Chapter 6: Money: Savings, Credit, and Grants After the Shock¶

On January 12, 2010, the earthquake that strikes Haiti does not spare Fonkoze, the country's leading microfinance institution. Ten of its forty-one branches collapse or are badly damaged. Five of its employees are killed; more than 450 of them see their homes badly damaged or destroyed. Its clients are mostly ti machann, market women. More than 19,000 of them see their home, their business or their marketplace swept away by the quake, or collapsing in the days that follow.

In the focus groups the institution organizes afterward, one finding keeps coming back: many clients had already restarted a small business in the days and weeks after the earthquake, because they urgently needed an income. A client from Bizoton, a member of a solidarity group, sums up what those weeks were like: "My credit agent was the first person to come looking for me after the earthquake. When I saw him, I had hope, because I knew Fonkoze had not abandoned me."

Fonkoze had just evaluated its lending program after the 2008 hurricanes and was preparing a permanent catastrophe insurance product for its clients. It decides to act as if that insurance already existed. Every affected client, meaning every client who lost a home, a business or both, receives the package planned under the contract in preparation: a one-time payment of 5,000 gourdes, about $125; cancellation of the balance on the loan taken out before the earthquake; and a new loan to restart the business, if the client is ready after a careful assessment of their capacity, and agrees to pay a retroactive fee of 2% of the loan amount. The fee is there to say one thing: insurance has a price. According to the institution's report, 19,811 affected clients receive this package, and 9,637 families hosting disaster survivors receive the same 5,000-gourde payment.

Here, the gift brings credit back to the clients.

What the scene teaches¶

Three things, which organize this whole chapter.

First: the money that matters most after a disaster flows through institutions that existed before. The Bizoton credit agent knew where his client lived, what he sold, how much he owed. No organization that arrived after the earthquake knew any of that. A local lender who knows its clients can act within a few weeks; a Rotary club trying to rebuild that knowledge would take months.

Second: wiping out a debt is not a trivial gesture. Fonkoze canceled the loan balances of nearly 20,000 affected clients, but not out of improvised charity. Using money from outside partners, it carried out an insurance contract it had been preparing, paired with a fee and a new loan granted after assessment. It even designed a post-disaster assessment form to check that each client could carry new debt; the Smart Campaign, an international initiative for client protection in microfinance, cited it as a model. A write-off that is planned and funded does not have the effects of a general write-off, which makes people believe loans no longer need to be repaid.

Third: disaster survivors need capital, and they know what to do with it. After the payments, Fonkoze surveyed 274 affected clients in four of the hardest-hit branches. According to this follow-up, 76% had invested at least part of the 5,000 gourdes in their business, 28% all of it; 95% had taken out a new loan; and the repayment rate on these new loans reached 97% as of December 31, 2010. These are the institution's own figures, and the report cautiously says these repayments "may indicate" sound credit management; Chapter 13 will come back to what figures like these prove.

Your club is not a bank. It knows the banks. This chapter explains what it can do with that knowledge.

What happens to money in the weeks that follow¶

For clients¶

The 1998 floods in Bangladesh are the best documented. According to a press article cited in the first practitioners' guide, more than 90% of the clients of microfinance institutions (MFIs) had to suspend their usual business for more than three weeks, and the loan recovery rate fell from 92% to 43%. The same article reports that about 65% of clients suffered losses or damage to their business assets.

A field study of these floods makes an observation every banker in your club will understand: the flood did not hit everyone the same way. Some were hit hard, others moderately, others not at all, and some even benefited, such as those who ferried passengers by boat. This is the central argument against any uniform measure.

For lenders¶

The practitioners' guide published in 1999 by the Microenterprise Best Practices project, funded by the United States Agency for International Development (USAID), describes four simultaneous behaviors: some clients stop repaying, others stop saving, others want to withdraw their savings, others ask for emergency loans. "All of these behaviors reinforce one another to create a drain on the MFI's liquidity": money flows out of the institution just as it stops flowing in. And donor money generally takes several weeks to arrive.

The Bangladesh figures show the scale of the problem. Small MFIs had to limit withdrawals of compulsory savings to 50 or 75% of balances. The national body that refinances MFIs, PKSF, estimated its partners' needs at 9,200 million taka; it disbursed 1,000 million.

Fifteen years later, in the Philippines, the MFI CEVI drew the same lesson from Typhoon Haiyan: "liquidity played a bigger part and so was more important than capital." It also notes something few donors know: "Relief funding in disasters is generally not available to fund lending." Hence a perverse chain of events: lenders cut back their lending, and if repayments exceed new loans, money flows out of the affected communities just when they need it most.

