Journal of Contingencies and Crisis Management

Volume 19 Number 3 September 2011

Factors Affecting Business Recovery Immediately after Hurricane Katrina Christy M. Corey and Elizabeth A. Deitch Department of Management, University of New Orleans, 2000 Lakeshore Dr., New Orleans, LA 70148, USA. E-mails: [email protected], [email protected] Factors contributing to business recovery 6–8 months after Hurricane Katrina were examined. Managers from 183 surviving organizations in the Greater New Orleans area rated the levels of storm preparation, amount of storm damage, severity of storm-related problems and organizational performance. Factors under management’s control such as having an emergency response plan, storm preparation and effective staff communication had no real impact on organizational performance. Significant predictors with a negative impact on organizational performance included variables such as storm damage and post-disaster problems. Complications arising from extreme population dislocation, specifically loss of customer base and staffing issues, had the greatest impact on organizational performance. The implications for disaster preparation and management are discussed.

W

hen a community experiences widespread disaster, one extremely important factor in the recovery efforts is the ability of local businesses to survive and thrive. Business survival is a vital part of community recovery after a natural disaster, as businesses provide jobs, goods and services, and tax dollars (Cochrane, 1992). No disaster in US history has been more destructive than Hurricane Katrina, which landed in Southeast Louisiana on 29 August 2005. The hurricane itself wrought major damage, and levee failures in New Orleans resulted in massive flooding of the city. Property damage has been estimated at US$81 billion (Knabb, Rhome, & Brown, 2006), and nearly one million residents were displaced (Institute for Southern Studies, 2009). Business owners and managers in New Orleans faced unprecedented difficulties in resuming operations, with the city essentially ‘closed’ for some time due to flooding. Residents of New Orleans were prevented from returning to the city for nearly a month after the storm, as Hurricane Rita on the heels of Katrina impeded attempts to drain and secure the city. Thus, local business owners and managers were strongly impacted by not just the physical damage of the hurricane and its aftermath but also by the loss of operations while the city was closed, employees were dispersed across the country and their customer base was dislocated. Even when the city was reopened, it was still experiencing the impact of a critical infrastructure breakdown of catastrophic proportions (Boin & McConnell, 2007). Gas and electricity were not available in most of New Orleans, city water was not potable and communications were spotty. Thus, business owners and

& 2011 Blackwell Publishing Ltd.

managers faced a monumental task to get back on their feet. In this investigation, we were interested in the factors that contributed to recovery and short-term organizational performance 6–8 months following Hurricane Katrina. The goal of this study was not to examine which factors predict post-Katrina business survival or death. Rather, we focus on those businesses that did survive Katrina and consider variables that predict current organizational performance that we define relative to performance before Hurricane Katrina. Individual business performance after disasters is relatively under-studied. What little research has focused on private businesses following natural disasters has often been at a very aggregated level (e.g., Cohen, 1993; Eguchi, Goltz, Taylor, Chang, Flores, Johnson, Seligson, & Blais, 1998; West & Lenze, 1994). In contrast, this paper uses data that were compiled from New Orleans area business owners’ and managers’ detailed accounts concerning their organization’s recovery progress and the pre- and post-disaster events leading up to the spring of 2006. As stated previously, the primary criterion variable of interest in the current study is organizational performance. Clearly, businesses included in this investigation had overcome the biggest obstacle, which is survival in a post-disaster economy. Surviving organizations can be sorted into three performance status categories; performance may be worse, the same or better compared with an organization’s preKatrina levels. In addition, within each performance status group, organizations have varying amounts of performance losses and gains. We consider both indicators of organizational performance in the current study.

DOI: 10.1111/j.1468-5973.2011.00642.x

170

1. Factors affecting post-Katrina organizational performance Some studies have attempted to determine features that predict the success of recovery for individual businesses post-disaster, but the results have been inconsistent regarding what matters the most. Chang and Falit-Baiamonte (2003) concluded that business sector, size and building occupancy tenure were the best indicators of business vulnerability after the 2001 Nisqually earthquake. In contrast, Tierney (1997a) found that the amount of operations disruption, size of the business and general economic decline in the area were most predictive of success after the Northridge Earthquake. Finally, Webb, Tierney, and Dahlhamer (2002) found that different features predicted successful recovery in Miami after Hurricane Andrew than in Santa Cruz after the Loma Prieta earthquake. The differences in these findings are likely due to what type of disaster was experienced and how widespread the damage was; each disaster is unique and an earthquake will not cause precisely the same problems as a hurricane. The sheer scale of the population dislocation for Hurricane Katrina in the New Orleans area was unprecedented; approximately 25% of the population had not returned 6 months later at the time of data collection in the spring of 2006 (US Census Bureau, 2009). Thus, for this particular disaster, population-related issues may play a uniquely significant role in business recovery. In the current study, four broad categories of potential factors influencing organizational performance were investigated. First, we examined organizational characteristics, including the age and size of the business and the industry in which it operates. Next, we assessed pre-disaster preparation, in terms of the existence of an emergency plan, insurance purchase and immediate pre-storm hazard mitigation measures. Then, we assessed the amount of physical damage incurred during the storm and subsequent flooding. Finally, we examined post-storm problems such as the disruption of communications and supply lines, staff and customer loss, provision of public services and problems receiving government aid. For each of the broad categories, we review previous research pertaining to these areas and offer hypotheses concerning their influence on organizational performance.

1.1. Business characteristics Business size has often been found to predict recovery after a natural disaster. Tierney (1997a) found that small businesses were more likely to report being worse off after the Northridge earthquake than their larger counterparts, and Kroll, Landis, Shen, and Stryker (1990) reported a similar finding for Oakland businesses after the Loma Prieta earthquake. Zhang, Lindell, and Prater (2009) point out that larger firms are more likely to have multiple locations, be located in newer facilities, have more financial and political capital and be better able to afford insurance for disaster recovery. Large firms may have been especially better equipped to deal with the population dispersion associated with Hurricane

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

Christy M. Corey and Elizabeth A. Deitch Katrina by having other locations at which employees could work or by having enough resources to continue paying employees while business was disrupted. The industry in which a business operates can have a large impact on how well it weathers a natural disaster. Disasters consistently create increased demand for some industry sectors. Kroll et al. (1990) found that construction firms in Oakland and Santa Cruz reported increases in business following the Loma Prieta earthquake. Cole, Corbett, and McCullough (2005) discovered that while retail and tourism industries reported substantial losses after Hurricane Andrew, building contractors saw demand for their services grow. Consistent with this trend, Dahlhamer and Tierney (1998) reported that the largest proportion of recovered businesses after the Northridge earthquake was found in the manufacturing and construction sector. Thus, we expected to similarly find that construction-related businesses would have improved business post-Katrina. Wholesale/retail firms have seemed to be the most vulnerable, in previous studies of post-disaster business recovery. Webb et al. (2002) found that retail and wholesale businesses in Florida were less likely to recover from Hurricane Andrew than other types of organizations. Chang and Falit-Baiamonte (2003) similarly found that a far greater percentage of retail businesses suffered major economic loss following the 2001 Nisqually earthquake in Seattle than did business in other sectors. And Boarnet (1998) also concluded that the highest percentage of businesses experiencing loss after the Northridge earthquake was in the retail sector. However, Runyan (2006), in post-Katrina interviews with small business owners along the Gulf Coast, reported that those businesses that were able to reopen soon after the storm saw large sales increases, due to two factors. First, there were far fewer competitors in the area than before the storm. Second, those towns had seen large increases in population concentrations due to being flooded with evacuees from New Orleans. While we would expect a similar effect for surviving New Orleans area businesses in terms of the lack of competition, the population migration would be to the detriment of our sample. Thus, we are unsure of the net effect on retail businesses and so refrain from hypothesizing regarding their performance. It is unclear whether business age has an impact on disaster recovery. The general business literature suggests that new businesses are more likely to fail than established ones (e.g., Aldrich & Auster, 1986). However, Webb et al. (2002) found no significant effect of age on recovery following the Loma Prieta earthquake, and although they did find an age effect in Florida after Hurricane Andrew, it was not in the expected direction – new firms actually recovered better than older businesses. They suggest that this may be due to inertia in older businesses, and a failure to adequately respond to changing environmental contingencies. In sum, regarding business characteristics, we expected that larger businesses and those in the construction/ manufacturing sector would report better recovery from

