LBJ faculty, students present research on environment and energy, COVID and learning, equity, reproductive health at APPAM | LBJ School of Public Affairs | The University of Texas at Austin
LBJ faculty presenting at #2021APPAM: Ken Flamm; Andrew Waxman; Gordon Abner; Abigail Aiken; Miguel Pavon; Paul von Hippel
LBJ faculty presenting at APPAM March 27–29. Left to right: Ken Flamm, Andrew Waxman, Gordon Abner, Abigail Aiken, Miguel Pavon and Paul von Hippel.

 

Professors and students from the LBJ School of Public Affairs at The University of Texas at Austin will lead conversations and present research across an array of policy areas at APPAM, the Association for Public Policy Analysis and Management (APPAM) Fall Research Conference in Austin March 27–29.

The multidisciplinary research conference, themed "The Power of Inclusion: Incorporating Diverse Voices in Public Policy Analysis and Management," attracts the highest quality research on a wide variety of important current and emerging policy and management issues. LBJ scholars will present work on environmental and energy policy, sustainable development, equity, reproductive health policy, broadband, climate policy, COVID and learning, and police accreditation.

Learn more from LBJ School faculty and students at the following sessions:

Faculty

Public and Nonprofit Management and Finance

Gordon Abner, Assistant Professor

  • Panel (in person): Risk, Resilience, Discretion and Accountability: Issues in Service Delivery
    Presented with Maria-Elena Giner
  • Panel paper: Risk, Resilience, Discretion and Accountability: Issues in Service Delivery — Does Private Accountability Create Public Value: A Look at Police Accreditation
    Presented with Jennifer E. Lake

    Police accreditation represents a significant investment of time and money for departments who pursuit it and yet its impact on accountability remains understudied. To address this research gap, we conducted semi-structured interviews with leaders of nationally accredited police departments to understand what they have done differently because of accreditation. Prior work has speculated that accreditation does not change the behaviors of accredited departments because departments who choose to pursue accreditation already have the right policies in place. Our findings challenge that hypothesis. Our data reveal that while many leaders of accredited departments feel that their departments where high performing prior to accreditation, those departments still had to make significant changes to their policies in order to be incompliance with national accreditation. Accredited police departments generally need three years to come into compliance with accreditation, sometimes more. We provide concrete examples of how police departments are changing their policies and procedures in important ways that enhance accountability because of accreditation. Our findings have important implications for government oversight and government accountability more broadly.

Family and Child Policy

Abigail R.A. Aiken, Associate Professor of Public Affairs

Innovations in Science and Technology

Kenneth Flamm, Professor of Public Affairs
Presented with Tyler Baines (MPAff '21), Lillian Hatcher (MPAff '21), Gina Hinojosa (MPAff '21) and Lindsay Hodge (MPAff '23)

  • Panel paper: Broadband Infrastructure in Pandemic Times: Challenges, Opportunities, and Constraints — Onlining in the Time of COVID: Determinants of U.S. School District Teaching Responses to the Pandemic

    There has been considerable speculation that factors like COVID incidence in schools, levels of computer and internet access in student households, political sentiment, and socioeconomic characteristics of school districts may have played a significant role in school district decisions to alter teaching methods during the COVID-19 pandemic. However, no research studies have yet comprehensively analyzed the factors that motivated teaching modality decisions.

    Using a variety of large scale datasets, we created a database covering more than 12,000 U.S. school districts, including district-level decisions about teaching modes to be used during the 2020–21 school year, socioeconomic demographics, school funding levels, past standardized test scores, student proficiency in main subject areas, COVID incidence, broadband adoption, price and quality measures, and 2020 voter decisions.

    Most of the included datasets were available at the school district level. However, data on COVID incidence; broadband price, availability, and usage; and 2020 election information were available at the county and ZIP code levels. For these variables, we constructed household-weighted estimates for school districts by using a Census crosswalk linked to FCC census block household estimates to calculate numbers of 2019 households in every intersection of a county or ZIP code with a school district. We then constructed household-weighted averages of county or ZIP code averages for school districts containing portions of those ZIP codes or counties, to produce school district estimates.

    We then estimated a linear probability model of a school district's choice to switch to remote learning and entirely online teaching methods. Our preliminary conclusion was that broadband infrastructure quality seemed to have a substantively and statistically significant impact on a school district's decision to switch to online teaching methods. Surprisingly, cumulative COVID case and death rates had very small and statistically insignificant effects on the probability of a school district choosing online teaching modes.

