Multilateral Development Banks

An MDB is an international financial institution (IFI) created by a group of countries that utilizes pooled contributions from national governments and additional resources, such as interest collected from loans, to finance development projects. MDBs have memberships that include donor and borrower countries and are arranged on a global or regional basis. They provide financing, through loans and grants, and advice for development-related projects and activities to a borrowing member for the purpose of executing development projects. MDBs do not execute projects or programs themselves; instead they lend funds to executing agencies in the borrowing member countries to carry out the project or program in accordance with procurement guidelines and regulations determined by the MDB (Government of Canada, 2018). After a disaster, nations with low capital reserve often request increased or additional emergency loans to fund long-term reconstruction and rehabilitation. In such contexts, MDBs are called upon to provide support through reconstruction financing and experience, international good practice, impact evaluations, and the task of coordinating the various donors and agencies engaged in the process (Fenglar, Ihsan, & Kaiser, 2008). Some of the largest MDBs include the World Bank, Asian Development Bank (ADB), African Development Bank, Inter-American Development Bank, and Caribbean Development Bank. Table 2 provides a snapshot of all major global and regional MDBs. Subregional banks such as the Central American Bank for Economic Integration, Development Bank of Latin America, and East African Development Bank are not included due to their limited geographical reach and small number of member states (i.e., less than 20; Faure, Prizzon, & Rogerson, 2015; Government of Canada, 2018). Most MDBs were established during and after decolonization. The European Bank for Reconstruction and Development was established following the collapse of the Soviet Union. The Asian Infrastructure Investment Bank and the New Development Bank were established more recently, 20 years after the last major new MDB. The United States and Japan hold the largest percentage of voting shares in a majority of the MBDs, including ADB (Faure et al., 2015).

Once a disaster occurs, MDBs such as the World Bank, for instance, may be called upon for assistance. Since the World Bank is not a relief agency, it does not participate in the initial response. However, a team from the Bank may assist in initial impact assessments, including an estimate of financial losses from the disaster and an estimate of reconstruction costs. Based on this initial assessment, the World Bank might restructure a nation-state’s existing loan portfolio or other projects with the Bank to shift funding toward recovery. For example, it can restructure existing loans to allow for expanded recovery projects or redesign projects that have not yet been approved to address the changes caused by the disaster. Moreover, an Emergency Recovery Loan (ERL) can be granted to meet specific long-term reconstruction needs (Global Facility for Disaster Reduction and Recovery [GFDRR], 2012; World Bank Independent Evaluation Group [WB-IEG], 2006). The World Bank’s postdisaster reconstruction portfolio has increased sharply since the 1980s. From 1984 to 2005, the Bank had 528 disaster-related projects worth US$26.3 billion in its portfolio, representing 9.4% of all loan commitments. However, the World Bank’s lending has been heavily concentrated in just a handful of countries. Ten countries account for 208 of the 528 disaster projects (39%) and 7.5% of projects, mostly related to restoring physical assets, received 32% of the financing (WB-IEG, 2006). Table 3 provides a list of key MDBs that have financed postdisaster humanitarian relief and/or long-term recovery with a snapshot of the financial instruments used and examples of assistance provided.

Characteristics of MDB Financing

Three aspects characterize transboundary financing through MDBs. First, the funds flow through a small number of large, influential MDBs that finance not only postdisaster reconstruction but also a wide variety of development-related programs. As a result, transboundary MDB financing for disaster recovery is often closely intertwined with existing development loans and/or debt obligations of a nation-state. Following the 1985 Mexico City earthquake, the government of Mexico established four different housing programs. The most extensive of the four programs was the Renovacion Habitacion Popular developed to rebuild or repair 48,800 housing units in the hardest-hit neighborhoods in Mexico City. Funding for the overall program came in part from Mexican government housing programs (40%) and the remainder from World Bank loans (60%). The Mexican government’s capacity to manage post-earthquake programs was helped by concessions from the International Monetary Fund. At the time of the earthquake, Mexico was to pay back about US$900 million in principal on its foreign debt. Instead, Mexico received a 90-day postponement, and then the International Monetary Fund agreed to lend the country another US$1.6 billion while indefinitely postponing principal repayments (Comerio, 1998).

