Market Assessment Report: Burundi Date: April 2016 Sector(s): Contact(s):

Economic Recovery and Development – Humanitarian Response Jackie MacLeod, Emergency Livelihoods Coordinator, IRC [email protected] Claudine Inamahoro, Technical Advisor – Markets, CRS [email protected]

Data Collection:

March 17 – March 25, 2016

INTRODUCTION AND JUSTIFICATION As the political crisis in Burundi reaches the end of its first year, there are significant signs that the humanitarian situation is worsening. Even prior to the political tensions, Burundi was considered the hungriest country on earth. According to the May 2016 situation update of the FAO-led Food Security and Nutrition Working Group, an estimated 3.6 million people are food insecure - of whom nearly 700,000 are severely so representing a dramatic increase over the past six months 1. Concurrently, with the suspension of a large percentage of international aid and the ongoing economic crisis, the national economy has slowed, with a -7.2% GDP growth rate at the end of 2015.2 Local market prices of basic food items have risen dramatically3 with an increase between 30-100% for rice and beans (depending on market location). In addition, the EU’s decision to suspend development aid, which accounts for approximately 50% of the national budget, has strongly impacted an already fragile economic system and will further weaken the economy4. While the economic situation continues to decline, El Nino impacts are worsening. The last Inter-Agency monitoring report for Burundi reported 5,068 houses damaged/destroyed by heavy rains and floods affecting 30,408 people.5 In addition, other factors like a risk of poor performance of the 2016 farming “Season A”, return of displaced people and refugees and continued price increases contribute to grim predictions for the Burundian population6. 1 http://www.fao.org/disasterriskreduction/east-central-africa/fsnwg/documents/detail/en/c/4368/ 2 https://www.imf.org 3 http://www.irinnews.org/analysis/2016/02/12/how-burundi%E2%80%99s-political-crisis-has-crippled-its-economy 4 http://www.english.rfi.fr/africa/20151103-business-usual-burundi-until-money-runs-out 5 http://reliefweb.int/report/burundi/burundi-inter-agency-monitoring-report-29-january-2016 6 Analyse de la Securite Alimentaire d’Urgence dans 6 provinces du Burundi les plus affectees par l’instabilite socio-politique. Novembre 2015. [Available online] at http://documents.wfp.org/stellent/groups/public/documents/ena/wfp281049.pdf?iframe

From Harm To Home | Rescue.org

IRC | CRS | Market Assessment

2

STATEMENT OF INTENT Objective(s) The IRC and CRS proposed to conduct an assessment that would gather the information on market functionality as well as the needs of the affected population; and build an evidence base to feed into the programming strategy for the Rapid Response Mechanism (RRM) and provision of assistance to the crisis affected, vulnerable households in 10 provinces of Burundi7 through a market based intervention. Households were considered crisis affected if they have been displaced, have suffered exceptional economic hardship in the last 12 months and/or have suffered a loss of livelihoods or assets. The assessment had the following objectives:  To assess the feasibility of a cash based/market based approach to aid delivery; that can address the needs for and access to material assistance (to include, but not limited to, non-food items, hygiene materials and food)  To assess whether the needed items are available and accessible in the local markets.  To gain an understanding of what the needs of the affected population are and to what extent different members of the affected population require assistance to meet those needs.  Identify opportunities to implement cash based/market based programs with dual benefits to both host communities and IDP population. Core Questions To design and guide the assessment, we developed the data collection tools and methodology around the following 5 core questions: 1. What has been the impact of events of the last year on the functionality of the markets? 2. Are food insecure households well connected to local markets? 3. Do supply or demand, or both affect the market performance? 4. Do local traders behave competitively? 5. What are the best entry points for humanitarian assistance? METHODOLOGY Tools and Sampling The methodology for the assessment was focused in three parts, in order to assess the needs of the affected population, the functionality of the markets in the potential areas of intervention as well as the feasibility to respond to the needs of the affected population through a market-based approach.

