other financial services such as insurance and loans. These restrictions affect their ability to work or start their own business.
The extent to which their legal status is uncertain, or their rights are curtailed, also determines what kind of investments refugees will make.
Forced migrants also struggle due to lack of information about their rights and face discrimination. Employers are not always aware that they can hire refugees. When asked if refugees in Uganda were allowed to move freely or to work, only around 20 percent of employers surveyed gave the right answer. 13 In focus group discussions, Congolese refugees in Rwanda named employers' lack of knowledge of refugees' right to work as a major obstacle to their labor market integration. 14 Employers might also discriminate against refugees as well as IDPs for other reasons, including prejudices and stereotypes. Refugees and IDPs themselves might not always know their rights and be able to defend them.
Due to conflict and violence, or the displacement itself, forcibly displaced often lose assets, which may hamper their economic activities, especially if self-employment is an important source of income. A number of studies in low and middleincome countries-for example, in Colombia, Uganda, and Ethiopia-show negative association between forced displacement and level of household assets. 15 Refugees and IDPs have to leave assets behind, like land and livestock, and often have to deplete their savings to finance their journey and replace lost income. At destination, those forcibly displaced find it difficult to access or own land for farming.
The lack of assets can hamper self-employment and entrepreneurship. Liquidity constraints can also prevent those forcibly displaced from searching for (better) jobs.
Negative effects of physical asset loss are reinforced if households are separated from family members during displacement or face the death of a breadwinner. Representative refugee surveys in Ethiopia, Kenya and Uganda show that refugees were often separated from family members, had higher dependency ratios than hosts, and were more likely to live in female-headed households, reducing their ability to generate sufficient income. 16 Because of the loss of assets or inability of a male breadwinner to assure sufficient income, women might participate in the labor market for the first time after displacement. We see this in the case of Syrian refugees in Lebanon, for example. 17 Forced migrants often lack skills or certifications required in the new labor market. Compared to economic migrants, forced migrants usually have less choice over their destination, are less economically selected, and have less time to prepare for their move. Thus, they more often face a mismatch between the skills they have and those needed in the new labor market. One example is when agricultural workers are internally displaced to urban areas. Refugees and IDPs are also often forced to change occupation after displacement, as data from Iraq, Uganda and Cameroon show. 18 Obtaining new skills needed is likely to be easier for younger than for older refugees and IDPs. Acquisition of education and skills, as well as work experience, are likely to have been disrupted through conflict and forced displacement. Long periods of forced inactivity enact a toll on human capital formation. The formal or informal qualifications of those displaced are often not recognized at their destination, especially if they cross borders, which particularly affects those middle-and high-skilled. In general, the returns to education obtained in the host country might be different from those obtained abroad.
Forced migrants may also not speak the host community language, hindering employment or earning high wages. Research clearly finds a positive association between language proficiency and labor market outcomes in high-income countries. For middle-income countries, the importance of language is confirmed by a study finding that Syrian refugees with better Turkish language proficiency have a higher likelihood of being employed. 19 Refugees surveyed in Uganda identified language as the main obstacle to access employment (19 percent). 20 A lack of language skills may decrease the types of jobs the forcibly displaced can do, and may also represent a barrier to formation of social networks to aid in finding employment.
Refugees and IDPs are more likely than the host population to suffer poor mental health, which might prevent them from working. Refugees and IDPs often suffer traumatic experiences before, during, and after their journeys and suffer from the psychological effects, including depression, anxiety, and post-traumatic stress disorder (PTSD). 21 Stress and trauma negatively affect cognitive and socio-emotional functioning. 22 Mental health might add to physical health issues incurred during forced displacement, further hindering labor force participation. 23 Sometimes a vicious circle ensues, with mental health issues preventing refugees and IDPs from participating in the labor market, which exacerbates their mental health issues.
The experience of violence and forced displacement might also lead to a negative outlook on life and higher risk aversion, which can negatively affect economic activities undertaken.
