1 of 33

Business & Economic Resilience in Disasters

Alfredo Roa-Henriquez, Ph.D.

North Dakota State University

November 2, 2023

Providence Resilience Partnership

2 of 33

Small Business Profile at 2022

Source: Census and Small Business Administration

3 of 33

Small Business Profile at 2022

Source: Census and Small Business Administration

4 of 33

The Role of Critical Inputs in Disasters: Labor Disruptions

5 of 33

The Role of Critical Inputs in Disasters: Transportation�Disruptions

6 of 33

The Role of Critical Inputs in Disasters: Supply Chain�Disruptions

7 of 33

8 of 33

The Role of Critical Inputs in Disasters: Power Outages & Communication Outages

9 of 33

Key Concepts

Resilience, Mitigation, & BI

10 of 33

Resilience: Often refers to any action that reduces hazard losses
But there’s a perfectly good word for actions taken before the event – “mitigation”
Best use of “resilience” – actions taken after an event

can build up resilience capacity beforehand – it’s a process
(inventories, resource agreements, identify back-up locations)
but these tactics are not implemented until after the event

Can only prevent property damage before the event, but can reduce business interruption afterwards

BI begins when the disaster strikes & continues until recovered

Measured in terms of lost sales revenue, GDP, employment

Property Damage $

+

Business Interruption $

Disaster

Mitigation

(Pre-disaster)

Resilience

Key Distinctions: ‘Mitigation,’ ‘Resilience,’ & ‘BI’

11 of 33

Economic Resilience

Static Resilience:

General Definition: Ability of a system to maintain function when shocked.
Econ Definition: Efficient use of remaining resources at a given point in time to produce as much as possible.

Dynamic Resilience:

General: Ability & speed of a system to recover.
Economic: Efficient use of resources over time for investment in repair and reconstruction, including expediting the process & adapting to change.

Resilience essentially synonymous with efficient business continuity

Metric: losses avoided by using a resilience tactic as % of potential losses without it

12 of 33

Why BI Matters

September 11 World Trade Center Attacks

- property damage (PD): $25 Billion

- business interruption (BI): $100 Billion

Hurricane Katrina

- PD: $75B

- BI: >$100B

ShakeOut San Andreas Fault Earthquake Simulation

- PD: $100B

- BI: $68B

13 of 33

Resilience Tactic	Definition (Activities Involved)
Conservation	Maintaining intended production using lower amounts of an input or inputs
Resource Isolation	Modifying a portion of business operations to run without a critical input
Input Substitution	Replacing a production input in short supply with another
Inventories	Continuing business operations using emergency and ordinary stockpiles
Excess Capacity	Using idle plant or equipment in place of a damaged ones
Relocation	Moving some or all of the business activity to a new location
Management Effectiveness	Improving the efficiency of business operations in the aftermath of a disaster
Import Substitution	Importing needed production inputs when not available from local suppliers
Technological Change	Improvising the production process without requiring a major investment
Resource Pooling/Sharing	Recontracting, selective exchange of resources, creating new partnerships
Production Recapture	Making up for lost production by working overtime or extra shifts.

Resilience Tactics (Actions)

14 of 33

Survey

Design of Two Large-area Surveys to Empirically Measure Economic Resilience

15 of 33

Two random samples: Superstorm Sandy (NY and NJ) and Hurricane Harvey (TX)

Administered 60- and 13- months post-disaster for Sandy and Harvey, respectively

Stratified by firm size and property damage (oversampled PD and firms >10 FTE employees)

62% of firms in sample had both BI + PD

Median firm of 30 FTE employees, mean 2,696 FTE employees (max of 90K FTEs)

Sample size: 249 firms using 722 resilience tactics

Survey Methodology

16 of 33

Designed to Estimate Resilience Metrics (RM) & (BCR)

17 of 33

Results

Empirical Results from Surveys

18 of 33

Results

19 of 33

Results: PD versus BI

Superstorm Sandy (NY/NJ)	Property Damage Median	Business Interruption Median	Ratio
Among only firms physically damaged	$100,000	$180,000	180%
All firms	$25,000	$100,000	400%

Hurricane Harvey (TX)	Property Damage Median	Business Interruption Median	Ratio
Among only firms physically damaged	$56,000	$725,000	1,294%
All firms	$5,000	$575,000	11,500%

20 of 33

	Agriculture, Mining & Construction	Transportation, Communication & Utilities	Manufacturing	Wholesale & Retail Trade	Finance, Insurance & Real Estate	Service Sectors	All Sectors TOTAL
Business Interruption & Property Damage	3.2	4.7	4.1	2.8	3.1	3.9	3.6
Business Interruption Only	2.4	2.6	1.5	2.1	1.8	2.3	2.2

Results: How Many Tactics Did Firms Use?

