90156 July 2014 – Number 129 X1` Climate Change and Migration in the MENA Region Quentin Wodon and Andrea Liverani1 Introduction: Climate change and climate-induced migration (Foresight, 2011) are major global concerns. This is true for the MENA region as well. Yet empirical data on how perceptions of climate change and weather shocks affect migration in the region are scarce. To what extent are perceived and actual weather shocks and changes in the environment driving temporary and permanent migration flows? Do remittances reach households living in climate poor areas, and if so, what is their Egyptian Farmers – World Bank Photo Collection impact on poverty and human development? These are some of the questions considered in a study by In practice, the implication for this note and the Wodon et al. (2014) based on various data sources study it is based on, is that the results do not including new household surveys for climate provide new evidence on the direct relationship affected areas in Algeria, Egypt, Morocco, Syria, and between climate change and migration per se, but Yemen (the “five country sample” in this note). they contribute new evidence on three related issues: 1) the impact of weather shocks on migration; In a short summary note as this one, it is important 2) the impact of perceptions of recent climate change to be clear at the outset about what is measured and on migration; and 3) the impact of climate variables what is not. It is sometimes said that “Climate is what such as temperature and rainfall (but not directly we expect. Weather is what we get.” Simply put, climate change) on migration. climate relates to the distribution of variables such as temperature and rainfall over a long period of Survey and Census Data: Migration is often a time. This distribution is characterized by its strategy to cope with and adapt to changes in moments, including the mean and the variance of climatic and environmental conditions. In the key climatic variables. Climate change is then used surveys from the five country sample used for the to refer to the change in the distribution of rainfall study, the data suggest that 29.9% of households and temperature. However, it is difficult to tell if the have migrants, whether residents (current weather experienced at a point in time is due to household members who migrate temporarily) or climate change (the overall mean and variance of non-residents (former household members who rainfall and temperature) or part of an existing have migrated permanently). Rates are lower if distribution. migration is restricted to the last five years. At the individual level, 7.6% of individuals have migrated 1 temporarily; the proportion over the last five years is Quentin Wodon, Adviser in the World Bank Group’s Education 6.2%. For permanent migration, the rates are 8.0% in Global Practice (GEDDR). Andrea Liverani is a Senior Social Development Specialist in the Urban, Rural and Social the whole sample, and 5.7% in the last five years. Development Global Practice (GURDR). This MENA K&L Quick Most migration is internal, but the likelihood of Note was cleared by Franck Bouquet, Senior Adviser, MNAVP migration abroad is high in Egypt and higher for and Charles Cormier, Practice Manager in the Energy and individuals from higher welfare quintiles (as Extractives Global Practice of the World Bank (GEEDR). expected given the cost of international migration). and variation) affect migration from some districts For residents and non-residents, migration to urban to others, but in a limited way - socio-economic and areas, especially to large cities, is much more likely cost factors are more prominent. The data suggest than to rural areas. that migration flows are unlikely to increase sharply in the near term, but if conditions were to change To what extent are households migrating away from drastically, the effect of weather could be larger. climate affected areas, and is climate a strong push factor? Regression results for the five country Qualitative Data: Focus groups data also suggest a sample suggest that poor climate and extreme link between climate and migration, although again weather events lead to a higher migration the role of socio-economic factors is probably at least probability, but the role of climate is less important as important as that of climatic conditions in driving than that of socio-economics and urban job migration decisions. Respondents linked migration prospects. The data were used to build two indices to chronic droughts which lead to declining of household perceptions on changes in weather agricultural productivity. But few respondents patterns and the environment. The first captures the mentioned flooding or being displaced as a result of extent to which the climate is becoming dryer and natural disasters as the main reason for relocation or warmer, and is associated with droughts. The migration. Respondents appear to choose migration second captures the extent to which households after other strategies have proven unsuccessful. The suffer from excess water and is associated with qualitative work suggests that remittances are key floods. In regression analysis, higher values for both for livelihood, especially in Yemen where there is a indices result in higher rates of temporary tradition of migration to Saudi Arabia. migration, with statistically significant coefficients. The effects for permanent migration are positive but The qualitative work suggests that migrants are not statistically significant for the whole period only as always successful, which may explain why climate opposed to the last five years. So higher values for related