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery Hurricane Katrina. We could not make any prediction regarding the impact of business age, however, given the inconsistency of previous results. H1: Organizational size will be positively related to the amount of storm preparation, staff communication and organizational performance. It will be inversely related to direct storm damage and post-storm problems. H2: Organizations in construction sectors will have higher levels organizational performance compared with businesses in other industry sectors.

1.2. Pre-disaster preparation One might expect that those businesses that had taken the time to map out an emergency plan before the threat of Katrina would indeed be better prepared to withstand the disaster and recover afterwards. However, several studies of disaster impacts on businesses have failed to find any significant association between disaster preparation and business recovery (e.g., Chang & Falit-Baiamonte, 2003; Dahlhamer & Reshaur, 1996; Webb et al., 2002). The mere existence of a disaster plan does not guarantee that organizational members executed the plan successfully including all of its details at the appropriate time before a disaster. In addition, the scale of a given disaster can potentially overshadow the intended effects of disaster preparation. In the specific case of Hurricane Katrina, even though the storm’s approach was known for days, business owners still did not anticipate the size and scope of the devastation to come (Runyan, 2006). In Runyan’s interviews with business owners on the Gulf Coast, he found that owners reported pre-storm assumptions that this hurricane would be just like others they had lived through many times before, anticipating far less extensive damage than actually occurred. Boin and McConnell (2007) note that these attitudes are one of the most common barriers to organizational resilience, that officials have a tendency to think catastrophic infrastructure breakdowns ‘can’t happen here’. Furthermore, Zhang et al. (2009) point out that most businesses’ emergency plans, if they do exist, have a rather short-term focus. Attempts at hazard mitigation are typically focused on employees’ life safety and immediate damage prevention more than on the long-term continuity of business operations. Zolin and Kropp (2007) recommend that proactive organizations need not just a ‘disaster preparedness plan’, but also a ‘business continuity plan’ including protections for vital data and resources. Thus, in addition to the existence of an emergency plan, we examined some specific plan features. One short-term measure that may have helped businesses survive was physically preparing the business by boarding it up against storm damage. Another feature that would seem to be useful in surviving the immediate aftermath of a disaster like Katrina, with such major population dislocation, was the collection of emergency contact information for employees (e.g., alternative non-work email accounts, cell-phone numbers, out-of-

& 2011 Blackwell Publishing Ltd.

171 area relatives, etc.). Phone numbers and email accounts based in the New Orleans area were all non-functional during and after the storm. Bond (2007), drawing lessons from disaster medical assistance teams after Katrina, points out that a useful communication plan needs to contain multiple levels of redundancy. This need is illustrated by Zolin and Kropp’s (2007) New Orleans business interviewees who did not have non-local avenues of contact for their workforce, and reported experiencing severe difficulties even locating their employees. As one of their respondents stated, ‘anyone with a 504 [New Orleans] area code cell phone could not be contacted, and if anyone got a new phone number, no one knew what their phone number was’ (p. 187). Organizations had to resort to novel ways of attempting contact, establishing toll-free call-in numbers, check-in websites, even physically sending representatives to shelters with signs (Edwards, 2006). Certainly, those organizations that had planned some way to stay in touch with employees that was not dependent on local New Orleans infrastructure (preferably multiple ways) would be expected to have fewer troubles with poststorm staffing than those that did not. We also examined more long-term planning for disaster. As New Orleans was essentially ‘closed’ for nearly a month, arranging to function from afar should have been very helpful. This involves backing up important data and files someplace other than on-site. Organizations that arranged for an alternative out-of-area location for conducting business should also be better positioned for recovery. And finally, the purchase of adequate insurance would provide resources to help recover after the storm. Previous studies have found that a very small proportion of businesses tend to carry the relevant types of insurance for natural disasters. For example, Tierney (1997b) found that only 8% of businesses in Des Moines had flood insurance at the time of the 1993 floods, and only 20% of Los Angeles businesses carried earthquake insurance at the time of the 1994 Northridge earthquake. Nonetheless, even if few businesses did take long-term hazard mitigation steps, we would expect those businesses to recover more easily after Hurricane Katrina. H3: Organizations with an emergency response plan will have higher mean levels of storm preparation, staff communication and organizational performance compared with those organizations with no plan. Having a plan should result in lower mean levels of direct-storm damage and storm-related problems. H4: More storm preparation and better staff communication will positively predict organizational performance.

1.3. Physical damage incurred Several studies have found that the greater the amount of physical damage to a business’ structures, the more difficult their recovery. Webb et al. (2002) found that Santa Cruz businesses that experienced more disruptive physical damage as a result of the Loma Prieta earthquake were less likely

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

172 to recover in the long term. Tierney (1997a) reported a similar result for Los Angeles businesses after the Northridge earthquake. Businesses that experience more physical damage generally take longer to reopen than those that do not. There were several potential sources of damage for businesses in the New Orleans area during and after Hurricane Katrina. First, the wind and rain from the hurricane itself harmed many buildings, downing trees, damaging roofs, breaking windows and allowing rainwater to penetrate. Then, the levee failures after the storm had passed resulted in widespread flooding, damaging or even completely destroying many buildings that had survived the storm’s passage. Thus, many business owners returned to find multiple types of damage to their structures, inventory and equipment. We assessed all these types of damage, expecting to find that businesses experiencing more physical damage would be less likely to recover from the disaster.