7

Makamba, Rururi, Rumonge, Rutana, Ngozi, Kirundo, Muyinga, Karuzi, Bujumbura Mairie and Bujumbura Rural

IRC | CRS | Market Assessment

3

1. A focused market assessment examining the quality of the market and its capacity to provide goods (specifically focused on a few staple foods that were identified as important by surveyed vendors) and services (transportation) to the local and displaced population(s). RAM principles served as guidance, but given the assessment’s focus, and time limitations, a full RAM was neither necessary nor feasible. Market system mapping for pre and post crisis was not done and some tools for conducting interviews with market actors were combined for efficiency. a. Market level surveys were conducted with vendors, wholesalers who are principally selling food items8 and market representatives in at 20 data collection points in 10 provinces, (2 largest markets per province), during market days. A total of 104 market actors who are selling maize, rice, cassava, and beans (87 vendors, 15 wholesalers and 2 Market Representatives) were interviewed. A convenience sample was chosen as the sampling method due to time and human resources limitations as well as security restrictions present during the time of the assessment. b. Key informant interviews were conducted with the following: i. Local government authorities ii. Merchants and relevant business operators 2. A needs assessment to determine the sources of economic burden of the affected population and current levels of access to markets. This was accomplished by (a) household-level surveys and focus group discussions (FGDs) with members of the affected population and (b) key informant interviews with local government, local leaders and others as deemed relevant. a. Household level surveys were conducted with members of the affected populations over a 7day period. Households were surveyed at 20 data collection points in 10 provinces. A total of 76 household interviews were successfully conducted, representing a total of 507 individuals. A convenience sample was chosen as the sampling method due to time and human resources limitations as well as security restrictions present during the time of the assessment. b. 40 FGDs were conducted with in 20 villages neighboring the surveyed markets: 2 FGDs in each village, for men and women separately. 3. A cash feasibility snapshot study that evaluated the availability of and options for appropriate response mechanisms the social and political acceptability of cash, the potential risks of a CTP program, and the potential negative externalities of distributing cash assistance.

Most market actors sell a variety of goods in their shops, for example someone selling maize and rice will also carry basic home items such as soap and toilet paper. 8

4

IRC | CRS | Market Assessment

Limitations The sampling approach is limited in its ability to be representative, given the chosen method of convenience sampling. However, the results are not meant to represent a statistically generalizable sample of the population but are rather to gather information on the general functionality, integration and capacity for the markets to respond to cash or other types of market based programming. Ethical Considerations All surveys, key informants interviews and FGDs began with an informed consent and interviewees had the opportunity to opt out of the conversation. KEY FINDINGS How markets have been impacted: Demand The largest market disruptions over the last 12 months noted by market actors are a lack of demand and lack of capital. 72 of the 104 market actors interviewed noted that in the last year, traders have left their market. The most common number of traders who left was approximately 20, and the most commonly cited reason for businesses not being able to continue as normal (both wholesalers and retailers) was lack of capital. It was later noted that in the last year, in addition to some traders leaving, there are also a number of traders who have entered the markets, however this question was not asked during the quantitative survey. Table 1: Q: What factors prevent businesses from continuing as normal (last 12 months)? 120 100 80 60 40 20 0 lack of capital

increased taxes lack of demand Lack of security

corruption

other

Factors preventing businesses from continuing business as normal wholesalers

retailers

*Other reasons given include non-repayment of debts, lack of available credit, theft, transportation problems and physical relocation of the marketplace.