People's own money¶

Before any aid arrives, disaster survivors draw on what they have. In Sri Lanka, after the December 2004 tsunami, a study of damaged small businesses shows that in the first three months, 51% of the money spent to repair or replace their equipment came from their own savings, 15% from loans from family and friends, and only 2% from MFIs. In Haiti, seven weeks after the earthquake, a survey in five camps in Port-au-Prince found that 40% of residents had received informal help, most often money sent by family living abroad.

The first money of recovery is therefore that of the survivors, their relatives and their usual lenders. A club's role is not to replace it, but to help these three sources hold up.

What money spent locally does around it¶

A dollar paid to a disaster survivor goes to the grocer, who pays his supplier and his helper, who in turn buy things. That is the multiplier effect. Two studies put a number on it, and neither concerns a natural disaster.

The first is the Kenya experiment that Chapter 5 cited on prices: about $1,000 paid in a single transfer to more than 10,500 poor households in 653 villages, outside any disaster. The National Bureau of Economic Research researchers estimate the local multiplier at 2.4: each dollar transferred added about $2.4 to the area's output, the average of two methods (2.5 and 2.3). Households that received nothing benefit too: eighteen months after the payments began, their consumer spending is 13% higher, and their income rises mainly through the wages paid by local businesses, whose sales have increased. The average inflation measured is 0.1%. The paper states its limits: the estimate is imprecise, the hypothesis of a multiplier below 1 is rejected only at a marginal significance level, and one of the authors cofounded the NGO that paid out the money.

The second comes from Lebanon, in the winter of 2013-2014. A program run by the Office of the United Nations High Commissioner for Refugees (UNHCR) and its partners pays $575 by bank card to 87,700 Syrian refugee families, for heating. This is a refugee crisis, not a natural disaster. The evaluation commissioned by the International Rescue Committee finds no notable price increase across some fifty products; the money is spent in local shops. It estimates that each dollar spent by recipients generates $2.13 of GDP for the Lebanese economy. This figure is not measured: it is calculated with a formula, using the propensity to consume estimated for Jordan, for lack of a Lebanese figure, and the evaluators describe it as conservative.

Neither study says what happens to the effect when shops are destroyed or supplies come from far away. In Kenya, the authors link it to a largely closed local economy: a dollar spent outside the area no longer circulates there. For your club, the consequence is practical: when a project buys materials, services or labor, buy and pay locally whenever the local market can supply (see Chapter 8).

The principle: reschedule, case by case¶

A long-standing consensus¶

Since Hurricane Mitch and the Bangladesh floods in 1998, practitioners' guides have said the same thing. Loan rescheduling must be done right after the disaster and only in hard-hit areas. It is decided client by client, or borrower group by borrower group, rather than through a blanket measure. The Consultative Group to Assist the Poor (CGAP) put it in writing in February 2005, after the tsunami, in a note addressed to MFIs and their donors: donors "should encourage microfinance institutions to restructure, rather than forgive, loans." And it gives the reason: encouraging them to forgive loans or act as social safety nets will make it harder for them to operate sustainably, and so to reach more people once reconstruction comes.

The Minimum Economic Recovery Standards (MERS), published in 2017 by the SEEP Network, restate this doctrine almost word for word: "Writing off loans should always be a last resort as it can reduce the financial institution's capital and worsen credit repayment culture." The MERS also set out the exceptions: institutions must be ready to write off loans that cannot be recovered because of the borrower's death, permanent disability or disappearance. The other exception is the one in the opening scene: a write-off planned in advance through an insurance contract.

What rescheduling costs, and what not rescheduling costs¶

The Bangladesh figures speak for themselves. On a loan of 3,000 taka ($60) rescheduled by two months during the flood, the only cost to the MFI, measured in May 1999, was 2 to $3 in lost earnings. By contrast, in September 1998, a survey of more than 500 affected clients of the MFI BRAC who had not yet been granted rescheduling found that fewer than 20% managed to repay out of business income. Nearly half were borrowing from family or moneylenders, at rates of 120 to 140%, to stay current and keep access to BRAC's services. And 15 to 20% were selling assets or relief goods, withdrawing their savings or eating less.

In other words, an installment kept on schedule costs the borrower far more than a postponed installment costs the lender.

How to reschedule¶

The 1999 guide describes three forms: postpone principal only; postpone principal and interest while letting interest accrue; or postpone both and stop the clock. To catch up on missed installments, you lengthen the loan term, increase the remaining installments, or accept a lump-sum repayment after a specific event, "such as when the water recedes or after a harvest."