Second, the flow of funds takes place through specific instruments and procedures (Fenglar et al., 2008) that may impact predisaster infrastructure or other development projects and direct the funds in specific ways. In the case of the World Bank, postdisaster reconstruction financing has been mainly through ERLs, including International Development Association (IDA) credits and grants, and reallocations. The ERL is a three-year lending instrument that allows for expedited processing from project initiation to approval, quick disbursement, and flexibility for borrower country in meeting fiduciary requirements (WB-IEG, 2006). Reallocation, however, remains the primary fiscal response to disaster (Benson & Clay, 2004). Between 1984 and 2005, the World Bank approved ERLs worth US$9 billion and reallocated more than US$3 billion of existing loans (from 217 projects) toward postdisaster recovery. Once a nation-state requests assistance after a disaster event, the World Bank’s country staff first examines the country’s existing portfolio of loans and credits to identify loans from which funds can be reallocated for reconstruction. After the 2001 Gujarat earthquake, 12 existing projects financed through the IDA were restructured for reallocation, which provided US$416 million for reconstruction projects. Funding of US$10 million to 130 million per project was taken from the original implementing agencies and given to relevant reconstruction implementing agencies. The scope, components, and targeted sector for the projects changed substantially. Reallocations, however, divert funds from their original purposes. In 18 of the 217 World Bank projects, they have undermined the capacity to meet the original project objectives (WB-IEG, 2006).

Third, transboundary financing of disaster recovery is usually accompanied by transboundary flow of supporting services through global consultancy firms. For instance, as per the World Bank, consulting services contracted for a Bank-financed project have to be procured in accordance with the Bank’s procurement guidelines for selection of consultants. The IDA conducts prior review for each contract for a potential consulting firm estimated to be US$100,000 or more and US$50,000 or more in case of an individual consultant (World Bank, 2002b). Depending on an MDB’s procurement arrangement guidelines and on the experience, capacity, and expertise of domestic consultant firms to meet an MDB’s eligibility requirements, a number of global consultant firms can be contracted to assist in reconstruction tasks. Such global firms cross political and geographical boundaries to bring services and knowledge by way of consulting contracts during postdisaster recovery. In order to implement the Gujarat Emergency Earthquake Reconstruction Project (GEERP) after the 2001 Gujarat earthquake, for example, 18 consultancies were awarded to 15 consultants, including 12 domestic firms and 3 global firms with headquarters outside India.

Bilateral Development Agencies

Key national government agencies involved in bilateral assistance for immediate humanitarian relief and long-term term reconstruction can include overseas diplomatic missions, international development agencies (i.e., bilateral development agencies), national disaster management agencies, and the military. Among these agencies, all of which are well suited to respond to postdisaster needs due to their specific focus or expertise, bilateral development agencies (or bilateral agencies) play a key role in providing assistance for long-term reconstruction and recovery (Coppola, 2015). Bilateral agencies are institutions established by individual countries to provide development funding to nation-states. They often work closely with IFIs to advance project priorities and programs. However, unlike IFIs, bilateral agencies are responsible to a single government and are usually part of a government ministry (Government of Canada, 2018).

Transboundary funding flows from bilateral donors through their bilateral agencies and predominantly flow from the Global North to the Global South. The major bilateral agencies are from donor countries that are members of the Organisation for Economic Co-operation and Development (OECD). Formed in 1948 to administer American and Canadian financial assistance for the reconstruction of Europe after World War II, the OECD currently has 36 members, mainly from North America and Europe as well as countries such as Australia, Chile, Israel, Japan, South Korea, New Zealand, and Turkey. Some of the key bilateral agencies include the Canadian International Development Agency (now the Canadian Department of Foreign Affairs, Trade and Development), the European Commission (in the EU), the Department for International Development (in the United Kingdom), and USAID (Government of Canada, 2018).

After a disaster event, bilateral donors can provide assistance in multiple ways including direct cash, consumable products, equipment, building materials, transportation, technical assistance, or debt relief. Financial assistance for long-term reconstruction can be delivered directly to the recipient government, to UN agencies operating in the country, through multidonor trust funds (MDTFs) or via local and international NGOs (Coppola, 2015. Table 5 provides a snapshot of pledges by OECD member states, and their corresponding bilateral agency, for relief and recovery assistance after the 2004 Indian Ocean Earthquake and Tsunami. In all, 99 governments, including 22 OECD member states, 77 non-OECD member states, and 3 intergovernmental organizations (i.e., the EU, the Organization of Petroleum Exporting Countries, and the African Union), pledged 46% of the total pledged amount of US$14 billion (Flint & Goyder, 2006; Telford & Cosgrave, 2007).