factors do not seem to have a very large both indices lead to higher rates of migration, but effect on permanent migration. What happens to the the evidence is weaker for permanent migration. migrants once they leave? Many migrants hope to work and save enough to own their own business (a The estimates suggest that a steep worsening of the small store or shop). Finding jobs, however, is not an climate could increase migration rates by 1.5 easy feat amid widespread joblessness found percentage point. This represents up to a fifth of throughout the region given that unemployment current migration patterns (but again, only the levels hover around 15 %, and a much higher 20 % results for temporary migration are statistically to 40 % among youth 15 to 30 years old. While some significant for the last five years). This is of an order migrants find urban communities receptive, others of magnitude similar to that obtained when say they were met with hostility. Finding adequate considering reasons declared in the surveys by housing was also a challenge. households for the migration of their members. Summary Results: Table 1 summarizes the main Additional evidence for Morocco using a 2009-2010 results. The evidence suggests that worsening national survey suggests that weather shocks climatic conditions, or perceptions thereof, are a increase the likelihood of temporary migration by push factor for migration away from affected slightly more than one percentage point, an order of (mostly rural) areas towards urban centers. It must magnitude similar to that for the five country be noted that in an analysis such as that of census sample. In the Morocco survey, the impact of data which is at the level of a country as a whole, the weather shocks on permanent migration is again not impact of the climate on the overall patterns of statistically significant, but the impact of changing migration tends to be diluted. By contrast, in the structural conditions, i.e. reduced agricultural yields analysis based on climate affected areas in the five due to lack of water is, and it contributes to countries sample, as well as when looking at those permanent migration away from affected areas. directly affected by weather shocks with the national Morocco survey, the effects are estimated on those Combining Yemen census and weather station data, most affected, which also explains why impacts are the study also finds that climate variables (in means larger. In the most affected areas, it seems fair to July 2014 · Number 129 2 suggest that climate conditions account for at least on school enrollment. Thus, in areas with 10 % to 20 % of current migration flows, and this unfavorable climate, remittances seem to help for could increase in the future. But the evidence is meeting basic needs and escape poverty or stronger for temporary than for permanent malnutrition, while in areas with more favorable migration by individual household members. climate, remittances may be used more for investments, including in the education of children. Table 1: Summary Results Variables Country Migration type Size Conclusion: The study suggests that the impact on Temp. Perm. migration of weather shocks and perceptions of a Climate perceptions 5 countries + Weak Medium worsening climate is positive. Yet, the effects are not Weather shocks Morocco + Weak Medium Climate variables Yemen N.A. + Small very large and they are stronger for temporary than Focus groups 5 countries + + Medium permanent migration. In areas affected by climate Source: Adapted from Wodon et al. (2014). change, the analysis suggests that climate factors may account for between one tenth and one fifth of Role of Remittances: The study also looked at today’s level of migration. We would expect this to remittances, focusing on Yemen. The literature increase as the climate deteriorates. While many suggests that remittances increase in response to migrants appreciate the opportunities offered by climate shocks. Less clear is whether remittances are migration, living conditions and the ability to higher in areas suffering from poor climate in the integrate well in their area of destination is far from absence of weather shocks. The study used a Yemen guaranteed, given intense competition for few good national household survey and weather data to jobs. In terms of implications, five areas for policy assess the likelihood of receiving remittances by actions are outlined below (findings of the study on area. The results suggest that households living in coping and adaptation and their implications are less favorable climates (as measured by higher discussed in a separate note and not covered here). temperatures, lower rainfalls, more variability or seasonality in both, and larger differences in a given First, the study shows that while environmental and year between extreme temperatures) tend to be less climatic factors play a role in migration, other socio- likely to benefit from remittances in Yemen. economic factors are more important. Similarly, citing environmental degradation as key to trans- Finally the study used matching econometric border migration is likely misleading: in the study, techniques with the Yemen household survey and when environmental factors are at play, migration is weather data to measure the impact of remittances, mostly internal. These findings run against the both domestic and international, on poverty and received wisdom around climate-induced migration, human development outcomes (school enrolment, but are in line with results from other assessments, immunization, and malnutrition). Four main results such as the recently released Foresight (2011) report are obtained: (i) remittances – substantial in Yemen – on environmental change and migration. Identifying tend to have positive impacts on poverty, school climate migration as a domestic policy issue will enrollment, and measures of malnutrition; (ii) for lead to a different perspective on the problem for beneficiary households the impact of international MENA policy makers and donors alike. remittances tends to be larger than that of domestic remittances, probably because the amount of Second, migration can be a form of adaptation, but remittances received tends to be higher for it is often seen as a last resort by households, and international than domestic remittances; (iii) the especially so in the qualitative work. One reason is impact of remittances, especially international that migration may be perceived as costlier than remittances, on poverty and malnutrition tends to be other strategies, such as using savings, selling assets, larger in areas affected by high temperatures, and debt, or removing children from school. Material also to some extent in areas with lower levels of costs (travelling and housing) aside, migration has rainfall, which in both cases tend to be more high risks due to uncertain outcomes and vulnerable; (iv) by contrast, in areas with higher immaterial costs, including the uprooting of levels of rainfall or lower levels of temperatures, individuals, households, and even communities. where issues of poverty and malnutrition may be Also, those left behind at the household or less severe, remittances tend to have a larger impact community level may be deprived of migration’s July 2014 · Number 129 3 benefits when remittances are hampered by the cost reached 60% in 2000 and will reach 70% by 2015. of remitting or if migrants cannot find jobs. How MENA policy makers address the issue of climate induced migration is linked to how well Policy responses and development engagements they will promote an urbanization open to migrant’s must recognize that migration is a viable and contributions. legitimate option to address risks to livelihoods and wider well-being. It is also a means of adapting to Finally, the provision of education and training can climate change and its impacts. A key question for help potential migrants better utilize labor market migration policy is whether/when migration should opportunities, adapt to new conditions, and shift be treated as a risk to be managed and mitigated, more easily between jobs. An emphasis on basic and and whether/when it should be treated as an portable skills and human capital would be effective opportunity to be facilitated, even encouraged. regardless of the causes, timing, and destination of Enabling sending area communities to better the observed patterns of migration. It would benefit leverage the benefits of migration and increase their not only those leaving, but also those staying or adaptive capacity is a better alternative than eventually returning. progressive displacement. The economic insertion of migrants in cities leads to opportunities for the Ultimately, dealing with climate induced migration sending communities, mostly due to remittances. will require interventions specifically aimed at Evidence from the surveys and qualitative work migrants, yet the policy package needed to deal with suggests a positive impact for remittances in areas both climate change and migration is broader and affected by climate shocks, especially on human presents both a challenge and an opportunity. An development outcomes. But without an enabling opportunity because several levers can be used to environment, remittances too often turn to pure better leverage migration’s potential for adaptation consumption and the accumulation of non- and development; but also a challenge since an productive assets such as real estate. These can be of integrated policy response requires levels of little use for preventive and ex-post adaptation, coordination and commitment likely to arise only since investment contributions are limited and their through broader governance reforms and strong liquidity limited for use when climate impacts political leadership in MENA countries. strike, at which point their value can drop. Policy should focus on upping the impact of remittances by References encouraging productive use, for instance by Foresight, 2011. Migration and Global Environmental subsidizing de-fiscalization for remittances-funded Change, London: The Government Office for Science. investments and community saving schemes which facilitate financial integration and increase liquidity. Wodon, Q., A. Liverani, G. Joseph, and N. Bougnoux, editors, 2014, Climate Change and Third, urban development policy is a key Migration: Evidence from the Middle East and North component of policies to address climate induced Africa Region , World Bank Study, Washington, DC: migration. The study summarized ion this note The World Bank. focuses on sending areas, but the qualitative work in urban areas suggests that the migrants’ integration into major destination cities is not going very well. Climate induced migration tends to be to mostly large cities. The policy responses to climate shocks and migration are then to be found in cities as much as in sending areas. Migrants are concerned about employment and housing, with migration adding to existing pressures that can only be dealt with through broad-based economic development, not necessarily focused on migration. Climate induced migration should be part of a broader policy debate on urbanization. The MENA region is rapidly urbanizing. From a 48% urbanization rate in 1980, it July 2014 · Number 129 4