Controlling for current wellbeing and loss of assets, a study on internally displaced households in Colombia finds that those who had been exposed to more severe violence have lower perceived prospects of upward mobility relative to the mean of sampled IDPs. 24 In addition, discrimination and stereotypes in the host country can affect expectations and lead to self-fulfilling prophecies. The experience of conflict can increase risk aversion. 25 Lower hopes and aspirations, as well as higher risk aversion, can have particularly negative effects on entrepreneurship, as entrepreneurs should ideally be self-confident, have higher risk appetite, and rebound quickly after failures. 26 The time horizon of forced migrants influences if they are willing to make host country-specific investments, and they might take asset portability into consideration. Compared to economic migrants, the time horizon of forced migrants can be: (a) longer if they do not think that they will return or be resettled elsewhere; (b) uncertain if they might be forced to, or are planning to, move again; or (c) shorter if they expect unforeseeable changes to occur to them again in the future.
Depending on the situation and their time horizon, they might be less willing or more willing to make language, skills, or hard asset investments in their new location. Qualitative evidence from Colombia, for example, shows that a group of internally displaced persons who received new plots of land did not invest in the land and only grew subsistence crops, out of fear of being displaced again. 27 In Uganda, refugees planning to return to their country of origin seem less interested in looking for local jobs. 28 The pay-off period for investments plays a role, and if permanent status is only granted after a long period of uncertainty, specific investments may no longer be optimal. 29 Forced migrants might also take asset portability into consideration. They might, for example, prefer investments in human capital to investments in physical capital, and make sure that their human capital is transferrable. 30

Forced migrants also lack social networks in host countries to help them overcome information asymmetries and integrate into the labor market.
While economic migrants tend to go to where they have social networks, the decision forced migrants make on where to go is determined largely by violence, with less forethought. In addition to language barriers, living in remote areas in camps or in group accommodations, combined with restrictions on freedom of movement, increases social isolation and makes it difficult to build networks with hosts. Contacts with certain groups of co-nationals or other forced migrants can be helpful for better labor market outcomes but relying too exclusively on these networks might hamper long-term integration. Forced migrants often face a lack of demand in the host labor market. Forced migrants usually do not choose their first destination based on available labor market opportunities, and tend to move together in large groups, suddenly increasing labor supply. This excess supply adds additional challenges to their labor market integration and might influence both short and long-term labor market outcomes. Together with other barriers to their labor market integration, this lack of demand might lead to higher rates of self-employment among those forcibly displaced in their host countries.
The influence of different obstacles depends on a number of factors, and job interventions need to consider each context to address job challenges forced migrants face. The obstacles for labor market integration vary between refugees and IDPs.
They depend on the context in the host community, notably where forced migrants are located (in camps in remote areas vs. urban areas), if they have a stable legal status, are formally allowed to work and economic opportunities exist. Obstacles might also change with length of displacement and vary within and between sub-groups of forced migrants in the same host country depending on age, gender, skilllevel, language, culture, the situation in the country of origin and the causes of the displacement.

WHAT ARE WE DOING?
In general, jobs interventions have become more common in the context of forced displacement. The number of jobs interventions, notably for refugees, has recently increased. This is a result of a global policy shift from humanitarian to development interventions in situations of forced displacement, and the long-term nature of many displacement situations. Previous interventions focused on skills trainings and small grants or credits. Since the start of the Syrian refugee crisis, public works or cash for work, value-chain interventions, and marketbased approaches have become more widely used. Many programs, such as productive economic inclusion approaches, now include several types of interventions. Interventions on the demand side of the labor market, and interventions aiming to increase social networks, are less common. The use of wage subsidies has been limited to high-income countries. UNHCR is a key partner in the preparation and implementation of WBG operations in the context of forced displacement.

Support for jobs is an important part of the WBG's engagement for refugees and host communities.
Nearly half of the 15 projects approved under the GCFF, and about one-third of the IDA-18 RSW projects, aim to create economic opportunities for refugees and host communities. They provide access to capital and training, sometimes complemented with additional interventions as part of a more comprehensive productive economic inclusion approach. Access to capital in the form of grants, matching grants or loans ranges from small amounts for the self-employed to larger amounts for business start-ups and existing firms. Other projects aim to build socio-economic infrastructure for refugees and hosts and include labor-intensive works, mostly as part of infrastructure building. Under both the GCFF and the RSW, the WBG has supported policy changes facilitating the economic inclusion of refugees as part of Development Policy Loans (DPLs), for example with Colombia and Jordan, as well as Programs for Results (PforR) instruments in Jordan and Ethiopia. 31 This includes support to provide refugees more stable legal status, reforming laws and regulations limiting refugee access to the labor market, and granting the right to move and settle freely. Support for policy changes will play an even larger role under the IDA-19 host communities and refugees window. In addition, the International Finance Cooperation (IFC) is engaged in efforts to promote investments by firms in refugee hosting areas.