    A measure of broadband adoption/usage of 25/3+ service in 2019 had a positive and statistically significant effect, as did an estimate of 25/3+ broadband availability. School district size (student enrollment) also has a positive and statistically significant effect. When a political sentiment variable is added to the model, it had a substantial substantive effect in predicting a shift to online teaching and was highly statistically significant. With political sentiment included the contribution of 25/3+ broadband adoption to onlining probability within a district falls by 2/3 and is no longer statistically significant, but the size and significance of the broadband availability and school district enrollment effects persist. In all models, COVID incidence in a school district seemed to play little substantive role in shaping the school district response. Demographic differences across school districts seem to have had only small effects on school district policy. We conclude by discussing possible implications for future public health crisis responses.

    Related: LBJ School team makes research discoveries, places 2nd at Microsoft and ODI's Education Open Data Challenge (May 6, 2021)

Natural Resource, Energy, and Environmental Policy

Miguel Pavon, Adjunct Professor of Public Affairs
Presented with Maria-Elena Giner

  • Panel paper: Causes and Consequences of Pollution Exposure on Air, Water and Land — Implementing Water and Wastewater Infrastructure in Underserved Communities in Texas Along the U.S./Mexico Border: What Did We Learn?

    In Texas, a unique phenomenon occurred in the early '90s similar to that in developing countries: the formation of informal settlements called colonias that lacked the most basic municipal infrastructure such as water and wastewater. Consequently, incidences of Hepatitis A and tuberculosis in this region exceeded twice the average values for the state of Texas. Almost 2 billion dollars was authorized by federal and state agencies for first-time water and wastewater services for over 330,000 residents. The objective of this research was to assess at a high-level whether the funds were distributed equitably across the 31 counties adjacent to the Texas-Mexico border, identifying where the need continues, whether population growth occurred, and compiling programmatic and technical lessons learned. The research design included a mixed-method approach that incorporates over 100 interviews with agency program managers and utilities and both grant recipients and non-grant recipients. Also, a geographical information system was used to apply two regressions on the allocation of funds for the period 1995 through 2017.

    The results show that the funding was extremely successful, resulting in over 75% of the population receiving wastewater services. The funds were generally distributed equitably amongst the counties with smaller populations and most counties received the adequate amount of funding as compared to others. Yet those counties with the larger cities that extended service for the colonias were overfunded as both city and colonia needs were met. This could serve as a carrot for cities with much more institutional capacity and operational efficiencies provide a regional solution. Also, by 2018 the expected reduction in need was achieved in most counties based on funding received and original need. Yet, the largest need continues in two of the larger counties where much capacity exists. Even so, progress is underway. The challenge will be in the smaller more isolated colonias where overcoming the barriers to service will be costly.

    Important lessons were learned, specifically, the role funding agencies had in building capacity in the utilities, timely disbursement of funds, ensuring household connections were completed quickly, and tracking progress. Unintended consequences included oversized facilities as population growth did not result as projected in more rural colonias. Also, a strategy is required for replacing aging infrastructure. The relevance of this research is two-fold. First, it documents the combined results across multiple agencies and identifies where need continues. Second, the lessons learned can serve as a model for other countries in transition working towards increasing water and sanitation infrastructure, a United Nations Sustainable Development Goal 6 (SDG6). Colonias conditions can be like those of developing countries.

Education

Paul von Hippel, Associate Professor of Public Affairs

  • Panel paper: The Unequal Effects of COVID on Learning and Behavioral Outcomes — The Impact of COVID-19 School Closures and Policies on Student Achievement and Inequality

    The COVID-19 pandemic has clearly reduced children's learning, yet our current portrait of COVID learning loss is fragmentary and distorted. Only a small and unrepresentative fraction of students have taken standardized tests since the pandemic began, and some remote tests taken by young children were invalidated by parental assistance. Three evaluations have reported that COVID learning loss has been greatest among young children, poor children, black children, Hispanic children, and children in rural areas, yet two evaluations also suggest that learning loss, though substantial in math, has been practically nil in reading (Kuhfeld et al., 2020a; Renaissance Learning, 2020; but not Curriculum Associates, 2020)—a result that seems hard to believe. In addition, we know practically nothing about the impact on learning of school policies on online, hybrid, and in-person learning (Harris et al., 2020).