Case by case does not mean arbitrary. The guide recommends a written policy that gives loan officers parameters and a few model schedules. The MERS call for meeting each borrower to assess their repayment capacity after the disaster.

Durations vary: in Bangladesh in 1998, from three to ten weeks depending on the client; after Mitch, up to nine months, with the most vulnerable recovering more slowly.

The group trap¶

In Nicaragua after Mitch, the MFI ProMujer rescheduled the loans of groups of thirty women. In each group, about two thirds recovered quickly and repaid on the original schedule; one third followed the slower schedule. But the rule required all thirty to finish repaying before any new loan: the whole group was stuck. The lesson is direct: rescheduling must also adapt the rules for access to the next loan.

The order of money: transfers, savings, then credit¶

The MERS give a simple sequence: "organizations should prioritize inward remittances and cash transfers immediately following a crisis. Next, they should work with FSPs to increase access to savings and insurance. During a reconstruction or recovery phase, FSPs can reintroduce (or expand) loan products based on client demand for access to finance." FSPs are financial service providers. The Part III opener placed this sequence in the overall timeline of the levers; here is what it means for each form of money.

Migrant remittances¶

In countries with many emigrants, diaspora money arrives before aid. A 2010 World Bank paper shows that remittances increase after disasters where emigrants make up a large share of the population, substantially in about half of the countries studied, and recommends restoring money transfer services quickly.

After Mitch, the transfer service from the United States run by a Salvadoran federation of savings and credit cooperatives saw the number of transfers triple in the month after the hurricane.

A club in a country of emigration has a role here that no agency has: it can press the authorities and operators to reopen transfer points quickly, and make safe, low-cost channels known to the diaspora.

Savings¶

CGAP asks MFIs to lift compulsory savings requirements temporarily in affected branches "until the emergency stage has passed and clients have begun reconstruction," and to open withdrawals according to their liquidity. The 1999 guide suggests making compulsory savings available as a loan, with a date for rebuilding the balance set according to the scale of the disaster, on which access to future loans will depend.

Compulsory savings offer little protection. In Bangladesh, according to Grameen Bank, 95% of the compulsory savings of affected clients was withdrawn, and these balances, built from deposits of a few taka, counted for little against losses of about 10,000 taka per household. Voluntary savings protect better: MFIs that already offered voluntary savings with open access suffered less from liquidity crises than those that had to release compulsory savings in an emergency.

During the flood, a credit officer from the MFI ASA visited her members every other day with two options: withdraw their savings, or take a consumption loan of 1,000 taka. In one group, ten withdrew, ten borrowed, three did both. Those who borrowed knew the loan cost more; they wanted to keep their savings intact.

And aid money flows back into savings. According to the MERS, MFIs in the tsunami-affected areas of Sri Lanka reported higher savings balances in March 2005 than in March 2004, and in Batticaloa district, about 35% of the cash grants and cash-for-work wages received were deposited in MFIs.

Savings groups¶

Savings groups (Village Savings and Loan Associations, self-help groups) are often presented as the resilience solution par excellence. Let us be clear: the research gathered for this book found no rigorous impact study of savings groups after a sudden-onset disaster. In Dar es Salaam, households with a member in a group recover faster from floods, but that is a correlation, not proof. The social fund of these groups works like a small insurance scheme, with a limit the authors themselves spell out: "a community-based insurance mechanism will not be viable if all community members are affected by the same event." In Ethiopia, during the 2015 and 2016 drought, older groups held up better than groups created less than a year earlier.

The consequence for your club: support the groups that exist rather than hastily creating new ones after the disaster. When no financial institution guarantees a lasting presence, the MERS recommend working with community-based financial services, owned and run by their members, "such as ROSCAs and savings groups." The CED guide and Module 3 of the CED training program describe how they work in normal times.

Recovery credit¶

When credit comes back, it comes back first in its simplest form. The 1999 guide reports that Bangladeshi MFIs consider the best recovery product to be "the standard microfinance short-term working capital loan": the ordinary working capital loan. The reconstruction loan, for its part, raises a debt problem. CGAP says it bluntly: "The amounts of many such 'asset replacement' loans require the borrower to also go through three or four income generation loans to earn enough income to service these reconstruction loans." It recommends granting them only after the emergency stage, when MFI staff can assess each client's damage and creditworthiness.

Three experiences show what well-run recovery credit achieves, and its limits.