The WBG also implemented nearly 80 projects targeting IDPs between 2000 and 2019, of which over one-third support jobs. 32 As the recently developed World Bank IDP strategy shows, these projects were implemented in over 30 countries. One example is the support for living standards and livelihoods of IDPs in Azerbaijan. No impact evaluation using a control group was conducted, but the results from pre-post comparisons seem promising, notably for the development of income-generating activities with self-help groups. The location of the microbusinesses in IDP settlements limited the profitability and growth potential of such projects, however. studies also assess the jobs potential for refugees in certain value chains or in private sector firms. To increase knowledge about the specific challenges refugees might face, the World Bank has also helped conduct representative cross-sectional surveys on refugees and IDPs, and has promoted the integrating of refugees and IDPs into national household surveys run by National Statistical Offices.

WHAT WORKS?
As rigorous evidence is nascent, this section looks at emerging lessons from interventions targeting formal and informal employment, self-employment and entrepreneurship, as well as hypotheses requiring further testing. Rigorous quasi-experimental or experimental evidence on jobs interventions for refugees and IDPs is scarce and mainly focused on high-income countries. Table 2 presents an overview of the state of evidence for each type of intervention. Where we do not have enough rigorous evidence to credibly demonstrate impacts, we present first emerging lessons, factors that need to be considered, as well as ideas, hypotheses, and interventions that might be useful to test in the future. Given the restrictive legal framework and general lack of employment opportunities in the formal sector in many host countries, we use a broad definition of jobs as any income source, formal or informal, including employment, self-employment and entrepreneurship.
Jobs interventions for those forcibly displaced must take the legal obstacles into account and, if possible, support reforms in this area to allow other interventions to be truly successful. Legal frameworks are especially important in determining relationships of those forcibly displaced with the labor market. As evidence from OECD countries shows, the length of waiting before legally allowed to enter a labor market is associated with employment gaps, which take many years to fill due to scaring effects, deterioration of human capital, and lower motivation. 33 Decreasing the period of uncertainty about the prospects of stay leads to better labor market outcomes, possibly by giving refugees a longer time horizon and incentivizing country specific investments. In addition, settling those forcibly displaced where labor market opportunities are, or allowing them to move and settle freely has a positive effect on their labor market outcomes. One example for reforms in this regard is the new refugee law in Ethiopia, which grants refugees the right to work and allows them to reside out of camps. How lowering the regulatory barriers for formal labor market participation of refugees or helping them overcome the administrative hurdles will impact their labor market outcomes will depend on the context. 34 In Jordan the introduction of work permits had limited uptake and impacts at the beginning, likely due to the limitation of work permits to a small number of sectors that did not match the skills and aspirations of the refugees and the fact that work permits were tied to employers. 35 Follow-up reforms in Jordan now aim to grant more flexible work permits, waive work permit fees and allow Syrian home-based businesses, which is particularly important for Syrian refugee women. Even if refugees mostly continue to work in the informal sector, because of their skill levels and the size of the informal sector in the respective country, they will do so on a more equal footing with natives in terms of bargaining power if they have the right to work. Anti-discrimination laws and communication about rights of those forcibly displaced aim to address discrimination against refugees and IDPs, but there is a lack of rigorous evaluations on the benefits of such policies and interventions.
As refugees and IDPs usually face multiple barriers to enter the labor market, the evidence for using integrated interventions that address several constraints at once seems more promising. Such interventions range from the combination of training with cash injections or work experience to full-fledged productive economic inclusion-type of approaches. As integrated interventions are usually more expensive, more evidence is needed to make the "black box" more transparent to understand which program elements are most effective and efficient. Compared to other target groups, certain types of interventions that consider the specific situations of refugees and IDPs might deliver larger benefits. Table 2 summarizes the different types of interventions, their rationale in the context of forced displacement, and what we know about how they affect refugees and IDPs.