    In this project, we estimate the distribution of COVID learning loss in reading, math, and science scores from the Measures of Academic Progress (MAP) test. In 3rd to 8th grade, MAP was taken by 4.2 million students in fall 2019, before the pandemic. MAP was taken by just 2.5 million students during the pandemic fall of 2020, but we expect the number of tested students to increase in spring 2021 and return to pre-pandemic levels in fall of 2021.

    To estimate policy impacts, we will merge MAP scores with school characteristics and pandemic policies from a variety of sources, including school and district websites scraped during the pandemic. We will use multiple imputation to fill in missing test scores conditionally on scores from before and after the pandemic, school policies, and school and child characteristics.

    These are our specific questions:

    • On average, how much did the pandemic reduce student achievement over the short and medium term?

    Specifically:

    • In fall 2020, how much lower was student achievement than it would have been if the pandemic had not occurred and schools had stayed open? How much does the answer to this question change when missing scores are imputed?
    • To what degree will the impact of the pandemic persist or fade on later tests given in spring and fall 2021?
    • How much did COVID learning and loss vary across districts, schools, and students?
    • How much less was learned by students from disadvantaged groups?
    • Did districts with poorer broadband access have less learning during COVID closures?
    • Did younger children suffer greater learning losses than older children?
    • How much did school and district policies affect learning?
    • Did schools opening more in-person see less learning loss than schools opening remotely or in a hybrid fashion?
    • Did learning losses vary by school sector (public, private, charter)?
    • Was the variance associated with school policies large or small compared with the variance associated with school and student demographics?

Natural Resource, Energy, and Environmental Policy

Andrew Waxman, Assistant Professor

  • Panel (virtual): Applications of Big Data in Environmental and Energy Policy Analysis
  • Panel (in person): Causes and Consequences of Pollution Exposure on Air, Water and Land

    After decades of research into its health and economic impacts, pollution of air and water remain a persistent challenge through particulate matter, lead, ozone, toxics and a range of other chemicals. Economics proposes a range of solutions to this problem many of which pose political or implementation challenges. The research proposed for this panel engages broadly across the literature to examine the health and economic consequences of pollution, explore the challenges to implementation of pollution regulation and understand the potential unintended consequences of policy. The proposed work uses the localized exposure of populations to leaded gasoline from NASCAR races to understand its substantial educational impacts. Separate work seeks to gauge the unintended consequences of local regulator's behavior to comply with National Ambient Air Quality Standards. Another paper uses air sensors attached to Google Street View vehicles to examine variation in the localized exposure to pollution and understand the pollution co-benefits of road pricing. A third paper considers the impact of the Resource Recovery and Control Act on toxic waste releases and the subsequent effects on health and housing. The proposed papers use state-of-the-art methods and datasets to better understand previously unaddressed aspect of pollution costs and regulation to better inform future policy.

  • Panel paper: Causes and Consequences of Pollution Exposure on Air, Water and Land — Using Big Data to Estimate the Environmental Benefits of Congestion Pricing in the Los Angeles Area

    To estimate the effect of traffic congestion on local air quality, we combine two unique sources of big-data: First, real-time data on speed and flow of vehicles in California freeways from the California Performance Measurement System (PeMS). PeMS collects real-time speed and flow data from finely spaced loop detectors on California freeways. Second, real-time spatially and temporally disaggregated data on a variety of air pollutants are obtained from Aclima. The Aclima/Google Street View vehicle mounted pollution monitors provide unprecedented spatial and temporal detail that stationary EPA detectors cannot provide. They also avoid potential substantial measurement error from satellite-based estimates. The data provide observations every three seconds, corresponding to many repeated observations within a single city block, but also including urban freeways. Using these data helps to remove potential concerns of measurement error in pollution due to extrapolation or aggregation, which prior work has shown may be substantial when prevailing winds disperse pollution asymmetrically in space (Sullivan, 2017; Heiblich, et al., 2018).

    We use the combination of Aclima car-based readings and EPA station monitors to estimate detailed models of the production, dispersion and conversion of nitrogen oxides (NOx) into ozone (O3). We also demonstrate substation heterogeneity in road-level pollution that deviates from EPA monitoring levels demonstrating that variation in the nature of non-point pollution from vehicles can translate into differential exposure levels that are not captured well by ground monitors or even predictions based on satellite data.

