Beyond the damage: probing the economic and financial consequences of natural disasters
by Charlotte Benson and Edward Clay, ODI July 2004

The reported global cost of ‘natural’ disasters rose 15-fold between the 1950s and the 1990s. During the 1990s, major catastrophes resulted in reported economic losses averaging an estimated $66 billion per year (in 2002 prices). Record losses of some $178bn were recorded in 1995, the year of the Kobe earthquake in Japan – equivalent to 0.7% of global GDP. Numbers of people affected have also risen sharply, with a three-fold increase between the 1970s and 1990.

These alarming increases have triggered growing awareness of the potential human, structural and economic threats natural hazards pose. However, there is only a limited sense of their broad economy-wide or macroeconomic significance, or their implications for longer-term development. This is partly because impact assessments often concentrate on the most easily measured direct losses occurring as a consequence of a disaster. This largely reflects immediate concerns to meet the most visible short-term humanitarian needs of affected people, and to determine replacement investment requirements and insured losses. There has been much less focus on indirect and secondary impacts, such as on the demand for goods and services and livelihood opportunities. These effects can also have significant humanitarian consequences, as well as implications for poverty reduction and other development objectives.

Issues of vulnerability

Underlying the impact of a disaster is the issue of vulnerability. There has been relatively little work on this from a national perspective, although there is a considerable body of work at a household level, particularly on drought in sub-Saharan Africa. Household-level research plays an important part in informing both risk reduction and post-disaster response efforts. National-level analysis should play a significant complementary role, placing vulnerability within a broader socio-economic, political and policy context, and capturing shifts over time more clearly.

Research at a national level shows that vulnerability to natural hazards is determined by a complex and dynamic set of influences, such as the economic structure of a country, its stage of development and prevailing economic conditions and policy. Vulnerability changes quickly, particularly in countries experiencing rapid growth, urbanisation and socio-economic change. Regular re-assessment of vulnerability is required to ensure that disaster management strategies, including the nature and form of post-disaster relief and rehabilitation, remain appropriate.

Screen Shot 2012-10-18 at 4.14.53 PM

Screen Shot 2012-10-18 at 4.15.03 PM

Reduced vulnerability in Bangladesh

In Bangladesh, vulnerability to cyclones and related coastal storm surges has been reduced through a combination of shelters, embankments that reduce impacts, enhanced early warning and evacuation planning. Meanwhile, the economy’s sensitivity to extreme monsoon flooding and drought has also declined. This is partly due to structural change in agriculture, with a rapid expansion of much lower-risk dry season irrigated rice, internal market integration and increased private food imports during disaster years. Despite extreme and long-lasting flooding in October and November 1998, cereal production was actually 5.6% higher year-on-year in 1998/99 (the government’s pre-flood forecast was for 2.4% growth). Initial assessments of losses, which anticipated a 10–11% decline in output, underestimated the country’s greatly enhanced capacity to increase dry season production when required.

Changes in the composition of productive activity have been another factor in Bangladesh’s increased resilience to flooding; export-oriented garment manufacturing has expanded, and to date has been relatively flood-proof. Other developments have included relative financial stability in recent years; the growth of micro credit, enabling the poor to better withstand disaster shocks; and labour migration. Remittance flows have increased post-disaster – for instance, rising by 18% following the 1998 floods – providing those affected with a new form of coping mechanism.

Increased vulnerability in Malawi

In contrast, Malawi’s economy has become increasingly sensitive to climatic variability (drought, erratic rainfall, extremely high rainfall and related floods). This too reflects a complex of factors, including:

  • non-sustainable agricultural practice;
  • structural changes in agriculture, where a shift to smallholder production has not been accompanied by the establishing of a viable credit system, support in providing seeds and other inputs and a supportive marketing structure for smaller producers;
  • institutional weaknesses in the agricultural sector;
  • de-industrialisation, with the reintegration of South Africa into the regional economy;
  • political instability and problems of governance;
  • the short-term behaviour of aid donors, contributing to the volatility of public finances; and
  • the effects of HIV/AIDS on human resources.

The unfavourable effects of high rainfall on maize and tobacco in 2001 were recognised only following the harvest. By then many costly decisions, notably the sale of grain reserves by a cash-strapped government, had been made on the basis of poor advice from the IMF and others. When the food crisis came, it was widely assumed to have been either caused or exacerbated by a drought in 2002. However, the supposed drought was not reflected in meteorological reports.

The implications for relief and post-disaster reconstruction

Changes such as those in Bangladesh and Malawi imply that a relief programme that was highly successful a decade ago may no longer be appropriate. Food aid requirements can alter. New pockets of vulnerable groups can emerge, not least amongst the urban poor. The nature of community support networks can shift. The appropriate balance of channels for providing aid can also change, reflecting relative strengths and capacities of government and civil society. Low-income countries are most vulnerable to disaster shocks, financially and economically, when there are severe and growing problems of governance and weak fiscal and monetary management.

Post-disaster reconstruction needs serious contingency thinking prior to a potential event, and its implementation needs to be orchestrated so that it exploits risk reduction possibilities in rebuilding a country. This is a particularly critical moment of opportunity. A disaster creates a significantly heightened awareness of the need for risk reduction, but this interest rapidly fades as communities and economic systems are re-established and other more immediate priorities reassert themselves in ministries of finance and aid agencies.