1.4. Post-disaster difficulties Hurricane Katrina and its aftermath had features of both natural and social disasters (Baade, Baumann, & Matheson, 2007); both infrastructure and social structure broke down in the wake of the storm. Every aspect of life in the New Orleans area was altered. Thus, the potential problems experienced by businesses in the area were many, and likely had cumulative effects on business recovery. First of all, basic ‘lifeline’ services were non-functional for varying lengths of time across the region. Most businesses were without power, natural gas or clean water for extended periods of time, ranging from 2 to 8 weeks. Many roads were impassable even after the flooding was drained, due to debris and extensive damage from sitting underwater for weeks. Communications were massively disrupted, partly due to the loss of electricity, but also from downed telephone lines, destroyed cell towers and flooded electronic equipment. Other services, such as postal delivery and garbage pickup, could not be carried out due to a lack of personnel. The loss of these types of services may be an even larger barrier to rapid business recovery than disaster damage. Tierney (1997b) found that utility disruption after the 1993 Midwest floods was a more important cause of business closure than direct damage. Even though only 15% of Des Moines businesses experienced flood damage, 42% of them were forced to close for some period of time due to lack of water service, sewer service, electricity or phone. Tierney (1997a) found that businesses located in areas where damage and disruption were widespread after the Northridge quake had more problems getting back on their feet, noting the influence of larger neighbourhood and ‘ecological’ factors in business recovery. Clearly, even if a business itself is unharmed by a disaster, damage to offsite lifelines will still have a negative impact on recovery. The cost of losing these basic services for businesses can be very large. Rose, Benavides, Chang, Szczesniak, and Lim

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

Christy M. Corey and Elizabeth A. Deitch (1997) used computer models to estimate the productivity loss that would be associated with a loss of electricity for the Memphis area in a major earthquake. Using a recovery timeframe in which 83% of the population saw restoration of power within 2 weeks, and 97.5% within 5 weeks (a far shorter timeframe than the post-Katrina power restoration), they estimated the regional economic loss over the recovery period to be as much as 7% of the gross regional product. And that was for electricity disruption alone. The amount of economic loss due to all services being cut over such an extended period of time after Hurricane Katrina must be astronomical. We also examined perceptions of government response and aid as government responses to disaster in the form of community aid and business loans have been shown to impact the recovery of businesses and neighbourhoods (Zhang et al., 2009). However, the immediate government response to Hurricane Katrina at all levels, local, state and Federal, has generally been viewed as inappropriate and uncoordinated (Olejarski & Garnett, 2010), and Runyan’s (2006) interviewees specifically blamed FEMA for actually impeding their attempts at recovery (e.g., by poaching workers from their rebuilding sites). Area businesses commonly faced cash flow problems after Hurricane Katrina, as no funds were coming in, but routine payments still had to be made (Runyan, 2006; Zolin & Kropp, 2007). Thus, many hoped to use external aid to help cover this (hopefully temporary) shortfall in funds. The results in this area from studies of other disasters are mixed, however, as Webb et al. (2002) found no significant relationship between the use of post-disaster aid and recovery outcomes after Hurricane Andrew and the Loma Prieta earthquake. Furthermore, in this case, many businesses were unable to even fulfil application requirements for aid from organizations like the Small Business Administration (SBA) due to their storm damage. Runyon’s (2006) respondents complained that the SBA was asking them for 3 years of financial statements and tax returns, despite the fact that most of these documents had been ruined by the very disaster that led them to seek aid. Also, there is ample documentation that the process for receiving an SBA loan after Hurricane Katrina was far too slow to make a real difference even when documentation had survived or could be obtained from the IRS, with fewer than 1% of the 66,819 emergency loan applications sent to the SBA being approved by mid-October (Loten, 2005). Thus, it is unlikely that direct financial aid from the government was a substantial factor in business recovery postKatrina. We also must consider that the fortunes of area businesses are tied to those upstream in the value chain. Even businesses that have successfully weathered disaster may fail due to the failure of their suppliers or difficulties transporting supplies where needed. Consequently, failing suppliers can cause a ‘domino effect’ to other businesses (Zhang et al., 2009). Therefore, we expect that businesses that report problems with supply lines will have more recovery difficulties.

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery Finally, staff and customer loss may have large impacts on post-disaster performance. A distinguishing feature of Hurricane Katrina was the astounding amount of population dislocation it created. As of July 2009, the greater New Orleans (GNO) area still had only about 90% of the population compared with pre-Katrina levels (US Census Bureau, 2009). Population decrements cause problems for businesses in two ways. First, the local population is a critical source of customers for many businesses. Thus, customer loss after dislocation may be a major contributing factor to business failure post-disaster (Tierney, 1997a). Second, the local population is a valuable labour source. When populations are displaced, area businesses may lose existing employees and have difficulty in recruiting new hires (Zhang et al., 2009). While there may be little that businesses can do to hold the general population in an area to maintain a customer base, they can take actions to retain their employees starting with effective post-disaster communication. Thus, we also examined the extent to which organizations were able to maintain contact with their employees in the immediate aftermath of the storm. We would expect that organizations that communicated effectively with employees would retain a greater proportion of their staff, and that such staff retention would be a positive influence on recovery. H5: The amount of direct-storm damage and post-storm problems will negatively predict organizational performance. H6: Organizations with greater levels of staff loss will have lower organizational performance. H7: Population-related problems, including staffing issues and loss of customer base, will have the greatest impact on organizational performance. In sum, there are a multitude of factors that can influence the extent to which a business is able to recover following a disaster. We examined several of these factors in relation to Hurricane Katrina, a rather unique disaster in terms of the dual nature of the natural-disaster impact (both hurricane and later flooding), the social chaos that ensued and the amount of population dislocation involved. We expected to find that those organizations that recovered best would be larger firms in the construction/manufacturing sector with dispersed markets, those that engaged in more long-term preparation, those with the least damage and those with the fewest post-disaster complications.

2. Method 2.1. Participants and procedure Managers from 183 businesses in the GNO area participated in the study by completing a 30-question survey; they were recruited using the convenience sampling methodology. To be eligible, businesses had to be situated in one of the five parishes comprising the GNO area. The business also needed to exist before Hurricane Katrina and still be

& 2011 Blackwell Publishing Ltd.

173 operating at the time of data collection. Finally, managers needed to be familiar with the history of the organization during the specified time period in order to complete the survey. Data were collected in Spring 2006, 6–8 months post-Katrina. The sample of organizations in the GNO area is uniquely diverse. It includes mainstays of the New Orleans business community such as Cafe´ du Monde and P&J Oyster Supply, which are 144 and 133 years old, respectively. Many businesses had difficult experiences evacuating from and returning to the GNO area. Approximately 37% of organizations in the current sample operated in an alternate, outof-town location until they were allowed to return to the city or until necessary repairs had been made and utilities had been restored. Some owners and managers returned to find that their organizations had been looted (21.3%), flooded (32.2%) or even occupied by the military or law enforcement (15.8%). Managers from those organizations that had flood water damage reported flood levels anywhere from 1 inch to 30 feet. Only 64% of organizations in the sample had flood insurance; 68% had business interruption insurance and 38% had business recovery insurance.

2.2. Measures 2.2.1. Business characteristics Organizational size was operationalized as a two-group categorical variable with large and small groups. Small organizations were defined as companies that employed o50 people, and large organizations employed 50 or more individuals. To measure the age of an organization, managers reported the number of years that the business had been in operation. 2.2.2. Emergency response plan Participants indicated whether the organization had an emergency response plan in place before Hurricane Katrina. Responses were indicated in a dichotomous yes/no format. 2.2.3. Storm preparation Efforts to prepare for Hurricane Katrina were rated across the six areas of activity. Short-term planning measures included preparing the business building by boarding up, creating an emergency communication plan, collecting emergency out-of-area contact information for employees and collecting alternative non-work email accounts for employees. Long-term planning included backing up electronic data somewhere other than on-site location and arranging alternative out-of-area location for doing business. Managers indicated whether the organization had performed each of the areas of preparation, as well as anything in addition to those listed. A summary measure of amount of storm preparation was computed by summing up the total number of actions taken in preparation for Hurricane Katrina. The number of preparation scores ranges from 0 to 7.