5

IRC | CRS | Market Assessment

93 of 104 market actors interviewed noted a decrease in customers in the last year. When asked to select the main reasons for the decrease of customers9, by far the most commonly cited reason was a lack of purchasing power from the population (73 responses). The other reasons given were an outward movement of population to other locations (26 responses) and clients using other markets (15 responses). Outward movement of population and relocation of physical markets were mentioned in Bujumbura Mairie and Bujumbura Rural, mainly caused by lack of security in the capital. Supply Over the last 12 months, the impacts on the markets and disruptions to business have been, by and large, monetary. Though, it should be noted that due to flooding in some locations (Rumonge and Bujumbura Rural) as well as insecurity in urban areas notably, Bujumbura Mairie, there have been physical impacts on the market as well. This includes degradation of roads due to flash flooding and in some cases the market location has moved entirely. Some market actors noted that this has caused a decrease in their clientele because people have difficulty finding the new market location. Table 2: Q: What have been the most noted disruptions to the market in the last year? 100 90 80 70 60 50 40 30 20 10 0 reduced demand

price rises

reduced supply transportation problems

security problems

infrastructure destruction

Other disruptions

*Other disruptions include change of market location, increase in theft, flooding, displacement of population and increase in cost of transportation.

Reduced supply for the main food commodities (rice and beans) is mostly seasonal, March-April and September-October being the most deficient periods of the year. However, households reported running out of food stocks earlier than usual due to lost harvests caused by the heavy rains that fell out of season. The Answers were not mutually exclusive, therefore it was possible to have one respondent choose a number of reasons for the decrease in cliental if needed. 9

6

IRC | CRS | Market Assessment

normal alternative solution during the lean period is to feed on tubers (cassava, sweet potatoes) for most households. However, it was reported during focus group discussions that the supply of substitutes is also insufficient in the markets due to heavy rains, which caused crop failure in the last year. Wholesalers and retailers reported to resorting to imports (mainly from Tanzania) or regional suppliers instead of local suppliers during the lean periods. Secondary data reports regional and trans-border trade flows, confirming that the markets are integrated 10. However transport and customs duties push up prices for consumers when supply sources are far from local markets. In addition to the physical damage of roads due to flooding, the transportation problems noted were mostly regarding the higher prices of transportation over the last year. In general prices for all goods and services have increased dramatically in the last 12 months: data show an average increase of 38% for food items and 24% for NFI11. When asked whether the vendors needed to change where they source their goods over the last year, very few needed to change their source dramatically. There were some vendors who noted changing from one province or even commune to another when they needed to obtain a better price to maintain their business, but only 25% of those interviewed needed to make a change in where they sourced their goods (e.g. local vs regional vs international). It is noteworthy to mention that most traders do not source their goods via Bujumbura, so even though some supply routes have been disrupted due to road degradation they have been able to change the source of their goods with little effect on the price. When asked if they would be able to meet an increased demand for their key commodities, overwhelmingly the answer from vendors was yes. For example, 59 vendors reported that one of their key commodities is beans, and 46 of those vendors reported that they would be able to meet an increase in demand if needed. The vendors believe they could increase their stock from existing sources if their access to capital increased, which they believe would increase with the increase of demand. Table 3: Q: Retailers, what are your top two key commodities? Beans 59

Rice 46

Cassava (Manioc) 41

Maize 22

Table 4: Q: If the demand for those products increases, would you be able to meet the demand? Beans 46

Rice 35

Cassava (Manioc) 21

Maize 21

*Table captures the number of retailers who stated that they would be able to meet increased demand.

10 11

WFP Emergency Food Security Assessment, Nov 2015 Prices for 2015 and 2016 were collected for main food and non-food items in all surveyed markets for the month of February.

7

IRC | CRS | Market Assessment

In general, the market performance was affected by supply and demand factors, which, at the end pushed up prices in the face of unchanging or reduced income sources for consumers. Other market factors that influence the functionality of markets like competition, integration and basic market amenities were not impacted by the crisis. Access to local markets Data shows that households have easy access to markets, within walking distance and with no specific access restrictions for any group of people in the communities, apart from bad but functional roads. Temporarily, one local market visited in Bujumbura Rural (Gitaza Commune) is not accessible by road, due to recent landslides caused by heavy rains. A few markets changed their location, specifically in Bujumbura Mairie, due to instability. However, no major disruptions or damage were caused to the market infrastructure, and all markets observed during the assessment are fully functional. Prices of transport on major roads increased due to macroeconomic factors like inflation and national currency devaluation since last year. As a consequence, the prices of goods have increased, which in turn high contributes to the erosion of the purchasing power of consumers in the whole country. Population members with no steady income, nor productive assets like land or livestock, are hard hit. Table 5: Reasons for lower than normal household food consumption 25 20 15 10 5 0 Lack of money

