Disaster recovery – evidence from 100 natural disasters

This paper examines the factors that affect the recovery process following 100 major natural disasters. The study aimed to measure the speed and quality of recovery and to begin to answer the question of why some places recover faster and better than others. This worldwide study, conducted over a 15-year period, covers four natural hazards: major floods, storms, earthquakes, and tsunamis.

Methods
Methodology abstract

A number of indicators were considered to measure the speed of recovery, based on a “return to normality”.[i] Measures of the quality of recovery were based on the concept of “building back better” and a change in resilience.[ii] The speed of recovery was measured separately for ‘economy’ and ‘society’, based on the time taken in months to return to ‘normal’ to comply with the conditions listed below. The quality of recovery was measured in terms of changes in economic growth, safety and amenity. Improvements in these factors result in improved resilience, while a deterioration results in impaired recovery. They are each measured on a five-point scale: from 1, much worse to 5, much better.

Dependent variables - Factors measuring speed and quality of recovery

Speed (months)
Quality (scale 1-5)

Economy
Livelihoods ≥90% back in work; Economy ≥90% businesses back in operation; Services ≥90% telecoms, water, power restored
Economic growth

Society
Access fully restored; Temporary shelter completely cleared; Permanent housing ≥90% rehoused; Schooling ≥90% children in full-time school
Safety; Amenity

Independent variables - Factors that might affect speed and quality of recovery

The independent variables that we imagined might affect the speed and quality of recovery include both "givens", including antecedent factors and the hazard load, and manageable factors, that include resourcing, organisation and management. (Table 4)

Givens
Manageable Factors

ANTECEDENTS
RESOURCING

Economy Type (event year)
Self vs external finance

Income Group (2018)
Government aid

GDP Growth (event year)
Non-Governmental aid

GDP per Capita (event year)
Insured loss

GNI per Capita (event year)
Insurance penetration

IHDI Value (2017)
Speed of funding delivery

Gini Coefficient (2018)
Adequacy of funding

Corrupt. Index (2018)
ORGANISATION

Non-Life Insurance Penetration (event year)
State of Democracy

LOAD
Disaster Management Authority

Damage Severity (% damage buildings, infrastructure)
Preparedness

Economic Loss
Public Participation

Fatalities
MANAGEMENT

Displaced
Scientific basis of decision making

Affected Population
Decision Quality

Experience of previous disasters

Data collection

Various data collection methods were used in a complementary way to provide both quantifiable and qualitative data.

Remote sensing (6 cases)

Interviews (15 cases)

Workshops (5 cases)

Surveys (7 cases)

EM-DAT international database (103 cases)

A cautionary note

In interpreting the information from published sources a number of measures relied on the subjective judgement of the author. Specifically this applies to the assessment of the speed and quality of recovery, in those cases where definitive information was unavailable, and to an assessment of the quality of decision making by the authorities responsible for disaster management.

The sources for each of the variables tabled in Appendix.  Confidence level given for estimates of speed and quality and resourcing.

[i] Quarantelli, E.L. (1999), "What Is a Disaster: Perspectives on the Question", Disaster Prevention and Management, Vol. 8 No. 5, pp. 370-452. Routledge

[ii] Kim K., Olshansky R. B., (2014) The theory and practice of building back better. Journal of the American Planning Association 80 (4), 289-292