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

174 2.2.4. Communication Post-Katrina communication-related issues were measured in two ways. First, using a yes/no response format, participants indicated whether communication between management and staff was a problem after the storm. Next, they used a checklist to indicate which of the following actions their organization tried that worked in terms of communicating with staff members: used emergency contact information including email and phone, used company website to post messages to employees, company website allowed employees to check in and update contact information, posted messages on local radio or local television stations, and posted messages on staff-finder websites other than your company’s webpage. Finally, managers noted anything their organization did in addition to those activities listed. A summary measure of the total forms of communication used was computed by summing across the items that worked for each organization. Scores on this variable ranged from 0 to 6. 2.2.5. Direct storm damage Damage sustained to a business resulting from Hurricane Katrina was assessed in eight areas: physical exterior of the building, physical interior of the building, electronics and computers, important files/paperwork, customer files/ records, accounting records, electronic data and inventory/supplies. Participants rated the extent of damage on a five-point summated rating scale ranging from ‘1 ¼ Unharmed’ to ‘5 ¼ Totally destroyed’. Higher scores indicate greater levels of damage. Although factor analysis showed that these eight areas of damage were related to one factor (a ¼ .93), we sorted these variables into three damage categories so that we could explore whether certain areas of damage were differentially related to the main study variables. Respondents’ scores on the first two items were averaged to create a variable representing physical damage to the place of business. The five items related to loss of electronic equipment and files were aggregated into one variable. Finally, the loss of inventory and supplies was tested alone. Higher scores represent greater levels of damage. 2.2.6. Post-storm problems Participants rated the degree of problems that occurred post-Katrina. Problem areas included loss of staff, loss of customer base, recruitment of new staff, reliability of suppliers, insurance claims and payments, availability of federal assistance (e.g., small business loans), unreliability of postal service and infrequent garbage pick-ups. Responses were made on a five-point summated rating scale ranging from ‘1 ¼ Not a problem’ to ‘5 ¼ Very problematic’. Higher scores indicate more problematic areas. 2.2.7. Staff loss Participants reported the percentage of staff loss for their organization. Observed scores on this variable ranged from 0% to 100%.

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

Christy M. Corey and Elizabeth A. Deitch 2.2.8. Organizational performance Organizational performance was operationalized as both a categorical and a continuous variable. First, managers indicated whether their organization’s performance was better, worse or the same compared with pre-Katrina performance levels. This information was used as a categorical variable representing the performance status of the business. Next, managers estimated the percentage of the increase or decrease in business since Katrina. Those reporting no performance change scored as 0. This percentage increase/decrease was used as the continuous dependent variable for analysis of variance (ANOVA) and regression analyses.

3. Results There are many factors that had a profound impact on business recovery that occurred before and after the storm. Pre-storm preparation gave way to storm response to the complications presented by Katrina. Later, the aftermath left many organizations with tremendous rates of staff loss and increasing labour shortages in the surrounding community. The remainder of this section will address the role of these and other various factors in contributing to organizational performance either directly or indirectly. First, we will examine the influence of business characteristics on the main study variables. Next, the role of an emergency response plan will be assessed in terms of the potential advantage that it confers to an organization to overcome obstacles created by a disaster. Finally, an assessment of the predictors of post-disaster organization performance will be discussed.

3.1. Descriptives Descriptive statistics for the main study variables are presented in Table 1. The age of organizations in the sample ranged from 1 to 144 years, averaging 26.71 years. The findings in the current study are consistent with previous research, which has found inconsistent effects of age on disaster recovery. Here, age was correlated with communication variables including problems communicating with staff (r ¼.16; po.05) and total forms of communication used (r ¼.15; po.05). Older organizations did not have significantly greater organizational performance. In the area of pre-storm preparation, respondents indicated different forms of preparation that they engaged in before evacuating the city of New Orleans. Some organizations focused on short-term preparation measures; they prepared their building(s) by boarding up (46%), created an emergency communication plan (42%), collected emergency out-of-area contact information for employees (38%) and collected alternative non-work email accounts for employees (12%). The area of short-term preparation used the least turned out to be one of the more important tasks that New Orleans business owners and managers should have completed, as many organizations lost access to their internet

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery

175

Table 1. Descriptive Statistics and Intercorrelations for the Main Study Variables Mean SD 1. Age 2. Storm preparation 3. Communication problems 4. Forms of communication 5. Physical damage 6. Damage to electronics/files 7. Loss of inventory 8. Staffing issues 9. Lost customer base 10. % Staff loss 11. Org performance

1

26.71 27.76 – 1.97 1.26 .05 .68 .47 .16* 1.82 1.07 .15* 2.68 1.18 .01 1.92 1.22 .05 2.69 1.60 .02 2.85 1.44 .06 2.76 1.41 .04 32.32 32.47 .05 4.70 67.44 .02

2

3

4

5

6

7

8

9

10

11

– .07 .02 .02 .04 .22** .00 .15* .14

– .03 – .07 .71* – .19* .62** .69** – .16* .16* .14 .26** – .10 .20** .19* .27** .19* – .06 .14 .13 .35** .61** .26** – .08 .22* .17* .18* .29** .47** .30** –

– .42** .07 .08 .08 .08 .08 .00 .07 .06

Note: n ¼ 172. *p  .05; **p  .01.

servers and employees’ access to their work email ceased. Long-term preparation efforts included backing up electronic data somewhere other than an on-site location (62%) and arranging alternative out-of-area locations for doing business (26%). Because there were no significant differences between short- and long-term planning across the main study variables, we aggregated across all areas of storm preparation to create one variable representing the total amount of storm preparation activities. Another area of preparation that we considered included whether businesses had various types of insurance policies including flood, business interruption and recovery in place before Katrina. One-way ANOVA showed no differences in organizational performance between insured and non-insured businesses. In the area of communication, 68% of respondents reported that their organization had problems with communication between management and staff members after the storm. Respondents indicated all of the communication methods they used that worked when attempting to contact staff members and, on average, organizations in this sample successfully used two forms of communication. Most organizations used emergency contact information to call or email employees (85%). Approximately 30% of businesses posted messages to employees on their company website. Company websites that allowed employees to check in and update their contact information were utilized by 24% of organizations and 11% ran radio ads in areas where their staff members were likely to be. Regarding direct damage resulting from Hurricane Katrina, the most damage was sustained to inventory and supplies (M ¼ 2.69; SD ¼ 1.60), followed by damage to the physical structure (M ¼ 2.68; SD ¼ 1.18) and damage to electronics and files (M ¼ 1.92; SD ¼ 1.22). In the stormrelated problem areas, problems with the post office reliability ranked as the most problematic issue (M ¼ 3.64; SD ¼ 1.28). Next, problems recruiting qualified applicants (M ¼ 2.97; SD ¼ 1.39) and other staffing issues (M ¼ 2.85; SD ¼ 1.44) were major factors inhibiting post-Katrina organizational recovery. Problems with a loss of customer base