Poland, 1997. After the floods, the MFI Fundusz Mikro opens a temporary window under a separate brand: a 24-month loan at 10%, with six months' grace. On-time repayment reaches 93%. The product is limited to six months, which, according to the World Bank, preserved credit discipline.
Philippines, after Haiyan (November 2013). CEVI does the opposite of its competitors: it lends more. Its recovery loan is disbursed within one to two weeks, with a schedule matched to each borrower's cash flow. According to the lender and its funder, 4,889 loans averaging $430 were repaid better than the institution's average. This is a self-assessment, without a control group.
Nepal, after the earthquake of April 25, 2015. A grant from the Asian Development Bank finances, through an agricultural development bank and 53 small farmers' cooperatives, recovery loans at 5% a year over three years, including six months' grace. According to the Asian Development Bank's report, 15,700 households benefit, with women making up 76% of borrowers, and in July 2018 the on-time repayment rate among recipients reaches 99%. But the loans were disbursed from July 2016 to October 2017: fifteen to thirty months after the earthquake.

That last figure matters as much as the repayment rate. Recovery credit arrives late. In the months before, something else is needed. And it does not suit everyone: the MERS point out that lending "is not an appropriate intervention for individuals who are not economically active."

Chapter 5 showed the other risk: the subsidized rate. In Sri Lanka, loans to victims at 6% were followed by abnormal defaults among borrowers who had not been affected by the disaster. The interest rate is the lender's decision, made with full knowledge, not a club's.

Keep relief and credit apart¶

One rule runs through all the sources. CGAP: "MFIs are not relief agencies." Donors should separate emergency aid and microfinance as soon as possible: "Clients should not perceive mixed messages, so that the credit culture is not damaged." An MFI that temporarily provides relief should signal it with visible signs and keep separate accounts.

In Aceh, after the tsunami, an expert who evaluated microfinance in 2008 draws a harsh conclusion: about 400 donor organizations, poorly coordinated; of 1,502 officially registered financial cooperatives, 85 active. He draws two rules from it: "Grants and loans must not be provided by the same institution lest both staff and recipients will be confused, and any emerging, or pre-existing, credit culture will be undermined," and "Disbursement does not equal effectiveness." He recommends starting by mapping the financial landscape, local rotating savings groups included.

CGAP draws an instruction for funders from this: avoid setting MFIs numerical targets for clients served, because this can push them to lend to clients who will not be able to repay. Your club will therefore never measure its success by the number of loans granted.

Grants to small businesses¶

Between the emergency and the return of credit, there is a gap of several months that direct grants to small businesses can fill. Two randomized studies, the only ones found for this book after a disaster in a developing country, show what they are worth.

Sri Lanka, 2005. Researchers from the World Bank and two universities follow very small businesses with no employees in the districts of Kalutara, Galle and Matara. One hundred twenty of them, drawn by lot, receive 10,000 or 20,000 rupees (about 100 or $200), in cash or in equipment chosen by the owner. Result: for retail businesses, each 100 rupees received raises monthly profit by 19.6 rupees, a return of 19.6% a month. For manufacturing, the effect is zero on average. The authors conclude that after the tsunami, lack of capital was not the main obstacle for manufacturers. They suggest likely reasons: coconut fiber production was waiting for the retting lagoons to be cleared, lace making was waiting for the tourists. They support cash grants, "but only in limited cases," and observe that aid to businesses was "not well correlated with reported damage."

Beira, 2019. Seven months after Cyclone Idai, researchers from UNU-WIDER send 6,000 meticais (about $100) by phone to 121 small manufacturers drawn by lot, in Beira, a very hard-hit city, and in Chimoio, less affected. The amount matches the median annual investment of a Beira microenterprise before the cyclone. The effect is significant in Beira only. It is concentrated in one trade: "Treated carpenters stand out as their income increased by 102 per cent compared to the remaining firms." No measurable effect among tailors, blacksmiths or food producers. The authors explain it by the demand for repairs and furniture after the cyclone.

The two studies agree on one point: the same amount does not produce the same effect from one trade to another. A grant restarts a business whose customers and suppliers have come back; it replaces neither missing inputs nor vanished demand.

The United Nations Development Programme (UNDP), in its 2013 guidance note on emergency employment, adds three rules: the restart grant is one-time only, and announced as such; preferably, it is not paid by the institution that lends; it stops when markets pick up again. The International Labour Organization (ILO) warns along the same lines: any grant can weaken the local financial sector and create dependency on aid.

What the club does¶

A financial partner, not a teller window¶

The CED guide sets the rule for normal times: your club does not lend itself; it prepares borrowers, connects them with a professional lender and tracks the results. After a disaster, the rule becomes stricter, because the risk of over-indebtedness rises just as incomes fall.