There is a clear related need for greater longitudinal analysis of the impact of particular disaster events in order to help facilitate better understanding of the factors determining vulnerability, and how these can change over time. National or economy-wide disaster impacts, including total financial losses, should be reassessed 12–18 months after an event, as the fuller economic consequences only become apparent over time. Later, say after five or ten years, there should be a review of the longer-term consequences, including investment in disaster mitigation. Such exercises should be complemented by studies exploring impacts on affected households and communities. How have they fared, on both met and unmet needs arising as a consequence of a disaster? How could they be supported, both in reducing vulnerability to future events and recovering as rapidly as possible from them?

Reallocation of public finance

Public finance is widely seen as obscure, unexciting and the domain of a few experts. Yet it is a critical dimension in understanding the wider impacts of disasters. Post-disaster sourcing of finance can have a significant influence on both the short and longer-term impacts of a shock.

The behaviour of broad fiscal aggregates, total annual expenditure and revenue and the budgetary deficit, suggest misleadingly that disasters have little discernible impact in many countries. (The notable exception is many low-income sub-Saharan countries with a weak revenue basis and high aid dependence.) However, more disaggregated investigation of current and capital expenditure by sectors and sources of revenue shows how disasters can create significant budgetary pressures. The apparent insensitivity of fiscal aggregates in many countries in fact reflects successful post-disaster efforts to remain within the overall budgetary envelope established before the disaster, leaving levels of expenditure and fiscal deficits unchanged. In other words, disasters often result in widespread, but largely non-transparent, reallocations of resources.

Reallocations are typically poorly documented and cannot be easily quantified. The available evidence suggests that the brunt of financial reallocations appears to fall primarily on capital expenditure and in the social sectors. This may have a severe impact on the poor, and often on those worst affected by disaster. Donor agencies providing humanitarian support post-disaster need to be aware that existing welfare and benefit schemes and other pro-poor programmes may have been eroded, indirectly intensifying impacts on the poorest segments of society. The post-disaster diversion of resources can hamper longer-term efforts to reduce poverty – and, ultimately, vulnerability – and to achieve sustainable development. Decisions on post-disaster reallocations of budgetary resources are typically made with no system in place to protect priority areas of expenditure.

The available statistics similarly suggest that disasters have little impact on trends in total aid flows to disaster-affected countries. It is widely believed that the international community responds to disasters by increasing its aid, particularly in the form of emergency relief and food aid. However, some donors appear to respond to disaster crises by reallocating resources within projects and from existing projects, and by bringing forward commitments under existing multi-year country programmes and budget envelopes. Reflecting these practices, total aid commitments typically increase in the year of, or immediately following, a major disaster. Reallocations can be an appropriate way of quickly responding, in the sense that they involve a lesser administrative burden than the negotiation of fresh aid commitments. However, reallocated resources are typically not subsequently made good, with new commitments instead falling back after the crisis to reflect longer-term trends in aid.

After a disaster, there is often a substantial gap between projected and actual aid disbursement, reflecting management constraints – such as procedural difficulties, procurement delays and lack of local counterpart finance (problems that can also delay the disbursement of development aid). Rapid disbursement of food aid may be particularly important as even relatively short delays can prejudice post-disaster agricultural recovery and cause financial pressures.


A more complete assessment of the impact of disasters, exploring indirect and secondary effects as well as direct impacts, can contribute to ensuring that post-disaster response is effective and timely, enhancing the success of humanitarian response and reconstruction efforts. Analysis of the longer-term impacts of disasters and underlying determinants of vulnerability from an economic perspective helps to highlight a number of clear lessons for reducing hazard risk, and for making post-disaster assistance more appropriate.

Charlotte Benson ( and Edward Clay( are Senior Research Associates at the ODI.

This article is based on Benson and Clay, Understanding the Economic and Financial Impacts of Natural Disasters (Washington DC: World Bank, 2004). The study was undertaken on behalf of the World Bank’s Hazard Management Unit with the financial support of the UK Department for International Development’s Conflict and Humanitarian Aid Department (CHAD). Financial and professional support from these two organisations is gratefully acknowledged. The views expressed in this article and other outputs from the study listed below are those of the authors.

Further reading

C. Benson and E. J. Clay, The Impact of Drought on Sub-Saharan African Economies: A Preliminary Examination, World Bank Technical Paper 401 (Washington DC: World Bank, 1998).

C. Benson and E. J. Clay, Dominica: Natural Disasters and Economic Development in a Small Island State, Disaster Risk Management Working Paper Series No. 2 (Washington DC: World Bank, 2001),

C. Benson and E. J. Clay, Bangladesh: Disasters and Public Finance, Disaster Risk Management Working Paper Series No. 5 (Washington DC: World Bank, 2002),

C. Benson and E. J. Clay, Understanding the Economic and Financial Impacts of Natural Disasters, Disaster Risk Management Series No. 4 (Washington DC: World Bank, 2003).

E. Clay, L. Bohn, E. Blanco de Armas, S. Kabambe and H. Tchale, Climatic Variability, Economic Performance and the Uses of Climatic Forecasting in Malawi and Southern Africa, Disaster Risk Management Working Paper Series 7 (Washington DC: World Bank, 2003),