    Our preliminary results show the dramatic effects of idling at speeds at or below 5 MPH on localized pollution levels: NO levels can be as much as 30 times higher than at normal highway speeds, and ozone levels can be three times as high. Pollution levels generally decline past the idling point as engine combustion becomes more efficient. As shown below in Table 1, regression results show that for a 10% increase in speeds away from idling, NO2 levels fall by 0.33% and O3 by 0.16%. Also of note is the persistence of NO2 pollution reflected by the positive and significant coefficients from the hours prior as well as the effect of slower speeds during the hour before observations are taken. All of this provides a basis for demonstrating the effects of low speeds on pollution levels as well as their persistence in downtown areas.

    We build on these results in further analysis by including EPA monitor data and PeMS freeway speeds (when the vehicle is on the highway) to control for background pollution levels and to address concerns regarding selection bias based on vehicle location and speed.

    With these estimates in hand, we simulate partial equilibrium travel demand responses to time-varying cordon pricing around central corridors in Los Angeles that would vary by congestion levels, weather conditions, time-of-day. These results are key to informing new Low Emission Zones, which are becoming commonplace in European Cities (e.g., London, Milan, Stockholm) and have been proposed for Santa Monica.

 

Students

Innovations in Science and Technology

Tyler Baines (MPAff '21), Lillian Hatcher (MPAff '21), Gina Hinojosa (MPAff '21) and Lindsay Hodge (MPAff '23)
(Presented with Professor Kenneth Flamm

Social Equity and Race; Education

Francisco A. Castellanos-Sosa, Ph.D. candidate

Natural Resource, Energy, and Environmental Policy

Maria-Elena Giner, Ph.D., Commissioner for the International Boundary and Water Commission (IBWC)
Presented with Miguel Pavon, Adjunct Professor of Public Affairs

  • Panel paper: Causes and Consequences of Pollution Exposure on Air, Water and Land — Implementing Water and Wastewater Infrastructure in Underserved Communities in Texas Along the U.S./Mexico Border: What Did We Learn?

    In Texas, a unique phenomenon occurred in the early '90s similar to that in developing countries: the formation of informal settlements called colonias that lacked the most basic municipal infrastructure such as water and wastewater. Consequently, incidences of Hepatitis A and tuberculosis in this region exceeded twice the average values for the state of Texas. Almost $2 billion was authorized by federal and state agencies for first-time water and wastewater services for over 330,000 residents. The objective of this research was to assess at a high-level whether the funds were distributed equitably across the 31 counties adjacent to the Texas-Mexico border, identifying where the need continues, whether population growth occurred, and compiling programmatic and technical lessons learned. The research design included a mixed-method approach that incorporates over 100 interviews with agency program managers and utilities and both grant recipients and non-grant recipients. Also, a geographical information system was used to apply two regressions on the allocation of funds for the period 1995 through 2017.

    The results show that the funding was extremely successful, resulting in over 75% of the population receiving wastewater services. The funds were generally distributed equitably amongst the counties with smaller populations and most counties received the adequate amount of funding as compared to others. Yet those counties with the larger cities that extended service for the colonias were overfunded as both city and colonia needs were met. This could serve as a carrot for cities with much more institutional capacity and operational efficiencies provide a regional solution. Also, by 2018 the expected reduction in need was achieved in most counties based on funding received and original need. Yet, the largest need continues in two of the larger counties where much capacity exists. Even so, progress is underway. The challenge will be in the smaller more isolated colonias where overcoming the barriers to service will be costly.

    Important lessons were learned, specifically, the role funding agencies had in building capacity in the utilities, timely disbursement of funds, ensuring household connections were completed quickly, and tracking progress. Unintended consequences included oversized facilities as population growth did not result as projected in more rural colonias. Also, a strategy is required for replacing aging infrastructure. The relevance of this research is two-fold. First, it documents the combined results across multiple agencies and identifies where need continues. Second, the lessons learned can serve as a model for other countries in transition working towards increasing water and sanitation infrastructure, a United Nations Sustainable Development Goal 6 (SDG6). Colonias conditions can be like those of developing countries.

