This update is part of the post-Harvey Economic Recovery and Resilience project funded by the U.S. Economic Development Administration (EDA). The project team consists of Texas A&M University-Corpus Christi’s College of Business, Coastal Bend Business Innovation Center, and South Texas Economic Development Center.
Download this report in PDF.
It’s been nine months since Hurricane Harvey wreaked havoc on Texas. This historic storm left unprecedented property destruction to a widespread region along the state’s Gulf Coast. Although it’s still too early to know its full impact, public data allow us to understand the extent of Harvey’s immediate impact on the regional economy other than property losses. This report takes a look at the changes of economic activity so far, both negative and positive, in the local communities devastated by Harvey.
Among the 41 counties designated as Harvey disaster areas, the Federal Emergency Management Agency (FEMA) has rated 22 counties as those sustaining medium or high property damage. This report draws on popular measures of economic activity, such as unemployment and employment, for those 22 counties that bore the brunt of Harvey’s impact. These data are not publicly available for cities or towns.
MAP: Harvey Disaster Areas in Texas
Harvey made its first landfall in Aransas County, devastating its cities of Rockport and Fulton, and nearby communities, particularly Port Aransas in Nueces County and Bayside in Refugio County. After continuing to head north toward San Antonio, the storm detoured toward Houston. Whereas record amounts of rainfall caused catastrophic flooding in southeastern Texas inland cities, such as Houston and Beaumont, most property and infrastructure damages in the western half of the Harvey impact zone were caused by destructive winds over 130 miles per hour and storm surge over 7 feet above ground.
Harvey is widely expected to be one of the costliest natural disasters in U.S. history. The National Oceanic and Atmospheric Administration (NOAA) has estimated this storm’s total costs to be at least $125 billion. But how a storm’s economic impacts are calculated is not an exact science.
The most obvious of those costs represent direct damage to structures and other properties, such as vehicles, and infrastructure, such as roads, bridges and power lines. Those losses alone do not show up in the government’s regular economic reports because they represent economic activity or spending that has already taken place. Put simply, property damage or destruction alone does not affect the conventional measures of economic activity, such as output.
In addition to direct damage, natural disasters like Harvey cause indirect economic losses, which include disruptions in business and other productive activities due to evacuation and lost utilities, damage to structures and equipment, displaced employees and so on. These affect the economy indirectly through reductions in business sales, lost jobs and wage earnings.
But changes in the economy of an affected area in the wake of a natural disaster are the outcomes of not only losses but also gains as a result of subsequent rebuilding activity. The inflow of disaster relief funds from various government agencies and insurance payments in the wake of the storm have raised overall spending that would have spurred job and income growth. For instance, FEMA has committed more than $30 billion for Harvey-related relief and the Small Business Administration (SBA) has approved more than $3.3 billion in disaster loans in Texas.
The following describes Harvey’s local economic impact and recovery that take into consideration of these indirect losses and gains.
One way to understand the impact of Harvey on an affected area’s economy is to look at changes in its labor market conditions following the storm. The following scatter plot shows FEMA’s scores for the storm’s impact on individual counties (horizontal axis) against the scores for each county’s own rebuilding capacity (vertical axis). The bubbles depict the extent of changes in the local labor market.
CHART: Harvey Impacts & Rebuilding Capacity
The primary factor for the impact score of a county is the number of applicants for FEMA’s individual assistance relative to the total number of households within that county. With more than 80 percent of local residents registering with FEMA, Aransas, Jefferson and Orange Counties received the highest score of 3.
Factors for determining the rebuilding capacity of a county include its local governments’ own budget sizes and other financial resources. A score of zero indicates the lowest institutional capacity to recover. Among the 22 counties, Aransas County has the lowest score closed to zero, whereas Harris County has the highest score above 1.5.
In the above scatter plot, a location closer to the lower-left corner of the plot means the area sustained more property damage from Harvey but it has less rebuilding capacity. As the smallest of these 22 counties by area, Aransas County is not only most affected by Harvey but it has the least financial capacity to rebuild without external assistance.