& 2011 Blackwell Publishing Ltd.

ranked as the fourth most problematic issue (M ¼ 2.76; SD ¼ 1.41), followed by problems receiving insurance payouts (M ¼ 2.53; SD ¼ 1.50) and FEMA assistance (M ¼ 2.50; SD ¼ 1.64). The least problematic area concerned the reliability of suppliers (M ¼ 2.36; SD ¼ 1.24). In an effort to minimize the number of variables in the area of storm-related problems included in the subsequent multiple regression analysis, we examined correlations between the eight problem areas and the criterion variable organizational performance. Only three of the eight problem areas were significantly related to performance; these included problems with loss of customer base (r ¼ .47; po.01), staffing issues (r ¼ .29; po.01) and FEMA-related issues (r ¼ .26; po.01). Loss of customer base and staffing issues are both problem areas rooted in the severe population dislocation that occurred after Hurricane Katrina. These results support H7 in which we anticipated that problems related to population issues would have the greatest impact on organizational performance compared with other predictor variables. Clearly, these populationrelated problems were the most important factors among the eight problem areas. Now, our focus will be on these two particular areas of storm-related problems so that we can compare their effects with other types of variables when predicting organizational performance.

3.2. Organizational size and recovery advantage Large organizations (n ¼ 62) comprised 34% of our sample; 65.6% were small organizations (n ¼ 120). As stated in H1, larger organizations were expected to have an advantage over smaller organizations because they typically have more capital and resources to draw from during post-disaster recovery. Contrary to the hypothesized effects, organizational size was not related to organizational performance (F(1, 166) ¼ .18; p4.05). However, large organizations were significantly more likely to have an emergency response plan in place before Katrina (w2(1) ¼ 24.08; po.01) compared with small organizations. Eighty one percent of large organizations had an emergency response plan, whereas only

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

176

Christy M. Corey and Elizabeth A. Deitch

43% of small organizations were equipped with a plan. In addition, organizational size was related to the amount of storm preparation (F(1, 178) ¼ 10.68; po.01), such that large organizations were likely to take more pre-storm preparation measures (M ¼ 2.37; SD ¼ 1.30) compared with small organizations (M ¼ 1.75; SD ¼ 1.18). There were also two significant post-storm differences between large and small organizations: total forms of staff communication (F(1, 180) ¼ 48.09; po.01) and problems caused by loss of inventory (F(1, 181) ¼ 4.81; po.05). Specifically, large organizations (M ¼ 2.48; SD ¼ 1.13) tried more methods of communicating with staff members compared with small organizations (M ¼ 1.45; SD ¼.84). In addition, large organizations (M ¼ 2.34; SD ¼ 1.48) were less likely to have problems related to loss of inventory (F(1, 180) ¼ 4.81; po.05), compared with small organizations (M ¼ 2.88; SD ¼ 1.64).

3.3. The role of industry in recovery The organizations in this sample fell into 16 of 17 super sectors utilized by the Department of Labor for sorting companies into industry categories (see Table 2). Approximately two-thirds of the sample was located in five of the 17 super sectors: Retail trade (n ¼ 33), Accommodation and Food Services (n ¼ 32), Professional Business Services (n ¼ 23), Finance/Insurance (n ¼ 18) and Healthcare/Social Assistance (n ¼ 16). Other sectors had o10 organizations each. The data in Table 2 have been sorted according to the mean level of organizational performance. Consistent with H2, construction companies led the way, with an average 162% increase in performance, followed by Transportation/ Warehousing (36% increase) and Retail Trade (28.67%). Those sectors with the most severe dip in performance Table 2. Industry Sector, Sample Size and Organizational Performance Industry Sector

N

% Gain/loss mean (SD)

Construction Transportation/warehousing Retail trade Wholesale trade State/local government Real estate Natural resources/mining Manufacturing Other services Professional and business services Accommodation and food service Finance/insurance Health care/social assistance Information Arts/entertainment/recreation Education Federal government

5 5 33 5 3 4 4 9 6 23 32 18 16 6 6 5 0

162.00 36.00 28.67 18.75 16.67 3.33 2.50 1.11 2.00 4.57 5.52 12.00 13.13 21.50 29.17 35.00 –

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

(177.54) (94.23) (80.20) (40.49) (28.87) (95.04) (24.66) (51.03) (52.03) (35.85) (54.50) (18.40) (31.25) (39.82) (61.68) (47.26)

levels were Education (35% decrease), Arts/Entertainment/ Recreation (29.17% decrease) and Information (21.50% decrease). ANOVA with the top five sectors revealed significant differences on organizational performance (F(4, 110) ¼ 2.57; po.05). The lowest-performing sector, Healthcare and Social Assistance (M ¼ 13.13; SD ¼ 31.25), was significantly worse than the best-performing sector, Retail Trade (M ¼ 28.67; SD ¼ 80.20). Because analysis of these five sectors limits the sample size to two-thirds of the original sample, we also sorted all businesses in our sample into three broader industry categories (Construction, Retail and Other). Analyses of the main study variables showed no statistically significant effects using this broader three-group industry classification. Therefore, industry was excluded from further analyses.

3.4. Role of an emergency response plan For organizations in historically storm-prone areas, an emergency response plan seems essential. However, if the trend in the current sample of organizations is accurate, almost half of the businesses in the New Orleans area were operating without an emergency response plan in place before Hurricane Katrina. Slightly more than half of the sample or 55% (n ¼ 102) had a plan; 45% (n ¼ 81) faced the threat of Katrina with no plan for what to do before evacuation and what steps to take post-disaster. A series of one-way ANOVAs were used to test H3 concerning differences between organizations that had an emergency response plan before Katrina and those without a plan in the areas of storm preparation, communication, direct storm damage and storm-related problems (see Table 3). H3 received partial support in that organizations that had an emergency plan in place before Katrina showed differences on two variables. They were likely to engage in more storm preparation activities (F(1, 179) ¼ 10.68; po.01) and try more avenues for communicating with employees (F(1, 180) ¼ 48.09; po.01) compared with businesses without a plan. Clearly, the benefit of a plan is not limited to a simple organizational tool for emergency response. It facilitates more action in key staff-related areas like communication.