D. Cale Reeves (Ph.D. '19), Postdoctoral Research Fellow and Lecturer
Presented with Ariane Beck, Research Fellow

  • Panel paper: Local Sustainable Development: Policy Design, Program Participation and Distributional Impact — Who Makes It in the Long Run: Estimating Equity Impacts in Policy and Land Development Scenario Projections

    Time horizons at the intersection of urban development and climate change tend to run long. For example, about half of buildings are older than 30 years, and IPCC projections The Intergovernmental Panel on Climate Change projections provide guidance to policymakers through the year 2100. As urban policy makers and planners craft long term sustainability solutions for their cities, it is increasingly clear that they must also place equity impacts at the fore of their decision-making. However, complex individual interactions among a dynamic urban population make predicting long-term equity impacts difficult.

    Here, we use an agent-based model of technology diffusion overlaid on long-term land use planning scenarios to incorporate complexity and harness population dynamics. This approach contributes to an understanding of the interplay between regulatory (building code changes) and redistributive (technology subsidies) policy and a range of individual-level attributes including wealth, home ownership, and home tenure. This research sheds light on how city-scale sustainability improvements manifest over time, but also illuminates who ­— at an individual-level — shoulders the burden of these improvements. Understanding the long-term equity impacts of sustainable city planning enables policy makers to co-develop support systems aimed at alleviating disproportionate burdens that arise.

Public and Non-Profit Management and Finance

Jennifer E. Lake, Ph.D. student
Presented with Gordon Abner, Assistant Professor

  • Panel paper: Risk, Resilience, Discretion and Accountability: Issues in Service Delivery — Does Private Accountability Create Public Value: A Look at Police Accreditation

    Police accreditation represents a significant investment of time and money for departments who pursuit it and yet its impact on accountability remains understudied. To address this research gap, we conducted semi-structured interviews with leaders of nationally accredited police departments to understand what they have done differently because of accreditation. Prior work has speculated that accreditation does not change the behaviors of accredited departments because departments who choose to pursue accreditation already have the right policies in place. Our findings challenge that hypothesis. Our data reveal that while many leaders of accredited departments feel that their departments where high performing prior to accreditation, those departments still had to make significant changes to their policies in order to be incompliance with national accreditation. Accredited police departments generally need three years to come into compliance with accreditation, sometimes more. We provide concrete examples of how police departments are changing their policies and procedures in important ways that enhance accountability because of accreditation. Our findings have important implications for government oversight and government accountability more broadly.

Politics, Media, and the Policy Process

Megan M. Morris (MGPS '21) and Mark Hand, Ph.D.

  • Panel paper: Influencing Public Awareness: Policy Design By Tool Selection, Framing, and Narratives — Stories of a Crisis: Predicting the Adoption of Diverging Policy Narratives in an Extreme Climate Event

    In February 2021, an extreme winter weather event blanketed Texas and much of the South with unusually low temperatures, snow and ice. Power grids shut down and two-thirds of Texans were left without power and heat, some for multiple days. As the storm wore on and while Texans dealt with the fallout in the days and weeks after, diverging narratives on the causes and effects of the winter storm began to emerge in local, state and national media, and on social media sites. This paper studies the emerging and diverging narratives of the storm, focusing on what factors influenced their adoption among the general public, and why.

    The Narrative Policy Framework (NPF), a policy process framework, emphasizes the importance of narratives in the policy creation cycle. The NPF suggests that policymakers develop and promote narratives with generalizable content that are both persuasive and coherent with their proposed policy solutions, and that these narratives influence the eventual policy outcomes. The NPF allows researchers to study why some narratives are adopted while others fail to captivate or convince audiences. The present study uses the NPF to explore which policy narratives were more compelling to audience members following the Texas winter storms. Using text clustering on answers to questions about core narrative elements from the NPF (including characters, plot, and moral), we identify the dominant narratives that emerged immediately after the storm. We then ask how demographic characteristics, first-hand experience of the storm, and partisanship are correlated to the narrative elements that people adopt and the sources of narration that they look to. The study utilizes a novel data set from the Texas Politics Project and UT Energy Institute survey that was conducted in March 2021. We hypothesize that political partisanship and choice of information sources (narrators) are strong predictors of narrative elements that policy audiences adopt. However, we suspect that these may become less important if a person experienced extreme hardship during the winter storms, such that first-hand experience becomes a more important factor in the adoption of policy narratives. In an era in which public understanding of crises is shaped online and in real time, this paper reveals the pathways that shape the public formation of narratives, and what first-hand experiences can shake them.