Compared to firms observing only BI, firms observing Property Damage used, on average, 1.4 additional tactics

Generally, capital-intensive sectors such as transportation, communications, utilities and manufacturing used the greatest number of tactics

21 of 33

Results: Overall Metrics (BCR and RM)

Overall, firms avoided $4.57 for every dollar spent on resilience tactics

Overall, firms avoided 6% of max potential losses, considerably lower RM results for higher-severity Harvey results (2.4%) than Sandy results (34.1%)

22 of 33

BRC Decision-support Software

Provisional Patent No. 63/314,650

23 of 33

Policy Implications

What does this all mean? And, for whom?

24 of 33

Policy Implications

First wide scale attempt to measure the cost effectiveness of resilience actions
Results help tailor resilience recommendations; recommendations have a theoretical foundation in production theory
Public Sector: DHS/FEMA, SBA, US Coast Guard, Economic Development Administration, State/County Office of Emergency Management, National Guard
Private Firms: Predominantly SMEs without resilience officers/resilience analytics
Insurance Industry: Insurance/Reinsurance
Natural and human-induced hazards are increasing in economic disruptions 🡪 mitigation has not kept pace
BI > property damage

Downtown Houston after Harvey. Photo credit: The Texas Tribune

25 of 33

Thank you

26 of 33

Supplemental slides

27 of 33

28 of 33

Temporal Experiment: Numeracy & Evacuation Time

Highest Information Treatments (3 & 4)
Among professional subjects
After controlling for information received and random assignment
Numeracy resulted in an additional 8.79 hours (p<0.1) given to coastal populations to evacuate
Numerate managers are more than 1.7 times (p<0.1) more likely to call for early evacuation

Treatment 4 Numerate Managers (43.69 Hours)

Treatment 4 Less Numerate Managers (34.9 Hours)

Treatment 3 Less Numerate Managers (31.22 Hours)

Treatment 3 Numerate Managers (40.01 Hours)

29 of 33

Spatial Experiment: Population Evacuated

Subjects asked to select zones to evacuate based on probabilistic flood/inundation maps
Established Hurricane Evacuation Zones (Houston-Galveston Area Council)

Treatment 4 Numerate Managers (43.69 Hours)

Treatment 4 Less Numerate Managers (34.9 Hours)

Treatment 3 Less Numerate Managers (31.22 Hours)

Treatment 3 Numerate Managers (40.01 Hours)

30 of 33

Spatial Experiment: Population Evacuated

Inundation projections mirror the hydrodynamic models given to emergency managers

31 of 33

32 of 33

Spatial Experiment: Population Evacuated

Among all treatments
Professional subjects & Student subjects
On average, less numerate professionals evacuated almost 50K more people than numerate managers
On average, less numerate students evacuated almost 85K more people than numerate students

Treatment 4 Numerate Managers (43.69 Hours)

Treatment 4 Less Numerate Managers (34.9 Hours)

Treatment 3 Less Numerate Managers (31.22 Hours)

Treatment 3 Numerate Managers (40.01 Hours)

33 of 33

Spatial Experiment: Population Evacuated

Post-experiment Survey Question
Did you evacuate any zones even though you thought the likelihood of human exposure to flooding was low?
Answered yes:
Professionals evacuated 168K more people
Students evacuated 123K more people

Treatment 4 Numerate Managers (43.69 Hours)

Treatment 4 Less Numerate Managers (34.9 Hours)

Treatment 3 Less Numerate Managers (31.22 Hours)

Treatment 3 Numerate Managers (40.01 Hours)