Too expensive manioc

not available in market maize

beans

problems accessing market

other

rice

*Other reasons include switching to less preferred commodities and reducing number of meals eaten per day

When asked about their ideal versus actual consumption of the following staple commodities: beans, rice, and cassava 47% of households reported consuming less than their ideal amount of at least one of the commodities and 30% were consuming less than their ideal amount of two or more of the commodities. The most commonly cited reasons for lower than normal consumption were high prices and lack of money at the household level. Only one respondent indicated that physical access to the market was a constraint to purchasing the needed or desired commodities.

IRC | CRS | Market Assessment

8

Households reported less income in the last 12 months, combined with the increasing prices of food and nonfood commodities, all contributing to a stressed economic situation for already vulnerable households as well as reduced purchasing power for many households who were previously food secure. Food commodities are readily available in the markets, however the prices have increased significantly over the last year so for poor households, financial access is reduced or nearly eliminated. 87% households interviewed reported that their monthly expenses exceeded their monthly income. Only 10 households reported that they were able to save any of their harvest from the last year, with a mode of 25 kg of beans saved. Security As previously noted, instability was most notably mentioned in Bujumbura Mairie and Bujumbura Rural in terms of its impact on the market. Additionally, informal interviews reveal that the labor market (mainly the tourism industry) in the capital city shrunk, causing major lay-offs and as a consequence, reduced remittance sources from city workers to support their rural-based families in expenses related to medical care, agriculture inputs, children’s education, etc.12. Security was not noted as a problem in most of provinces with regards to accessing markets for the purchase of needed food and other goods for their households, again with the notable exception of households in Bujumbura Mairie and Bujumbura Rural. Entry points for humanitarian assistance The nature of needs reported by both household level interviews and focus group discussion participants, related to the most recent shocks (political crisis and EL Nino impact) range from food, agriculture inputs, capital for small trade, shelter repair and non-food items. More sustainable needs like capacity building on land management techniques to reduce El Nino impact and livestock replacement were reported as well. When asked about their preferred forms of aid a clear preference was shown for cash, followed by food13, from households who had both received and not received aid in the past. Voucher and in-kind (non-food) distribution modalities were the least preferred. It is however noted that more female than male respondents preferred food, perhaps to avoid intra-households conflicts related to the use of money. The most common reason for preference of cash was cited as freedom of choice and need for supplementary income in order to be able to respond the various and changing needs of their families, such as health care, food and unexpected expenses. Tables 6 shows modalities preferences by sex of respondent.

12 13

Conversations with a prominent Catholic Parish leader, March 2016 No specific preference between imported and local varieties were noted

9

IRC | CRS | Market Assessment

Table 6: Q: If you were to receive aid, what is your preferred form of assistance? Aid Preferences: Women

Aid Preferences: Men

2% 2%

7%

9%

3% 6%

42% 26%

58%

45%

money

food

vouchers

NFI

money

other

food

vouchers

NFI

other

*Other responses included distribution of livestock and housing

Respondents were asked to list their top three household expenses and the answers varied widely, indicating that responding to household needs in order to best alleviate the economic burden would be best served by a flexible market based solution. Food, hygiene items, clothing, and medical expenses were the four most reported answers followed by education expenses and mobile communications. Table 7: Top Household Expenses Other Housing costs (including rent) Transportation repaying debts Mobile communications education expenses clothing health and medicine Hygeine Items Food 0

10

20

30

40

50

60

70

80

Number of Responses * Other expenses included repayment of debts, transportation and housing costs including rent.