Where combined interventions are not possible, financial capital injections seem particularly important as the loss of assets and income holds those forcibly displaced back. Cash and other capital-based support has an encouraging track record, even in situations of fragility, conflict, and violence (FCV). One way in which refugees and IDPs differ from other types of migrants or poverty-affected groups is their particular need to make up for the loss of assets and income associated with their [unplanned] migrations. Notably in low-income countries, but also in middle-income countries, self-employment is an important source of income usually requiring certain assets. Liquidity constraints can also prevent those forcibly displaced from finding (better) employment. 36 In response, financial capital injections are even more important compared to other contexts. Skills training should be linked to cash injections or access to credit to be more effective. Where relevant, this would need to include granting access to land for farmers, which, besides financial capital, requires the availability of land that can be purchased or used. We would, however, expect short time horizons or uncertainty about length of stay, as well as very restrictive legal frameworks for refugees, to limit the benefits of such cash injections.

Skills training alone does not have a promising track record, and training needs to adapt to the specific constraints of those forcibly displaced.
Because refugees and IDPs are usually not able to choose their locations based on host community labor market needs, there is a higher likelihood of skills mismatch. Training programs could help address the lack of skills those forcibly displaced might face in destination labor markets, but the track record of "skills only" interventions is not promising. 37 Training programs in forced displacement context also need to tailor trainings to skills that are appropriate and marketable in view of the legal framework and location of forced migrants, which often limit economic opportunities. Training programs should directly address the greater skills mismatch by focusing on people that need to change occupation after displacement. Qualitative and descriptive evaluations show positive results for IT skills such as coding, but the potential for scaling up programs teaching these kind of skills needs to be proven. 38 Besides taking general gender-relevant aspects into account-such as gender-barriers to access certain jobs, or social norms that make it difficult for women to leave the house-interventions need to adapt to a potentially increased number of female-headed households and higher dependency ratios. Trainings need to design implementation modalities to allow female participants to attend, for example including childcare or adapting the hours of the training. Language training can also be helpful to improve employment opportunities, but to be effective this training needs to adapt to different education levels and link to job opportunities. Assessment and recognition of skills and qualifications could be helpful, notably in more formalized labor markets and for those with middle or high-skill levels, but more evidence is needed.
Interventions addressing the psychological and mental health effects of forced displacement are very promising. Several rigorous impact evaluations have shown that short-term therapy interventions can improve the mental health of refugees and IDPs. 39 First evidence-for example from RCTs in Iraq, Uganda, and Thailand-also shows that therapy can be successfully implemented cost effectively even in low-resource settings lacking professional counselors and psychotherapists. 40 We need to better understand, however, which types of support can facilitate labor force participation, and how such support can be best integrated into broader jobs interventions. Jobs interventions in general need to understand changes in mental health and economic behavior due to forced displacement-which can affect risk appetite, aspirations, and willingness to invest, as discussedand incorporate them into intervention design. Job interventions also need to address the potential scaring effects of long periods of inactivity and consider that beneficiaries may no longer be used to work.
"Matching" interventions can be promising but need to be adapted to the specific labor market. Job search assistance or "matching" programs are associated with positive effects on refugees' employment in high-income countries, where job opportunities in the formal sector exist. 41 The little existing evidence from low and middle-income countries seems less promising. Beyond public matching services, refugees and IDPs need support to build up social networks. In addition, first evidence shows that intensive coaching and individualized assistance can have positive effects on labor market outcomes even in low and middle-income countries, but these types of assistance tend to be more costly. 42 Interventions to overcome spatial mismatches, such as transportation and housing subsidies, seem promising and need further testing and evaluation. 43 While not an effective instrument to create jobs, subsidized employment could help forced migrants increase their employability. Wage subsidies in the private sector show promising effects for refugees and other migrants, but evidence is limited to high-income countries and short-term effects. 44 Other monetary benefits for employers need to be adapted to firm profiles and forced migrants' skills and preferences. The 2016 EU-Jordan Compact on Syrian refugees represents prominent example, where only a few firms took advantage of the opportunity to have rules of origin relaxed when exporting to the EU if Syrian refugees made up a certain percentage of the workforce. Public works and cash-for-work programs have become popular in the context of forced displacement. They can confer important short-term benefits, providing temporary employment and increasing income and assets even in very fragile and poor environments, and where refugees face legal obstacles to integrate the labor market. Medium to longer-term effects on employment, however, seem less promising. 45 Value-chain and other market-based interventions seem promising, but rigorous evaluations are needed. In the context of forced displacement, humanitarian and development actors increasingly implement interventions aiming to develop or strengthen links along value chains and provide market information, often combined with other measures. 46 These interventions aim to increase demand for goods and services of the self-employed. As for public works programs, rigorous evaluations are lacking.

All interventions should include those forcibly displaced as well as their hosts on a needsbased approach and promote social cohesion.
A recent review of the literature shows that the majority of studies do not find significant impacts of a forced displacement inflow on host community labor market outcomes, but there is evidence that certain subgroups in the host community can be negatively impacted. 47 Jobs interventions should help to address such potential negative impacts for host community members and target those forcibly displaced as well as their hosts on a needs-based approach, to avoid increasing social tensions. Interventions should also align benefits for those forcibly displaced with those provided to the host community. Ideally, effective interventions and systems should be developed for the local population, that can then be expanded to refugees and IDPs. Promoting social cohesion will help those forcibly displaced build up social networks and decrease discrimination, which could in turn help them find jobs. 48

WHAT'S NEXT?
The design of jobs interventions must address the specific obstacles refugees and IDPs face in accessing labor markets. This Note drew on a nascent but growing body of evidence to highlight the specific obstacles refugees and IDPs face on the labor market compared to locals and economic migrants. Low and middle-income countries need more data-notably panel data-to better understand the challenges forcibly-displaced individuals face. Even more than in other contexts, thorough assessments of both labor market supply and demand need to inform program design. Assessment must include legal context and the perceptions and aspirations of the forcibly displaced. Programs such as skills trainings have often not assessed if and in which type of activities and sectors those forcibly displaced are allowed to work nor market demand where refugees and IDPs were located, which has limited the benefits from such programs. Jobs interventions also usually do not seem to consider how forced displacement experience impacts economic behavior and investment calculations of their target group.
More rigorous evidence is needed on how to design jobs interventions for the forcibly displaced in low and middle-income countries, with more focus on IDPs. For many types of interventions, rigorous evidence only comes from high-income countries. The vast majority of evaluations focus on refugees. Future evaluations should include more low and middle-income countries and interventions working with IDPs, but also should look at longer-term impacts. Evaluations should focus on interventions, but where rigorous evaluations are missing, such as value-chain and other market-based and economic-inclusion approaches. Future research can exploit recent changes in legal frameworks to improve refugee labor market access in countries such as Jordan, Colombia, and Ethiopia to broaden our knowledge of policy impacts in low and middle-income settings. Evaluating the impacts of interventions in the forced displacement context should not be limited to basic needs outcomes, such as food security: evaluations should also consider employment and jobs outcomes, which may indicate more substantive integration into a host community with longer-term sustainable benefits. Moreover, these interventions should also consider the quality of work, including wages and exposure to hazardous conditions. This is because programs such as cash transfers, which show little to no benefits on employment rates, may have hidden positive effects in increasing quality of work conditions. Self-reliance indices might help measure intervention effects over time. 49 More experimentation is needed to evaluate promising interventions that have not yet been widely implemented. This includes interventions that aim to: • Enact policy reforms addressing legal challenges refugees face. Evidence from OECD countries shows that such policy reforms could also confer large benefits in low and middle-income countries. As the experience from the Jordan Compact shows, experimentation is needed to understand how to best design these policies in specific labor market and political contexts, and to address specific refugee population profiles and aspirations.
• Expand social networks of the forcibly displaced to help them find jobs. Notably in low and middleincome countries, public matching services cannot replace private networks and refugees and IDPs might benefit from support to build up such networks.
• Cover the costs of job search and overcome spatial mismatches. Even when allowed to move freely, forced migrants might need support to actually do so. Programs supporting transportation or providing housing allowances for urban centers can allow forced migrants to move to access economic opportunities. One first adaptive targeted field experiment found positive impacts of a cash-transfer labeled to pay for costs of job search among Syrian refugees in Jordan. 50 Programs aiming to overcome spatial mismatches in terms of employment opportunities have proven quite successful in other contexts. 51