Closest to the upper-right corner of the plot is Harris County. As part of the Houston metro area, Harris is the most populous county in Texas. Despite no more than 25 percent of this county’s residents applied for FEMA assistance, the number of applicants exceeded 130,000 due to the sheer size of its population.
The size of each bubble in the scatter plot reflects the immediate impact of Harvey on a county’s unemployment rate. This impact measure is the difference between the actual unemployment rate in September and a baseline for comparison. A baseline is a modeling projection for periods following Harvey under the counterfactual assumption of no hurricane event in August.
Not surprisingly, Aransas County, where Harvey made its first landfall, experienced the most economic impact. Its unemployment rate nearly doubled from 5.4 percent in July and August to 10.3 percent in September.
Counties with higher FEMA impact scores also tended to face larger deterioration in their labor market conditions (or larger bubbles). But unemployment did not rise immediately in all affected areas. Such counties as Bastrop, DeWitt, and Jasper actually experienced lower unemployment in September than expected.
The plot also suggests that larger counties tended to be more able to absorb Harvey’s impact on their local economies. Relatively smaller counties, particularly Aransas, Refugio and Jefferson, faced a sizable jump in the unemployment rate in September. Meanwhile, the largest counties, such as Harris and Nueces, saw little change.
ROAD TO RECOVERY
Changes in local unemployment rates since September help us visualize how the Harvey impacted communities have responded to the storm so far. The following set of charts displays historical unemployment rates for each county along with our baseline projection (dash line) under the assumption that the storm event did not occur. The projected data reflects the unemployment rates consistent with pre-Harvey conditions. For most counties, the baseline beginning September 2017 is relatively flat except for movements that capture seasonal changes.
CHART: Unemployment Rate by County vs. Baseline
This is not the case for actual unemployment. For Aransas County, the unemployment rate has been declining steadily after a dramatic surge in September. By April, its local unemployment rate reached 6.1 percent, compared to the projected rate at 4.5 percent. This pattern of persistent improvement is shared by its neighboring Refugio County, which also experienced a steady decline in unemployment from 7.5 percent in September to 4.9 percent in the following April. These two counties consist of neighborhoods where Harvey made landfall.
The majority of other counties also saw declining unemployment rates since September, but the sizes of improvement were relatively smaller. Still Harvey’s negative impact on local economies is not noticeable for some counties, such as Bee, DeWitt and Goliad. In these cases, the actual unemployment rates are exceptionally close to the baseline.
EMPLOYMENT AND EARNINGS
The observed unemployment impacts over time can be translated into losses in jobs and wage earnings. For each county, the employment loss in a particular month after Harvey is first measured by the size of its existing local labor force times the difference between the actual unemployment rate and baseline unemployment rate. The loss in total wage earnings equals the county’s average employee earnings times the estimated employment loss.
The following bar chart shows the estimates for cumulative local wage impacts by the end of April. While half of the counties experienced the widely expected employment and wage losses, the other half counties saw net gains by April.
CHART: Wage Earnings Impacts
By April, the 22 counties collectively experienced a net loss of $107 million in employee earnings. Harris County stood to lose $46.7 million due to Harvey’s impact on its overall employment level. Nueces County, which encompasses Port Aransas that was devastated during Harvey’s landfall, has lost $11 million so far.
On the contrary, Brazoria County, another part of the Houston metro area, has experienced a net income gain of nearly $6 million so far. This gain was a result of a higher than expected employment level during seven of the first nine months following Harvey.
The observation that half of the 22 counties had already experienced a net employment or income gain by April underscores the role of recovery and rebuilding efforts across the Harvey impacted communities. Yet the labor force in those communities in the wake of the storm might have changed for different reasons. Destroyed homes might have reduced the size of the local labor force due to the loss of displaced workers who left the affected community permanently. Other things being equal, the expected loss of population and thus labor force in the affected counties would have reduced their unemployment rates without creating any new jobs.
On the other hand, losses in labor force due to out-migration may be offset by temporary gains of workers coming from the rest of the nation. Along with the influx of workers for restoring utilities and infrastructure and cleaning up debris, Harvey-related relief and reconstruction funds from the government and other sources might have kick-started the recovery process before the end of 2017.