USAID, for instance, coordinates United States’ response to disasters globally. USAID’s Office of Foreign Disaster Assistance (OFDA) is divided into four subunits: Disaster Response Division; Prevention, Mitigation, Preparedness and Planning; Operations Support; and Program Support. The Disaster Response Division handles American assistance provided for disasters outside the United States. Depending on the scale of the disaster, a Disaster Assistance Response Team (DART) is deployed to the country to assess the damages and recommend the level of assistance that should be made by the U.S. government. The DART team coordinates U.S. relief supplies; provides operational support; coordinates with other donor countries, UN agencies, NGOs, and the host governments; and monitors and evaluates projects carried out with U.S. funds. Following the 2001 El Salvador earthquakes, the USAID/OFDA provided US$20.5 million in emergency assistance through financial and material supplies. For long-term recovery, the USAID created the Earthquake Recovery Program to address rural reconstruction. Through this program, the United States provided a budget of about US$75 million in fiscal year 2001 and US$100 million in fiscal year 2002 to provide assistance to more than 1,000 rural communities. The program built 26,872 homes for disaster-impacted households, 53 schools, 19 city halls, 5 health units, and 3 markets, as well as provided assistance to small farmers and small and micro enterprises (USAID, 2018).

Characteristics of Bilateral Agency Financing

Three aspects characterize transnational transboundary financing through bilateral agencies. First, the funds provided by bilateral agencies may be tied or untied. The OECDs Development Assistance Committee reached an agreement in 2001 on a recommendation to untie official development assistance to the least developed countries. Untied aid is financial assistance that is freely available with no restrictions on where the recipient country can buy the goods and services it needs for a project. Tied aid specifies that the recipient country must procure goods and services from the donor country. However, tied aid has been found to raise the cost of goods and services by 15% to 30%; to provide goods, technology, and services that do not conform to the priorities and specifications of the recipient country; and to favor capital intensive or donor-based technical expertise rather than small poverty-focused programs (OECD, 2001). Most of the major bilateral agencies provide untied funds, which means that there are no eligibility restrictions based on a company’s nationality (Government of Canada, 2018). After the 2014 Indian Ocean Earthquake and Tsunami, donor pledges were flexible and assistance was untied and often paid in full “up front” (Flint & Goyder, 2006).

Second, bilateral agency funding for postdisaster reconstruction are increasingly channeled through MDTFs. MDTFs have emerged as one vehicle for channeling and coordinating reconstruction resources especially for disaster events where a significant percentage of financing comes from bilateral and multilateral donors (Fenglar et al., 2008). MDTFs are financial and administrative arrangements with external donors to finance high-priority needs, provide assistance to increase donor confidence in countries with weak fiduciary systems, and reduce coordination and administration costs through joint arrangements. The funds come from donor countries, foundations, the private sector, and sometimes the IFI’s own grant resources. The IFI is responsible for administering and allocating the funds (Fenglar et al., 2008; Government of Canada, 2018). The World Bank, for instance, has been both a trustee and administrator of MDTFs. Donor conferences have become an important mechanism for mobilizing such international assistance. In such a forum, the preliminary estimates of damage/loss and needs assessments are presented, together with initial key policies of the government for directing the reconstruction (including the potential establishment of an MDTF; Fenglar et al., 2008).

Third, the financial and other terms on which bilateral assistance is offered can vary widely. Following the 2004 Indian Ocean Earthquake and Tsunami, instead of official development assistance, most bilateral assistance was provided in the form of grants, cash or in-kind. The remaining support was provided in loans, mainly as soft loans on below-market term loans. A small part of the total funding was provided immediately in the form of grants with few strings attached, and a larger amount was offered with various conditions attached in a mixture of grant and loan terms over an uncertain period. The conditions for grants and loans set down by bilateral donors were wide ranging, including attaching onerous procurement conditions to the provision of financial assistance, setting up different reporting requirements that were difficult for recipient agencies to meet, and looking for local partnership commitments from national institutions that placed a heavy burden on an overstretched local administrative system (Jayasuriya & McCawley, 2010).

Coordinating Financial Assistance

According to the UN Resolution 46/182 of 1991, a disaster-impacted nation-state has “the primary role in the initiation, organization, coordination, and implementation of humanitarian assistance within its territory.” Nevertheless, the potential for international aid agencies to undermine or inappropriately substitute the authority of the state has sometimes led to tense relations between nation-states and international actors. There is some tendency among international agencies to assume that governments will be too corrupt to deliver financial assistance effectively. While there is a growing move toward the establishment of MDTFs and other mechanisms to deliver financial assistance to recipient governments directly for postdisaster long-term recovery needs, funds are often still channeled through international aid agencies, the UN, and/or international NGOs. It may be more appropriate for donors to fund governments directly, yet, bilateral donors often bypass and marginalize governments partly due to lack of trust in the ability of states to deliver effective and accountable assistance (Harvey, 2009).