3.5. Organizational performance Our primary criterion variable, organizational performance, was operationalized two ways: first as a continuous variable representing performance gains and losses, and, second, as a categorical variable with three groups indicating the performance status of organizations (better, same or worse) compared with their pre-Katrina performance levels. The first half of this section will focus on the continuous measure that represents performance as a percentage where poorer performance has a negative value and better performance has a positive value. Zero scores indicate that the organization is performing the same as pre-Katrina levels. Although all businesses in the sample had survived Katrina, there was a

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery

177

Table 3. Effects of Having an Emergency Response Plan on the Main Study Variables Dependent variable

F

Plan mean (SD)

Amount of storm preparation Staff communication Total forms used Direct storm damage To physical structure Loss of electronics/files Loss of inventory Population-related problems Staffing issues Loss of customer base Staff loss Org. performance (% loss/gain)

50.48**

2.49 (1.17)

1.30 (1.04)

44.16**

2.25 (1.15)

1.26 (.63)

.08 .04 3.08

2.70 (1.23) 1.90 (1.19) 2.51 (1.57)

2.65 (1.10) 1.94 (1.26) 2.93 (1.62)

1.04 .94 1.28 .02

2.95 2.86 29.89 3.98

(1.37) (1.46) (29.62) (65.41)

No plan mean (SD)

2.73 2.65 35.44 5.53

(1.52) (1.35) (35.75) (70.10)

Note: n ¼ 169. **p  .01.

variation in their reported success. Forty-six businesses (25%) reported that their business was performing at the same level as pre-Katrina. The remaining businesses were split, with 61 (33.3%) reporting business was worse and another 70 (38.3%) reporting it was better. For those doing worse, the average reduction in business volume was 48.3% (SD ¼ 28.00). For those reporting improved performance, the average increase in volume was 56.9% (SD ¼ 77.70). Thus, approximately two-thirds of the sample had fully recovered or grown in terms of business performance 6–8 months after Katrina. In Hypotheses 4 and 5, we specified that more storm preparation and staff communication would predict greater levels of organizational performance and that more direct storm damage and storm-related problems would predict lower levels of performance. To test these expected effects, the relationships between the predictor variables and organizational performance were examined via correlational and multiple regression analyses. The bivariate correlations in Table 1 show that the continuous indicator of organizational performance was inversely related to five of the nine predictor variables. Contrary to H4, factors that managers have direct control over, including storm preparation and staff communication, were not significantly related to organizational performance. In contrast, storm-related factors such as direct-storm damage and post-storm problems had the largest effect on organizational performance; all five significant correlations were in the expected negative direction, which is consistent with H5. Because of the incredible population dislocation from the GNO area resulting from Hurricane Katrina, post-storm problems related to population issues (e.g., loss of customer base, staffing issues) were expected to have a greater impact on organizational performance compared with other types of predictors. Consistent with H7, loss of customer base (r ¼ .47; po.01) and staffing issues (r ¼ .29; po.01) had the greatest impact on organizational performance. In addition, all three areas of storm-related damage were significantly related to organiza-

& 2011 Blackwell Publishing Ltd.

tional performance such that increased damage led to worse performance. In support of H6, elevated rates of staff loss were associated with decreased organizational performance (r ¼ .30; po.01); this was the second highest correlation observed that included performance. The .61 correlation (po.01) between rates of staff loss and problems with staffing issues is a statistical indication of the conceptual redundancy in the items. To avoid this redundancy and the potential effects of multicollinearity, we excluded the variable rates of staff loss from the multiple regression analysis. We chose to include problems with staffing issues because it is one of those key population-related predictors that we expected would matter the most when predicting organizational performance. To explore the additive effects of the predictor variables on organizational performance, they were entered in blocks in a step-wise multiple regression analysis in the order in which they appear in Table 4. In further support of H7, the two storm-related problems most related to population issues, staffing problems and loss of customer base, were the only significant individual predictors of organizational performance, although damage as a block was a significant predictor. The model accounted for 30% of the variability in the criterion variable. We also examined group differences in the main study variables relating to organizational performance status using a series of one-way ANOVAs. For the performance status variable, organizations were sorted into three groups, better, same or worse, based on managers’ reports of their current performance status compared with that before Hurricane Katrina. The means of the three performance status groups were compared across the areas of storm preparation, communication, direct-storm damage and storm-related problems (see Table 5). Pairwise post hoc comparisons were conducted using the Student–Neuman– Keuls method. Similar to the results with the continuous performance measure, there were no mean group differences for performance status on the amount of storm

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

178

Christy M. Corey and Elizabeth A. Deitch

Table 4. Predictors of Organizational Performance Variable Amount of preparation Staff communication Problems reaching staff Total forms used Direct storm damage To physical structure Loss of electronics/files Loss of inventory Population-related problems Staffing issues Loss of customer base

Unstandardized coefficient

SE

Standardized coefficient

DR2 for block

5.07

4.29

.09

.00

13.64 1.51

10.33 5.25

.09 .02

.04

9.83 1.56 4.82

5.91 6.20 4.49

.17 .03 .11

.06*

9.78* 20.34*

3.60 3.53

.20 .42 Total R2

.20* .30

Note: n ¼ 172. *p  .05.

Table 5. Mean Differences in the Main Study Variables Relative to Organizational Performance Status Dependent variable Preparation Amount of prep Staff communication Total forms used Direct storm damage Physical structure Electronics/files Loss of inventory Storm-related problems Staffing issues Loss of customer base Supplier reliability Insurance issues FEMA Postal delivery Rate of staff loss

F (df)

Better

Same

Worse

.69 (2, 174)

1.81a

2.07a

2.02a

1.40 (2, 175)

1.65a

1.89a

1.95a

5.46 (2, 176)** 1.40 (2, 176) 2.12 (2, 176)

2.31a 1.73a 2.53a

2.80b 1.92a 2.50a

2.93b 2.09a 3.03a

173)** 172)** 171)** 162)* 148)* 171)* 171)**

2.53a 2.00a 2.49b 2.43a 2.18a 3.34a 27.51a

2.64a 2.48b 1.84a 2.16b 2.40a 3.63ab 21.31a

3.28b 3.74c 2.58b 2.96a 2.94a 3.97b 44.38b

5.03 35.53 5.38 3.82 3.11 3.83 8.13

(2, (2, (2, (2, (2, (2, (2,

Note: Within each row, group means that share a common superscript do not differ significantly; if two group means do not have a superscript in common, their difference is statistically significant (p  .05). *p  .05; **p  .01.

preparation or post-storm communication issues. Significant differences were observed in terms of damage to an organization’s physical structure (F(2, 176) ¼ 5.46; po.01). Organizations indicating the same or worse performance compared with pre-Katrina levels had significantly more storm damage compared with those performing better. In addition, performance status differences were observed in all six post-Katrina problem areas. Compared with organizations with better performance status, those in the worse performance status group had significantly more problems in the areas of staffing issues, loss of customer base and postal delivery (see Table 5). Although the effect of performance status was significant in the other three problem areas, observed mean differences between the better and worse groups were not in the expected direction. Organizational performance status was significantly related to another important outcome variable: rate of staff loss (F(2, 171) ¼ 8.13; po.01). Organizations with a worse performance status (M ¼ 44.38; SD ¼ 35.43) lost approximately 20%

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

more staff compared with organizations in the better (M ¼ 27.51; SD ¼ 31.03) and same status groups (M ¼ 21.31; SD ¼ 23.94).

4. Discussion This study attempted to identify the factors that mattered in surviving businesses’ recovery after Hurricane Katrina, a disaster unlike any before experienced in this country. Many important factors were identified that were similar to those found in research on previous disasters. Those areas of difference highlight the uniqueness of every individual disaster. Nonetheless, the findings do have implications for business preparedness for future disasters, as well as policy direction in helping communities recover from disaster.