How much has the local labor force changed in the Harvey impacted areas? The following bar chart shows the difference between the labor force size of each county over the post-Harvey period so far and the size during 12-month period immediately before Harvey (August 2016 to July 2017). While 13 of the 22 counties saw a net decline in the local labor force during the post-Harvey period, other counties actually saw a net gain as much as 3.4 percent.
CHART: Post-Harvey Labor Force Changes
The range between -4 percent (Tyler) and 3.4 percent (Colorado) is remarkable. This highlights the importance of restoration and reconstruction efforts in our measure of Harvey’s full impact.
LABOR FORCE CHANGES
Now we consider the effect of labor force changes in our measure of Harvey’s impact on local employment. The following charts compare for each of the 22 counties the employment impact in percentage terms based on the counterfactual assumption of no effect of Harvey on the labor force against estimates that take into account net labor force changes following Harvey.
For each month, the first column (blue) represents Harvey’s impact on employment without possible effects of the storm or the subsequent relief and rebuilding activities on the labor force. The second column (orange) depicts Harvey’s employment impact that accounts for the net change in the local labor force associated with the effects of displaced workers (which reduced the labor force) and an influx of workers to the county due to relief and reconstruction-related spending (which would have raised the labor force).
CHART: Employment Impacts (%) by County
The differences between the two patterns of employment impact estimates are remarkable. Take Aransas County for example. Without considering labor force changes in response to Harvey, the county appears to have bounced back steadily but slowly, with a loss of employment by 1.6 percent in April in comparison to 5.1 percent in September.
But despite an expected permanent reduction in population by local officials, the county’s labor force has expanded by a monthly average of 308 individuals, or about 3 percent. Some of the new workers might have come from its neighboring counties. Since September, Nueces, San Patricio, and Refugio counties have all seen a net decrease in their local labor force.
As for Aransas County, the two counties in the Houston metro area, Brazoria and Harris, saw a net increase in labor force. Harris County alone has so far attracted more than 18,000 workers per month on average. The net increase in the labor force related to relief and rebuilding activities has helped reduce the employment loss by an average of about one full percentage point during that 9-month period. The positive impact of mitigation programs and rebuilding activity on the local economy can also be seen in other counties.
The following charts compare the two alternative measures of employment impacts for the 22-county Harvey impact zone as a whole. Without adjusting for net changes in the labor force, the estimated employment impact was positive only between October and December, when restoration and cleanup activities occurred.
CHART: Employment Impacts With and Without Labor Force Change
An influx of workers into the affected communities after Harvey has helped raise the region’s overall employment level. The second chart, which includes the effects of net labor force changes, indicates that employment across the region has in fact expanded since last November. By April, the region as a whole stood to gain more than 50,000 jobs due in part to the simulative effects of the post-Harvey recovery and rebuilding efforts. The two counties in the Houston metro areas, Harris and Brazoria, accounted for most of those job gains (92%).
This area-wide view nevertheless ignores much of the differences across counties: While about half of those counties have added jobs during the recovery process, the other counties have struggled with outmigration of workers. Other economic factors, such as increasing Texas oil production in the Eagle Ford Shale, might have also helped improve the economic performance of some local economies more than others within the region.
Now it seems clear that local recovery and rebuilding efforts in the wake of Harvey have already more than offset the storm’s toll on economic activity across the disaster zone. The next question is whether the observed improvement in labor market conditions has also been supported by recovery in the business sector. We address this question by looking at sales tax revenues collected local taxing authorities. The column charts below show for each county year-over-year percentage change in its sales tax volume beginning July 2017. The timeline indicates the months of sales tax collection, which was returned by the state two months later.
CHART: Year-over-Year Changes in Sales Tax Revenues by County
The above charts reveal the impact of Harvey on businesses especially in late August when most businesses in the affected areas were closed. But since then sales tax revenues have grown for most counties. Hardin County, part of the Beaumont-Port Arthur metro area, is a notable exception.
The business sector in most counties is not fully restored yet. In Aransas County, Harvey damaged nearly 80 percent of its business establishments, but over three-quarters of businesses in its Rockport-Fulton area are now back in operation. Meanwhile, home repair and other reconstruction activities are a boon to local retail sales and thus sales tax revenues.