Moreover, scholars (Jayasuriya & McCawley, 2010) note that donor governments and aid agencies do not coordinate their activities well—with each other or with the national government that should ultimately have control over operations when disasters occur. For instance the World Bank’s ERL, a lending instrument for expedited processing during an emergency, is usually implemented within three years. However, postdisaster recovery often comprises large-scale complex projects spread over a large area, the implementation of which can be challenging to complete within a three-year period. Following the 2001 Gujarat earthquake, the GEERP’s original implementation time doubled from three to six years (ADB, 2008). The Bank did not realize during the early stage of the GEERP that the initial project period of three years was highly unrealistic and that it would impact urban housing reconstruction (World Bank, 2009). The three-year deadline translated to a tight implementation schedule, put in place to ensure timely disbursement of funds, which was incongruent with the slow pace of urban planning and reconstruction on the ground. In keeping with the GEERP schedule, GSDMA released the first of the three housing reconstruction assistance installments almost three years before the beneficiaries could actually apply for building permits. As a result, the first disbursement of housing installments did not match the timeline of planning and land readjustment processes, creating significant barriers for households rebuilding their homes (Ganapati & Mukherji, 2014). In more than 7.5% of cases, beneficiaries were unable to complete their homes within the required period and on budget (World Bank, 2009, p. 10).

The time limit of ERL projects has also led Bank staff to “rush certain activities unnecessarily” (WB-IEG, 2006, p. 33). In India, the World Bank (2009) notes that, “the limited time stipulations of [operations manual] OP 8.5, rather than a realistic assessment of the time required for project implementation, drove the three-year project” (World Bank, 2009, p. 7). The World Bank acknowledges that this time limit was unrealistic (World Bank, 2009, p. 5). It had assumed that all beneficiaries would be able to build to the construction targets set within the three-year time period. However, it excluded beneficiaries who were unable to maintain program deadlines (World Bank, 2009). The World Bank’s Independent Evaluation Group (WB-IEG, 2006) also points out that in postdisaster recovery projects the World Bank has in general allowed too little time for implementation, especially since implementation times can range from two to seven years. Many crucial postdisaster recovery activities require more than three years for completion. On average, projects involving land acquisition took 7.5 years and those with infrastructure activities 6.5–7 years. This has led to revised implementation times for most disaster activities. The emphasis on strict time for task completion limits risks the exclusion of certain activities from projects that require a longer time-frame but would achieve long-term development goals (WB-IEG, 2006, p.33).

Channeling and Tracking Financial Assistance

Due to the unprecedented levels of global funding and the large number of diverse entities involved in postdisaster reconstruction, credible and integrated financial tracking systems have become more critical. Transboundary flows of postdisaster recovery funds, however, can be complex and difficult to track and coordinate (Amin & Goldstein, 2008; Augustina, 2008). Funding flows among traditional donors, international NGOs, UN agencies, local NGOs, and the government can be complex, as funds are transferred across multiple levels and between agencies before reaching implementing agencies for final disbursement (McKeon, 2008). Unlike the UN OCHA’s cluster approach introduced in 2005, one year after the 2004 Indian Ocean Earthquake and Tsunami, to provide a framework for international coordination of immediate humanitarian aid and action (i.e., to concentrate resources for immediate relief and response and ensure that gaps in humanitarian efforts at national and international levels are detected quickly) (Stumpenhorst et al., 2011), a similar approach to coordinate resources for long-term recovery after a disaster has been slow in coming. Following the 2004 Indian Ocean earthquake and tsunami, initial assessments were often slow, overlapping, poorly shared, and imprecise. Most agencies active in recovery efforts collect large amounts of data from a predetermined group of beneficiaries, but few were transparent and open in processing these data so as to permit monitoring (de Goyet, 2008). As a result, multiple assessments were conducted to satisfy agencies’ own requirements; they were rarely shared and hence did not influence collective decision-making. Moreover, disaster survivors had to undergo numerous interviews and fill out multiple questionnaires due to the duplication of efforts, which contributed to a sense of resentment and a feeling of exploitation among the population (de Goyet, 2008).

Following their trilateral joint declaration (EU et al., 2008), the EU, the United Nations Development Group (UNDG), and the World Bank have designed and developed the guidelines for the PostDisaster Needs Assessment (PDNA). This framework has since be increasingly used as the base document to determine and coordinate international assistance and to inform recovery plans for countries seeking external funding for recovery after a disaster. The PDNA, conducted sector-by-sector (i.e., agriculture, commerce, education, environment, health governance, housing, water and sanitation, and transport), is to recommend how funds and resources pledged at international donor conferences should be allocated to address recovery needs for each sector. There is little evidence, however, of coordination across sectoral boundaries (e.g., focus on links such as the connections among housing, land and infrastructure, and the local economy) (Mukherji, 2017). Moreover, these efforts are confined to the planning stage and not carried over into implementation, as there is no requirement that the recommendations be adopted.