The expenses that consume the most resources in the interviewed households are related to food consumption, followed by hygiene items. Additionally, nearly half of the interviewed households reported consuming less than ideal quantity and quality of food (substituting sweet potatoes to rice for instance). Table 8 confirms what qualitative data (from the FGDs) reveals in terms of fewer numbers of meals a day for affected

10

IRC | CRS | Market Assessment

households. This chart reflects the number of households who reported consuming less than their ideal amount key staple foods per month, in the last month. Table 8: Number of Households with less than ideal staple food consumption in last 30 days Frequency of response

35 30 25 20 15 10 5 0 Rice

Beans

Manioc

Maize

Food Commodity

Moreover, the data show a deficit balance between self-reported incomes and expenses for almost half of the sample population (33 out of 79 respondents). The affected households are currently resorting to negative coping strategies to cover urgent basic needs. The main coping strategy noted by the interviewed population is the reduction of daily meals, substitution of quality food to less preferred foods, early sale of harvest, taking on additional debt and sale of productive assets such as livestock. More dramatic strategies such as leaving the country for refugee camps in Rwanda and Tanzania, and selling plots of land were reported in the focus group discussions. In addition, due to change of rain patterns, some households resorted to changing their cultural practices by planting different crops (vegetables) and in different periods, to try and mitigate the risk of losing all of their harvest. RECOMMENDATIONS In addition to the markets’ ability to respond, we analyzed the extent to which the financial infrastructure and services available locally can support the programmatic option of distributing cash-based humanitarian assistance to the affected population. We have determined that a cash transfer program is an appropriate and feasible response. Given that the markets are still functioning (if somewhat depressed by the events of the last year) as well as the diverse needs of the affected population, the strong recommendation for intervention is via a cash-based approach, started on a small scale to test the market response and then eventually expanded to scale.

11

IRC | CRS | Market Assessment

Various options for this response were considered, including the provision of in kind aid, cash transfers and cash grants to support local vendors; and the response recommendations matrix and response options framework can be found in Annex V of this report. Summary of findings:  The economy is monetized and people are comfortable handling money;  Infrastructures and services both from financial institutions as well as markets are in place and functional; and  Cash is in line with the affected populations’ preference for aid. Following the analysis of response options, we identified three main payment delivery mechanisms that would be appropriate and feasible for delivering cash to the affected populations. The table below summarizes the primary pros and cons with each type of payment delivery mechanisms and their associated risks, when comparing them among each other: Payment Delivery Type E voucher

Remittances

Mobile money

PRO Fast distribution

CON

Highest KYC requirements of the payment delivery types Ensures appropriate pricing Highest requirement of User Knowledge Market based intervention Slow to set up and requires pre-selection of vendors – less of an issue for recurring transfers Limits freedom of choice for beneficiary only to a list of allowable commodities Recipients receive physical cash – freedom of Transfer fees are most expensive of the three choice Third-party provider manages distribution of If scale up, may require support to Remittance cash, no cash passes through our hands company for distribution. Low KYC requirements If scale up will require more support to country finance team Fast distribution Requires mobile coverage mapping Recipients receive physical cash – freedom of Requires beneficiaries to own phones – or choice need to provide handset Third-party provider manages distribution of If scale up will require more support to cash, no cash passes through our hands country finance team Low KYC requirements

IRC | CRS | Market Assessment

12

CONCLUSION This assessment and report has sought to answer the questions outlined above. The recommendations made in this report are related to the current and humanitarian situation and market situation Burundi. Knowing that the economic and political situation in the country is dynamic at the moment we do not propose to allow the findings from this assessment to be extrapolated for a response analysis to direct programming related to a massive displacement or return of population. While we, the authors of the report believe that the traders in this context have shown and continue to show incredible resilience we cannot comment on the capacity for these same traders to contribute to a mass response, should refugees return en masse. Any market based programming, be it cash based or in kind should be accompanied by regular market monitoring to ensure that we continue to do no harm to the beneficiaries we serve as well as the larger economy in which they live. ANNEXES 1. 2. 3. 4. 5. 6.

Assessment Forms Focus Group Question forms Market Price collection tool Response Recommendations Matrix and Response Options Framework Cash Feasibility checklist Maps