Despite Harvey’s widespread devastation to infrastructure and property, measures of changes in local employment and business activity have helped us better understand the transitory nature of its economic impacts. This observation is consistent with an economic impact analysis for the hurricane conducted by the Texas Comptroller of Public Account.
The Comptroller’s office expects the state to lose $3.8 billion in Gross State Product (GSP), the broadest measure of economic activity for the state, within the first year of Harvey. This estimate is the difference between the estimated losses of $16.8 billion in economic activity immediately after the storm and the estimated gains of $13 billion during the subsequent recovery process. In the second year, Harvey is are expected to generate a net gain of $2.1 billion in GSP.
These estimates depict net changes in Texas’ economy as a whole. In all, the state economy should have already bounced back to pre-Harvey level by the end of 2017. Our overall findings across the disaster zone bode well with this high-level perspective.
As another hurricane season is approaching, government officials and policymakers alike can rest assured that Harvey’s economic recovery is taking shape across Texas. Government disaster-related aid and rebuilding activity might have offset much of Harvey’s economic toll on the affected communities.
After all, evidence of wide-ranging impacts and recovery speeds among individual communities implies that we should pay also particular attention to the performance of local economies. Future studies may benefit from more granular data, such as individual sectors within a city, or an analysis of the effects of resource allocations during the recovery phase, such as the relief and reconstruction funds from government and other sources.
The views expressed in this article are those of the author and should not be attributed to Texas A&M University-Corpus Christi or any of its units. For questions, please contact Jim Lee (firstname.lastname@example.org), director of South Texas Economic Development Center.
This project benefits from collaboration with various government agencies and other organizations. For this report, the author is grateful to first-hand data provided by the following organizations and individuals:
- FEMA Community Planning and Capacity Building: Rick Martin
- Rockport-Fulton Chamber of Commerce: Diane Probst, Mike Woods
- Santos McBain Management & Planning: Maryann Carl, Ray De Los Santos
- Small Business Administration: David Elizondo, Richard Jenkins, Mark Randle
RELATED STEDC ARTICLES:
- “Hotel Performance after Harvey,” South Texas Economic Pulse, 2018, Issue 7.
- “Business Recovery after Harvey,” South Texas Economic Pulse, 2018, Issue 6.
- “Harvey’s Impact on Corpus Christi,” South Texas Economic Pulse, 2018, Issue 3.
 The 22 counties that bore the brunt of the devastation by Harvey are Aransas, Bastrop, Bee, Brazoria, Calhoun, Colorado, DeWitt, Goliad, Hardin, Harris, Jackson, Jasper, Jefferson, Newton, Nueces, Orange, Polk, Refugio, San Jacinto, San Patricio, Tyler, and Victoria. Except Brazoria and Harris of the 9-county Houston metro area, all counties has an overall FEMA storm impact score of 2, which represents a medium or higher storm impact. The impact score for Brazoria and Harris Counties is 1.7, but some of its cities have a score of 2 or higher.
 For each county, we made a baseline projection for its unemployment rate and labor force size between September 2017 and December 2018. These projected data, or forecasts, are constructed by applying a statistical model that best fits the actual monthly data between January 2000 and July 2017. This model takes into consideration of the historical trend over a long period and seasonal patterns within a year.
 For each county, the estimate is the total wage impact equals the 9-month equivalent of the county’s average annual wage earnings times the average of the monthly employment impact estimates during the 9-month period between September and April.
 The measure of employment impacts corresponds to the vertical difference between the actual unemployment rate and the baseline shown in the corresponding unemployment chart above. The measure without the net effect of labor force changes equals the negative of that difference times the average labor force size of the 12-month period immediately before Harvey (August 2016-July 2017). The measure with the net effect of labor force changes is the above measure plus Harvey’s impact on the labor force. For each month beginning September 2017, Harvey’s impact on the labor force size is measured by the difference between the actual size and its baseline projected by the same method as for the unemployment rates above.
 For counties that do not impose a sales tax, such as Newton, we used the data of it largest city.
 Texas Comptroller of Public Account, “A Storm to Remember: Hurricane Harvey and the Texas Economy,” Fiscal Notes, February 2018.