Need	Who handles it	What the club does
Current loans that can't be repaid	the lender, case by case	pay nothing on borrowers' behalf; talk to the lender
Lender liquidity	the lender and its refinancers	look for funders; check the lender
Savings frozen or lost	MFI, cooperative, savings group	support existing groups
Diaspora money	transfer operators, authorities	press for transfer points to reopen
Capital to reopen	a partner separate from the lender	fund targeted grants, if the rules allow
Recovery credit	a registered lender	prepare and support borrowers

Before touching anything that concerns existing loans, talk to the lender. Paying borrowers' installments from outside can break the repayment discipline on which all future credit in your area depends.

Funding a lender: the DAI manual method¶

The guide closest to the situation of a club or district is a manual published in 2007 by the firm DAI for funders that support MFIs with grants after a crisis. It proposes four steps.

Check three conditions: minimal political stability, enough cash-based economic activity, and a stable population. For this last point, it uses at least 18 months, the length of time "commonly recognized as the minimum period of time required for MFIs to make and recover their loans." Without these three conditions, "Providing grant support to MFIs should not be considered."
Check your own capacity: access to a microfinance specialist, and a funding commitment of at least three years, coordinated with other donors.
Choose the type of support, from simplest to most demanding: equipment and materials (a destroyed branch to re-equip), short-term technical assistance, capitalization of the loan portfolio, operating costs. Capitalization "Provides critical liquidity," but requires a more thorough assessment.
Assess the institution (governance, interest rates that cover costs, low arrears, accounts audited every year), then disburse in tranches, against performance benchmarks. Module 3 of the CED training program lists the documents to request.

The Rotary Foundation's rules¶

The rules below are those in the documents in force in July 2026. Check them against the current Terms and Conditions on My Rotary before any application.

Microcredit is eligible. The April 2024 policy statement for the community economic development area of focus lists access to financial services for poor communities, "including microcredit, mobile banking, savings, and insurance," among the activities of the area of focus.

In a global grant, microcredit is tightly regulated. The July 2026 Terms and Conditions require a registered MFI that has been operating for at least three years; capital lent out at least twice, over at least 18 months; an annual interest rate no higher than the lower of two caps, ten points below the national average or 36%; training for recipients that goes beyond loan management; a midterm visit by the Cadre of Technical Advisers; and a final report with a baseline and an exit plan. The April 2024 CED guidelines add an additional review for any microloan project.

A district grant can fund microcredit activities with a registered MFI that has been operating for at least three years.

Two prohibitions matter for recovery money. Neither the Disaster Response Grant nor the global grant funds cash donations not earmarked for a person or a cooperating organization, nor loan guarantee schemes. If your club is considering grants to small businesses on the Sri Lanka or Beira model, how to fit them into a Foundation grant is not settled by the documents read for this book: ask the regional grants officer before building the project.

Timing. A global grant pays for nothing committed before its approval. We conclude from this that it cannot provide liquidity to a lender in the first weeks: other sources are needed, club and district funds, the lender's partners, donors. Fonkoze itself was funded in two stages, a test with 500 clients and then an expansion; Chapter 10 details how this was put together.

Insurance, finally, is not decided after the disaster. It is prepared beforehand, at the level of the lender or the group: that is the subject of Chapter 15.

Pitfalls¶

Pitfall	Why it happens	Remedy
Paying affected borrowers' installments	you want to bring quick relief	talk to the lender; rescheduling costs little
Funding a general write-off	the gesture seems generous	write-offs only if planned, targeted or insured
Lending yourself, or lending to everyone	the club has money and goodwill	a registered lender that assesses each borrower
Lending to rebuild before restarting business	the house seems the priority	working capital first, for active people
Creating savings groups to receive aid	it is quick and visible	support existing groups
Same grant for every trade	it is simple and seems fair	check what is blocking: capital, inputs or customers
Giving and lending under the same banner	one partner is more convenient	keep the windows separate
Counting success in number of loans	it is easy to report	measure repayment and business recovery

What your club now has in hand¶

Your recovery file now gets its "money" section: the local financial partner chosen and checked, the form of support selected (lender liquidity, restart grants, support for savings, remittances) and the safeguards against over-indebtedness.

The Bizoton credit agent had no global grant and no recovery plan. He had a list of clients and he knew where to find them. That knowledge is what your club must help keep standing. But money only restarts those who already know how to work, and reconstruction will call for trades the area does not have in sufficient numbers: that is the question of the next chapter.