International partners often prefer to use a country’s public financial management (PFM) systems to channel financial assistance during postdisaster assistance, as this is where their power to exert influence within foreign government bureaucracies lies. For funds that go through regular budgetary systems, more flexibility is needed to allow for faster disbursement and reallocations, which is only possible if special fund-flow mechanisms are in place from the beginning. A large share of project implementation is however channeled outside regular budgetary processes, particularly if NGOs and the UN system are playing a significant role. Without robust monitoring and evaluation systems in place, such an approach can make it difficult to track the use and impact of the funds (Fenglar et al., 2008). Following the 2011 Haiti earthquake, official bilateral donors and MDBs pledged US$13 billion toward immediate relief and long-term reconstruction efforts. Almost 36% of the long-term recovery and reconstruction grants went to international NGOs and private contractors. Lack of a systematic tracking mechanism made it difficult to identify the final recipients and specific details about how the money was spent, including outcomes of projects, as financial reports and rigorous impact evaluations were lacking. For instance, about 60% of U.S.- and Canadian-disbursed recovery assistance was categorized as “non-specified,” while 67% of immediate humanitarian and 43% of long-term recovery funding from the European Commission went to international NGOs with no clear data on how such funds were utilized (Ramachandran & Walz, 2013).

While documentation on individual projects funded by MDBs, bilateral agencies, and UN agencies are usually available, comprehensive and aggregate information on projects after a disaster are difficult to find due to an absence of reliable financial information and the lack of well-functioning data systems (Augustina, 2008). The complexity of funding flows makes it difficult to track funding allocations and see whether they match specific needs. In Indonesia, after the 2004 Indian Ocean earthquake and tsunami, funding flows among UN agencies, local and international NGOs, and the national government were complex (Amin & Goldstein, 2008). Most large international NGOs were well funded from private sources and did not ask for official donations. As a result, bilateral funding flowed mainly through UN agencies (Telford & Cosgrave, 2007) and MDBs. Donors provided funds for specific sector allocation rather than geographical needs and were unaware which areas were receiving funds and which were not (Amin & Goldstein, 2008). Financial tracking could not present comprehensive, accurate, and up-to-date information on funding at any stage (Telford & Cosgrave, 2007).

Supply-Driven Financial Assistance

International financial assistance can flow in ways that support donors’ (including bilateral donors’) agendas rather than put the affected community in the driver’s seat (Jayasuriya & McCawley, 2010; Telford & Cosgrave, 2007). Because donors play a key role in supplying reconstruction financing, implementing entities such as international NGOs are often required to view donors as clients instead of partners, whose wishes are prioritized over the needs of the affected community. Bilateral donors can place restrictions on their funding by specifying how to disburse the funds (Stephenson, 2005). This can result in supply-driven, unsolicited, and inappropriate financial assistance that in turn creates inequities, gender- and conflict insensitive programs, unsuitable housing design and livelihood solutions, and waste (Telford & Cosgrave, 2007).

A key aspect of supply-driven financial assistance is the discrepancy between offers of international assistance and the capacity to absorb the assistance. Following the 2004 Indian Ocean Earthquake and Tsunami, the pressure by donors on organizations, such as India’s Church’s Auxiliary for Social Action (CASA), which received more than US$30 million in international assistance, to receive and spend funds exceeded their capacity to utilize funds well. CASA had to ask donors to hold back funding, because amounts were more than what they needed (Rempel, 2010). In Aceh, financial complexities were experienced because of the multitude of donors and other actors. Local NGOs, who already had significant funds of their own, received additional external funds for their districts. However, many of the NGOs had never managed such large quantities of funds before (McKeon, 2008).

Another discrepancy arises when allocation of funds is not based on needs. After the 2004 Indian Ocean Earthquake and Tsunami, world leaders vied with one another to announce spectacular aid pledges driven by media and political considerations, without regard for the actual needs or capacities of the affected countries. Though long-term reconstruction and recovery needs were the most important, immediate relief and long-term recovery received almost the same amount of funds. About half of the pledges made were for emergency assistance, one-quarter for recovery to be spent in 2006, and one-quarter for recovery during the period of 2006 to 2010 (Engelhardt, Hidalgo, & Sole, 2006; Telford & Cosgrave, 2007).