4.1. Predictors of organizational performance By far, the most important variables in business recovery seem to be those related to the massive population

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery dislocation that occurred in this disaster. The strongest single predictor was the extent of loss of customer base, with the amount of staff loss second in importance. Together, these population-related variables accounted for 21% of the variability in post-Katrina organizational performance. This suggests that post-disaster aid that focuses on helping the surrounding community recover may have as huge an impact on business recovery as direct aid to businesses. However, it is possible that the overriding importance of this issue is somewhat unique to Hurricane Katrina, as this level of population dislocation far surpassed that of any previous disaster in the United States. Interestingly, more recent data, collected more than three years post-Katrina (Deitch & Corey, 2011), show that population issues are the one factor still influencing the organizational performance levels in the area. These trends suggest that the residual impact of population dislocation on businesses’ customer base and labour supply is long lasting despite the fact that the New Orleans area population continues to recover and approach pre-Katrina levels of 1.31 million residents. US Census data following Katrina show growth in the GNO population from 1.11 million people in 2007, to 1.17 million in 2008, to 1.19 million in 2009 (US Census Bureau, 2009). Thus, the effects of the dispersed population not only caused the greatest impact immediately following the disaster but have also persisted long after all the physical damage has been cleaned up. Overall physical damage was also a significant predictor of organizational performance. When examining performance as a categorical variable, damage to buildings did differentiate between those firms performing better and those performing worse or with no change. This is not surprising, as previous research has usually found that to be the case (Tierney, 1997a; Webb et al., 2002). More damage typically leads to longer closure after a disaster, hampering recovery (Webb et al., 2002). In addition, those businesses with more damage likely experienced financial losses in repairing that damage, as many were underinsured, with only 64% reporting that their business had flood insurance.

4.2. Emergency preparedness Although it was unrelated to post-disaster organizational performance, the current study did show a clear benefit to having an emergency plan in place before disaster strikes. Those organizations that had such a plan undertook more preparatory activities to protect their businesses and maximize their ability to continue to do business than did those without a plan. Emergency plans also facilitated significantly more communication with employees. Unfortunately, the degree to which emergency plans were executed and the effectiveness of such plans were not verified in the current study. Although managers were asked to rate the effectiveness of their emergency plan if they had one, we were concerned that the subjective perceptions of emergency plan effectiveness 6 months after the storm would be confounded with how well the business was faring at the

& 2011 Blackwell Publishing Ltd.

179 time. Those managers in better performing organizations would be more likely to rate their plan as effective compared with managers in worse performing organizations. Future research should use more objective ways of measuring emergency plan effectiveness that allows one to tease out the potentially confounding effects of post-disaster performance when making that evaluation. The fact that 55% of businesses had an emergency plan in place before Hurricane Katrina, while obviously not sufficient, does compare favourably with the rates of emergency plan adoption in other locations. For example, Dahlhamer and D’Souza (1997) surveyed businesses in Des Moines (with flood risk) and Memphis (earthquake risk) and found that only 29% and 22% percent of businesses in those locales, respectively, had developed any emergency plan. Additionally, more recently collected data on emergency preparedness levels in the New Orleans area before the approach Hurricane Gustav in the fall of 2009 found that 76.5% of that sample had a plan in place (Deitch & Corey, 2011). Thus, it seems that New Orleans area businesses did at least learn from prior experience. This again compares favourably with other locations. For example, Flynn (2007) found that even after experiencing a major flood disaster in 1997, the percentage of businesses in Grand Forks, ND, with a disaster recovery plan only reached 11.83%. It is likely that the sheer magnitude of the Katrina disaster made a greater impression on business owners than the previous disaster experience cited in those other studies, which may not have been as severe. Having an emergency plan in place before the disaster and the amount of pre-storm preparation were not significant predictors of performance. This is likely due to the fact that storm preparation was completely unrelated to the actual amount of damage a business sustained. Perhaps the sheer scale of the disaster caused by Hurricane Katrina outmatched storm preparatory efforts. There may be a critical threshold for the scale of a disaster before which storm preparatory activities would have a noticeable impact on business recovery, but after which the beneficial effects are negligible. Perhaps Hurricane Katrina and the aftermath that followed surpassed that threshold. Therefore, our results should not be interpreted to indicate that preparation is unimportant. Regarding this particular disaster, the amount of damage sustained was unprecedented. Had the damage not reached such catastrophic levels, we might have seen a relationship between preparation and damage.

4.3. Organizational characteristics Regarding the performance of the various industry sectors, we were not surprised to find that construction businesses reported the most increases in business post-Katrina. That has been a consistent finding in previous disaster research, as construction businesses are directly needed to recover, and so obviously will see greater demand for their services. What was unusual, however, was the positive performance

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

180 reported by retail and wholesale businesses. This is in direct contrast to findings from other disasters, where retailers were found to be having the most difficulty recovering (Boarnet, 1998; Chang & Falit-Baiamonte, 2002; Webb et al., 2002), although it is consistent with Runyan’s (2006) findings on the Gulf Coast after Katrina. Because of the magnitude of this disaster and the slow pace of SBA assistance (Loten, 2005), many pre-Katrina businesses were not operating at all in the spring of 2006 when data were collected. Thus, the reported gains here are likely due to the fact that there were simply fewer available retail outlets, and so those retail organizations that were open with products to sell had far less competition. Despite the fact that New Orleans lost population while the areas studied by Runyan gained, the decimation of the competition was apparently large enough even to offset that greatly reduced local population. Also, contrary to findings from other disasters, organizational size did not predict performance or staffing outcomes. Even though larger organizations were more likely to prepare for disaster, this advantage was apparently not substantial enough to make a difference in organizational outcomes.

4.4. Conclusion and future directions The study of factors contributing to business recovery after Hurricane Katrina has made clear the undeniable impact that massive population dislocation has on businesses in the disaster affected area. No other disaster in US history has been coupled with a population dispersion of this magnitude. This fact makes the process of business recovery in the GNO area so unique and specific to Katrina that it creates problems when we try to use these findings to further develop theories of disaster management. In addition, the distinction between surviving businesses and non-surviving businesses is an important one because we cannot generalize the results of this study to understand the determining factors for businesses that closed in the wake of Hurricane Katrina. Most of the businesses included in this study were outside of the most heavily devastated regions of New Orleans. Given that the scale of the disaster faced by local businesses varied tremendously depending on the elevation of the business and the proximity to levee breech points, those businesses that survived post-Katrina were clearly able to overcome the scale of the disaster that they were forced to face. However, this study provides a significant contribution to the disaster management literature because it is one of the first empirical studies concerning Hurricane Katrina that considers factors affecting business recovery at the level of the organization, as opposed to more aggregated, macrolevel data. In addition, our data are rich in detail because they were provided strictly by business owners and managers who craved an opportunity to describe their harried journey in post-Katrina business recovery. In addition to objective data, this investigation yielded subjective data including

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011

Christy M. Corey and Elizabeth A. Deitch managerial ratings of significant problem areas and opinions of federal assistance programmes. Although we agree that more objective, macro-level studies of post-disaster business recovery are valuable, future research should include more micro-level approaches where the opinions of managers and owners are documented so that this body of literature will reflect the importance of the state of mind that business managers and leaders are in post-disaster and how this affects business recovery.

