Multiplier Effects of Humanitarian Assistance The Local Economy-Wide Impact Evaluation (LEWIE) Method and Its Findings among Refugees in Uganda and Rwanda J. Edward Taylor, Mohammad Alloush, Anubhab Gupta, Ruben Irvin Rojas Valdes, Heng Zhu University of California, Davis

Mateusz Filipski International Food Policy Research Institute (IFPRI)

Ernesto Gonzalez-Estrada and Jaakko Valli UN World Food Programme (WFP) *Funded by United Nations World Food Program (WFP) and UC Davis Migration Research Cluster

Three Research Questions WFP-UC Davis Project “Assessing the Impact of Market-based Interventions on the Local Economy: Case Studies of C&V Programmes in Refugee Camps”

• What do refugee camp economies look like? • How do refugees in these camps affect nearby host-country economies?

• Do the impacts change when refugee assistance is in cash and/or land to farm?

Three Papers • Economic Life in Refugee Camps (Mohamad Alloush, J. Edward Taylor, Anubhab Gupta, Irvin Ruben Rojas Valdes, and Ernesto Gonzalez-Estrada). World Development (2017, in press). • Economic Impact of Refugees (J. Edward Taylor, Mateusz Filipski, Mohamad Alloush, Anubhab Gupta, Ruben Irvin Rojas Valdes, and Ernesto Gonzalez-Estrada). Proceedings of the National Academy of Sciences (2016). • Economic Impact of Refugee Settlements in Uganda (J. Edward Taylor, Heng Zhu, Anubhab Gupta, Mateusz Filipski, Jaakko Valli, and Ernesto Gonzalez-Estrada). Working Paper, University of California at Davis (2017).

Global Refugee Crisis • More than 65.3 million people are displaced by conflicts all over the world and 21.3 million are refugees • Most are in low-income countries neighboring conflict zones • East and Central Africa have more than one-fourth of the world’s refugees Source: UNHCR (2015) http://www.unhcr.org/2014trends/

Sources: UNHCR (2016)

Refugees in Rwanda • Site of mass genocide and displacement in 1994 • Remarkable turnaround • Today Rwanda hosts about 148 thousand refugees • Mostly displaced from DRC and Burundi

• Refugee population increased more than 100% in last 3 years

A refugee mother with her child in Nyabiheke Refugee Camp, Rwanda (2015)

2 Aid Regimes • Camps are run by UNHCR, fed by WFP • Two aid regimes  In-kind aid

• Refugees receive a monthly food packet that cover 100 per cent of the daily energy requirement • Maize grain, pulses, fortified oil, and iodized salt  Cash aid • Cellphones; transfer the equivalent amount to purchase the food bundle • Refugees can withdraw all the funds, do purchases, and save

Camp Selection: Different Contexts, Aid Regimes Nyabiheke Camp switched to cash 2 months before the survey Gihembe Camp switched to cash 18 months before the survey Kigeme Camp still gave food aid in kind at time of study Local economy: defined by a 10km circle around each camp We also estimate impacts on trade outside these circles

Uganda Study: 2 Settlements, 3 Aid Regimes  As of March 17 Uganda hosted more than 800 thousand South Sudanese refugees (only 22 thousand before 2013)  Over 500 thousand new refugees from South Sudan in last 7 months  Other refugees from DRC, Somalia, Burundi & Rwanda  Rwamwanja settlement in South-West Uganda hosts mainly Democratic Republic of Congo refugees  Generous land redistribution policy  Adjumani settlement in Northern Uganda hosts mainly South Sudanese refugees  Limited land availability, unfavorable growing conditions  Geographically disperse, separated into 15 food distribution points

A weekly market in Rwamwanja Refugee Settlement, Uganda (2016)

Data for Our Study (Rwanda) • Detailed economic surveys in & around 3 refugee camps in June and July 2015 • Random sample of refugee households and businesses (not likely to be picked up by the household sample) in each camp • Host-country households and businesses within a 10 km radius of each camp (the “donut”). • Benchmark and input to refugee-impact simulations

• Location of all transactions inside and outside of this 10 km radius

• Sample sizes: • 546 Refugee households • 585 Host-country households • 459 Businesses (171 through HH surveys)

• Roughly equal across camps (slight oversampling of in-kind camp as baseline for possible follow-up study)

Data for Our Study (Uganda) • Adjumani and Rwamwanja Settlements • • • •

Two settlements, vastly different Random sample of refugee households receiving cash, food, and land Random sample of refugee businesses Random sample of host-country households and businesses in 15Km donut

• Sample Distribution (HH’s 1503, Businesses 385) Settlement

Inside Refugee Settlement

Host-Community

Cash HH

Food HH

Phased out HH

Business Households

Business

Rwamwanja

199

147

0

59

266

124

Adjumani

160

325

55

108

351

100

Total

359

472

55

161

617

224

Economic Life in Refugee Camps: Rwanda

Refugee Characteristics Differ Refugee Households are Larger, Healthier, Less Educated (but Have Higher School Enrollment Rates) than Host Households

Host-country Contexts Differ More Agricultural around Nyabheke Camp

• Host-country households engage in agriculture around all three camps • Participation in agriculture and livestock is highest around Nyabiheke camp • Agriculture produces food and employs refugees

Refugees Interact with Host-country Economies in a Variety of Ways • As workers • As consumers • As businesses • Interactions shaped by host-country economy • Are there jobs for refugees? • Do refugees have money to spend? • Is there local production to satisfy refugees’ food demands? • Are there markets to supply food and other goods?

• Wage employment depends on refugees’ human capital (skills, education)

34 - 47% Have Wage Income, Mostly Near or in the Camp

• Highest in (agricultural) Nybiheke, Lowest in (commercial) Gihembe • Low human capital may limit nonfarm employment

• Sectors of Employment Reflect Local Economies • • •

Agriculture in Nyabiheke Non-agricultural in Gihembe NGOs a big employer of refugees (but not host-country workers)

8 – 13% Have a Business in the Camp

Highest in (agricultural) Nyabiheke Lowest in (commercial) Gihembe • Competition from host-country businesses near the camp • High transaction costs w/ host country markets favor refugee business formation inside camps • Host-country businesses may not set up in camps

Businesses in Nyabiheke Refugee Camp

8 – 20% Get Cash Remittances

• Incidence of remittances is generally higher in refugee than host-country households • Highest in the old commercial camp (Gihembe) • Lowest in the agricultural camp (Nyabiheke)

Refugees at Kigeme Camp Use Local Markets to Turn Food Aid into Cash Sales of food aid in local economy, Kigeme Camp 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0%

% Selling in Local Economy % Sold

Maize

Beans

Oil

Salt

Figure 3. Nearly all Kigeme refugee households sell some food in the local economy surrounding the camp. Nearly one quarter of all maize they receive from WFP is sold outside the camp.

…But at a Big Loss

• The cash that Kigeme refugees could get by selling all of their WFP packet is only 64 percent of the packet’s value

Cash and Agriculture Make Refugee Diets More Varied (and Similar to Nearby Host Households) % Consuming

Households in Cash Camps are Measurably Happier & More Food Secure (Including Relative to Nearby Host Households*)

* For comparison with nearby host-country households see Alloush et al. (2017), “Economic Life in Refugee Camps," World Development.

Uganda Gives Refugees Access to Land Allocated land sizes for refugees remains small - Rwamwanja: 3600 m2 on average, most used for cultivation - Adjumani: