This blog covers advances in child health research and in mapping techniques. We apply these stories to central Texas.

The Suburbanization of Poverty: The Geography of Childhood Poverty in and around Austin
By Sam Tedford
 
In recent years, Austin has settled in near the top of the list of fastest growing cities in the United States. And though we know that the city is rapidly growing, there is still much to understand about the changing demographics of Austin. Historically, racist urban planning led Austin to develop into an intensely segregated city both economically and racially with large concentrations of urban poor communities and communities of color forced into certain neighborhoods on the east side. However, in the changing social and economic environment of Austin today, the geography of poverty and race is unlike anything this city has seen before. A quickly growing city means a growing demand for housing, which has led to a major affordability crisis in the Austin housing market, and the displacement of many low income residents due to gentrification. For my study I focused on one of the most vulnerable populations directly affected by poverty and displacement: children. The study analyzes the geographic distribution of economically disadvantaged children in Austin and the surrounding areas across ten years from 2005 to 2015.
 
 

Mapping Disadvantaged Populations

 
Using data from Texas Education Agency performance reports from the 2004-2005 school year and the 2014-2015 school year, I initially mapped both district level and campus level economically disadvantaged rates, defined as the percentage of the student population on free and reduced lunch, while denoting Title I qualifying schools in each time period.
 
In the 2004-2005 school year, 12 out of 27 of the school districts surrounding Austin had over 40% of their student population on free or reduced lunch. By the 2014-2015 school year, 20 out of the 27 districts had a population comprised of over 40% economically disadvantaged kids.


School Districts


Economic disadvantage rate by ISD 2005

Economic disadvantage rate by ISD 2015


 
In the 2004-2005 school year, 117 out of 200 campuses in the study area qualified as Title I. In the 2014-2015 school year, 164 out of 266 campuses qualified as Title I.

School Campuses


Economic disadvantage rate by campus 2005

Economic disadvantage rate by campus 2015


 
 

Mapping Change over Time

 


Economic disadvantage rate change over time by ISD

 
 

Mapping Hotspots

 

In an attempt to find patterns in the rise of economically disadvantaged student populations, I first looked for spatial auto-correlation, or clustering, of economically disadvantaged populations in the district-level data. By calculating Moran’s I, a spatial statistic which compares each target feature to its neighbors as well as to the mean of all features, the district-level data showed statistically significant clustering with a 95% confidence interval. This initial calculation compelled me to continue exploring patterns in geographic proximity of economically disadvantaged districts and calculate the Getis-Ord Gi* statistic for the district-level data, another spatial statistic which identifies hot and cold spots, for economically disadvantaged student populations.

The results of the hot spot analysis reveals hotspots in Elgin, Bastrop, Manor, and Del Valle, meaning those school districts had a statistically significant concentration of economically disadvantaged populations. This method of analysis also reveals coldspots, here meaning statistically significant concentrations of low economically disadvantaged percentages, in five school districts: Dripping Springs, Lake Travis, Lago Vista, Eanes, and Leander.


Economic disadvantage hot- and cold-spots by district 2015

 
 

Mapping Directional Distribution

 


Standard deviation ellipses showing distribution of economically disadvantaged students
Next, I tested for similar patterns in the campus-level data. I first calculated standard deviational ellipses (SDE) for the 2004-2005 data and for the 2014-2015 data. SDEs were calculated in order to measure directional trends in the distribution of elementary school campuses with high economically disadvantaged student populations by weighting campuses by the raw count of economically disadvantaged students. The results show an increase of 77.53 square miles in area from 2005-2015 with directional shifts to the Northeast and the South.
Again, this prompted me to continue to explore patterns in the geographic distribution of economically disadvantaged students over time, and so I again calculated statistically significant hot and cold spots using the GetisOrd
Gi* statistic. The results reveal strong hotspots in the southeastern parts of Austin with developing hotspots in the far south, San Marcos ISD. This map shows a distinct coldspot to the west of Austin and a neutral line of campuses that are likely statistically insignificant due to their extreme neighbors on either side.

Campus ED GI 2015 web

 
 

Mapping Distance from Austin City Center

 
Finally, in an attempt to visualize the relationship between schools with high economically disadvantaged populations and the city center, I created a simple graph with the campus’s distance from the city center on the X axis and the campus’s economically disadvantaged rate on the Y axis. This graph displays the line for both 2004-2005 and 2014-2015 and reveals that the highest average percentage of economically disadvantaged students increased from 5 to 10 miles in 2004-2005 to over 30 miles away in 2014-2015.


Rates of student economic disadvantage by distance from center of Austin

 
 

Conclusions

 
What this data shows is that:

  1. The percentage of children living in poverty rose in the majority of the school districts in the Austin area from 2005 to 2015.
  2. The largest percentage point increases in economically disadvantaged populations from 2005 to 2015 happened in districts increasingly distant from center of Austin.
  3. The major directional shifts in childhood poverty since 2005 have been to the Northeast and to the South with a distinctive majority of childhood poverty situated in the eastern half of Austin and surrounding districts, namely Austin, Del Valle, Manor, and Bastrop.
  4. Childhood poverty in the Austin area is not randomly distributed and is distinctly clustered.

 
These changes in the geography of childhood poverty are important because they have an impact on how (and where) we aim to fight poverty. The methods and systems we have developed over the years to combat the effects of dense, urban poverty may not translate well to the sprawling suburbs. Austin is not alone among cities experiencing the suburbanization of poverty and shares many challenges with other cities nationwide. And though the trend is recent enough that we know little about its long-term effects, there are a some key areas that we know will be directly affected by the shifting geography of poverty. The suburbanization of poverty will have a direct impact on where we locate and how we implement health and human services, how we plan transportation systems, and who has access to job opportunities. The rise in suburban childhood poverty will have an effect on schools that have not historically supported large populations of economically disadvantaged students. And even as we seek to understand and meet the growing needs of these communities in real time, we must also be thinking upstream and questioning, what caused this?
 
 
Sam was an intern for Children’s Optimal Health and recently graduated from The University of Texas at Austin from the Department of Geography and the Environment. She is currently working as a Planner for the City of Austin Comprehensive Planning Division of the Planning and Zoning Department and keeps a weather eye on the shifting demographics of Austin and the modern and historic social injustices driving them.

COH welcomes Dr. Anjum Khurshid
The health care delivery and health planning landscape in the Austin MSA is changing rapidly. COH desires for itself and for the benefit of its community partners to add value to the strategic initiatives being established at the Dell Medical School (DMS) and by other governmental, nonprofit, and health-focused entities in Central Texas. COH enjoys a strong reputation as a trusted data steward, provider of high-quality geospatial analysis, and convener of action-oriented health summits.

In order to build on that reputation and enhance value to Central Texas, Children’s Optimal Health (COH) is very pleased to have contracted for the consulting services of Dr. Anjum Khurshid, MD, PhD, MPAff.

Dr. Khurshid will provide insight into better positioning COH to interface with the multiplicity of groups at DMS and elsewhere that are pursuing health-related innovations in Austin. Dr. Khurshid has the background and experience to help COH develop relationships with new and existing partners that more effectively match its capacity to emerging needs. Dr. Khurshid can also help COH access funding and technical assistance from national programs (both governmental and philanthropic) that support community-oriented, data-sharing activities.

Anjum Khurshid
How do Austin’s children fare under the 10-1 plan?
Children’s Optimal Health (COH) has prepared this series of maps in which the Austin City Council districts under the 10-1 system are overlaid on maps previously produced by COH. This was done to spark discussion on how the changes to Council could impact, and be impacted by, child health needs. The districts are delineated by color except in the map of “Medicaid-eligible children by ZIP Code” in which the districts are bordered in red. Please contact us regarding any questions you may have about these maps.

Transportation-Related Child Injury

APD_mvc_ped_bi_noCounts

APD_collisions_MV_noCounts

APD_collisions_major

Child Oral Health

Untreated_Decay

Medicaid_Eligible

Child Obesity

EcoDis_Density

BMI_SO_Density

BMI_Density_MS

BMI_Density_ES

Novel method of mapping Social Influences
Humans are social creatures, and we unconsciously ‘normalize’ our behavior to match those around us. Understanding this adaptation can help communities create policies that influence people to make healthy decisions. Based on this idea, one group of researchers in New Zealand has developed a novel method to measure visible behaviors which can help target interventions and assess their effectiveness.[1] Using “viewshed analysis,” the team quantified the number of smokers that could be seen from any given location within a study area at various times and mapped it out.[2]

The researchers only used this study to compare smoking visibility between different times of day and to compare weekdays to weekends. However, Dr. Philip Huang, the Medical Director and Health Authority for the Austin/Travis County HHS Department, sees the value in time series mapping to measure the effectiveness of health policies: “a replication of the study described in the article could definitely be of value in Austin…a pre-/post-test could be done if there is ever any policy change addressing smoking in outdoor restaurant and bar patios.” According to the research article, this method could also be “applied to measure other health-related [behaviors] including physical activity or consumption of unhealthy foods.”

Sources:

[1] Pearson, Amber L, et al. “Measuring visual exposure to smoking behaviours: a viewshed analysis of smoking at outdoor bars and cafés across a capital city’s downtown area.” BMC Public Health 2014, 14:300. doi: 10.1186/1471-2458-14-300.
[2] Badger, Emily. “Researchers have figured out how to map the social influence of public smoking.” The Washington Post Wonkblog. 10 April 2014.

"How much visual exposure to smoking we're likely to experience on a Friday night in downtown Wellington from any given viewpoint"

“How much visual exposure to smoking we’re likely to experience on a Friday night in downtown Wellington [New Zealand] from any given viewpoint”

The link between Traffic Collisions and PTSD
Many children injured in traffic collisions develop subsequent symptoms of post-traumatic stress disorder (PTSD) according to a newly published study from Sweden.[1] As many as one-third of children brought to the emergency department (ED) after sustaining a traffic injury “fulfilled diagnostic criteria” for PTSD after one month, with 23% still showing symptoms three to six months after the crash. These psychological issues are often overlooked, as the severity of psychological symptoms is unrelated to the severity of physical injuries. Risk factors for developing symptoms of PTSD include “a perceived threat to life and high levels of distress during and immediately after” a crash. While PTSD symptoms are currently under-addressed, Texas trauma researchers are studying “prevention and control strategies” to understand mental health outcomes of children after involvement in traffic collisions.[2]

In Austin several thousand children are involved in traffic collisions each year as drivers, passengers, bicyclists, and pedestrians. While post-trauma control strategies are the responsibility of medical professionals, groups leading prevention efforts must know where to target resources for a community to create optimal impact. The patterns of where these occur can be seen in maps produced by Children’s Optimal Health for the second Transportation-related Child Injury project. From 2010 to 2012 there were 7,678 collisions involving children, and 434 collisions between motor vehicles and child pedestrians or bicyclists, in the jurisdiction of the Austin Police Department (APD).

Traffic collision hotspots form a chain down IH-35 from Wells Branch Parkway south to Slaughter Lane. Other areas of concern include along highway 183, Lamar Boulevard and Rundberg Lane, Riverside Drive and Pleasant Valley Road, and along many of the city’s east-west thoroughfares.

Collisions of motor vehicles hitting child pedestrians and bicyclists more often occur within neighborhoods, with particular areas of concern being along East Riverside Drive and in the Lamar/Rundberg area. Proportionally more of these are child bicyclists in southeast Austin neighborhoods than elsewhere in the city.


Sources:

[1] “Many Children Affected by Posttraumatic Stress Disorder after Traffic Accidents.” Retrieved 30 July 2014.
[2] “The association between positive screen for future persistent posttraumatic stress symptoms and injury incident variables in the pediatric trauma care setting.” Journal of Trauma and Acute Care Surgery 2012 June 72(6):1640-6. doi: 10.1097/TA.0b013e31824a4c75. Retrieved 30 July 2014.