References Aldrich, H. and Auster, E.R. (1986), ‘Even Dwarfs Started Small: Liabilities of Age and Size and Their Strategic Implications’, in Staw, B.M. and Cummings, L.L. (Eds), Research in Organizational Behavior, Vol. 8. JAI Press, Greenwich, CT, pp. 165–198. Baade, R.A., Baumann, R. and Matheson, V. (2007), ‘Estimating the Economic Impact of Natural and Social Disasters, with an Application to Hurricane Katrina’, Urban Studies, Volume 44, pp. 2061–2076. Boarnet, M.G. (1998), ‘Business Losses, Transportation Damage, and the Northridge Earthquake’, Journal of Transportation and Statistics, Volume 1, pp. 49–62. Boin, A. and McConnell, A. (2007), ‘Preparing for Critical Breakdowns: The Limits of Crisis Management and the Need for Resilience’, Journal of Contingencies and Crisis Management, Volume 15, pp. 50–59. Bond, B. (2007), ‘Lessons in Preparedness for the Response to Hurricane Katrina’, Journal of Business Continuity and Emergency Planning, Volume 1, pp. 426–431. Chang, S.E. and Falit-Baiamonte, A. (2003), ‘Disaster Vulnerability of Businesses in the 2001 Nisqually Earthquake’, Environmental Hazards, Volume 4, pp. 59–71. Cochrane, H.C. (1992), ‘Overview of Economic Research on Earthquake Consequences’, in The Committee on Earthquake Engineering, National Research Council (eds), The Economic Consequences of a Catastrophic Earthquake: Proceedings of a Forum, National Academy Press, Washington, DC, pp. 100–111. Cohen, M.J. (1993), ‘Economic Impact of an Environmental Accident: A Time-Series Analysis of the Exxon Valdez Oil Spill in South Central Alaska’, Sociological Spectrum, Volume 13, pp. 35–63. Cole, C.R., Corbett, R.B. and McCullough, K.A. (2005), ‘2004 Hurricane Losses: Testing the Lessons Learned from Hurricane Andrew’, CPCU eJournal, Volume 58, pp. 1–9. Dahlhamer, J.M. and D’Souza, M.J. (1997), ‘Determinants of Business Disaster Preparedness in Two U.S. Metropolitan Areas’, International Journal of Mass Emergencies and Disasters, Volume 15, pp. 265–281. Dahlhamer, J.M. and Reshaur, L. (1996), Businesses and the 1994 Northridge Earthquake: An Analysis of pre- and Post-Disaster Preparedness, Paper presented at the annual meeting of the Eastern Sociological Society, Boston, MA. Dahlhamer, J.M. and Tierney, K.J. (1998), ‘Rebounding from Disruptive Events: Business Recovery Following the Northridge Earthquake’, Sociological Spectrum, Volume 18, pp. 121–141. Deitch, E.A. and Corey, C.M. (2011), ‘Predicting Long-term Business Recovery Four Years After Hurricane Katrina’, Management Research Review, Volume 34, pp. 311–324.

& 2011 Blackwell Publishing Ltd.

Factors Affecting Business Recovery Edwards, F.L. (2006), ‘Businesses Prepare Their Employees for Disaster Recovery’, Public Manager, Volume 35, pp. 7–12. Eguchi, R.T., Goltz, J.D., Taylor, C.E., Chang, S.E., Flores, P.J., Johnson, L.A., Seligson, H.A. and Blais, N.C. (1998), ‘Direct Economic Losses in the Northridge Earthquake: A Three-Year Post-Event Perspective’, Earthquake Spectra, Volume 14, pp. 245–264. Flynn, D.T. (2007), ‘The Impact of Disasters on Small Business Disaster Planning: A Case Study’, Disasters, Volume 31, pp. 508–515. Institute for Southern Studies (ISS). (2009), Grading the Katrina Recovery. Available at http://www.southernstudies.org/ISSKatrina %204YearReport.pdf (accessed 1 February 2010). Knabb, R.D., Rhome, J.R. and Brown, D.P. (2006), Tropical Cyclone Report: Hurricane Katrina: 23–30 August 2005. National Hurricane Center. Available athttp://www.nhc.noaa.gov/pdf/TCRAL122005_Katrina.pdf (accessed 1 September 2011). Kroll, C.A., Landis, J.D., Shen, Q. and Stryker, S. (1990), ‘The Economic Impacts of the Loma Prieta Earthquake: A Focus on Small Business’, Berkeley Planning Journal, Volume 5, pp. 39–58. Loten, A. (2005), Feds slow with disaster loans, business owners say, 10 October. Available at http://www.inc.com/news/articles/200510/ sba.html (accessed 10 September 2010). Olejarski, A.M. and Garnett, J.L. (2010), ‘Coping with Katrina: Assessing Crisis Management Behaviors in the Big One’, Journal of Contingencies and Crisis Management, Volume 18, pp. 26–38. Rose, A., Benavides, J., Chang, S.E., Szczesniak, P. and Lim, D. (1997), ‘The Regional Economic Impact of an Earthquake: Direct and Indirect Effects of Electricity Lifeline Disruptions’, Journal of Regional Science, Volume 37, pp. 437–458.

& 2011 Blackwell Publishing Ltd.

181 Runyan, R.C. (2006), ‘Small Business in the Face of Crisis: Identifying Barriers to Recovery from a Natural Disaster’, Journal of Contingencies and Crisis Management, Volume 14, pp. 12–26. Tierney, K.J. (1997a), ‘Business Impacts of the Northridge Earthquake’, Journal of Contingencies and Crisis Management, Volume 5, pp. 87–97. Tierney, K.J. (1997b), ‘Impacts of Recent Disasters on Businesses: The 1993 Midwest Floods and the 1994 Northridge Earthquake’, in Jones, B. (Ed), Economic Consequences of Earthquakes: Preparing for the Unexpected, Multidisciplinary Center for Earthquake Engineering Research, Buffalo, NY, pp. 189–222. United States Census Bureau. (2009), County Total Population and Estimated Components of Population Change, April 1, 2000 to July 1, 2009. Available athttp://www.gnocdc.org/census_op_estimates. html (accessed 1 February 2010). Webb, G.R., Tierney, K.J. and Dahlhamer, J.M. (2002), ‘Predicting Long-Term Recovery from Disasters: A Comparison of the Loma Prieta Earthquake and Hurricane Andrew’, Environmental Hazards, Volume 4, pp. 45–58. West, C.T. and Lenze, D.G. (1994), ‘Modeling the Regional Impact of Natural Disaster and Recovery: A General Framework and An Application to Hurricane Andrew’, International Regional Science Review, Volume 17, pp. 121–150. Zhang, Y., Lindell, M.K. and Prater, C.S. (2009), ‘Vulnerability of Community Businesses to Environmental Disasters’, Disasters, Volume 33, pp. 38–57. Zolin, R. and Kropp, F. (2007), ‘How Surviving Businesses Respond During and After a Major Disaster’, Journal of Business Continuity and Emergency Planning, Volume 1, pp. 183–199.

Journal of Contingencies and Crisis Management Volume 19 Number 3 September 2011