Assignment 01: Project Proposal and Problem Statement

1 - Introduction: Overview of the City and Policy Issue

Baltimore, Maryland, serves as the focal city for this project, where the pressing policy issue of youth involvement in crime, particularly among children and adolescents, is explored. Recent crime data from Baltimore’s open data portal reveals alarming statistics, including instances of crimes committed by children as young as 9 years old. This issue is linked to systemic challenges such as the lack of extracurricular programs and community resources for under-served youth. Research shows that accessible, enriching programs for the youth can significantly reduce their likelihood of engaging in criminal activity, especially during after-school hours. As a long-time advocate of educational equity through the exposure of programming languages such as Python and Java for Baltimore’s students, I have personally witnessed the positive impact of skills-based learning on children and their futures. This project seeks to examine the relationship between youth crime and the availability of extracurricular activities, with a focus on education and programming as a means of intervention.

2 - Relevant Datasets from Open Data Portal

For this project, several datasets from Baltimore’s open data portal are relevant and will be analyzed to gain a deeper understanding of the issue. The Baltimore Crime Dataset, updated weekly, includes important details such as the nature of the crimes, the ages of the offenders, and the time and location of the offenses. Additionally, the Per Pupil Expenditures and School Attendance Data available through Maryland’s Public Use Data Downloads provide insights into the allocation of educational resources and student engagement. The inclusion of crime time data allows for an exploration of potential connections between youth crime and after-school hours, supporting the hypothesis that a lack of structured programs during this period may contribute to the problem. By cross-referencing this with school attendance and funding data, this project can explore whether schools with lower attendance rates and fewer resources see higher levels of youth involvement in crime.

3 - Problematic Conditions: Description of the Issues and Context

Youth crime in Baltimore, particularly among elementary and middle school-aged children, has proved to be a growing concern. Crime data shows that children as young as 9 years old are involved in offenses such as assault, theft, and property damage. These incidents point to broader social issues, including economic disparity, a lack of social services, and limited access to extracurricular activities that can engage youth in productive, skill-building activities. The reduction in funding for after-school programs and community centers has likely influenced this issue, leaving many children without constructive outlets for their time. National research from NIH has explored the importance of structured after-school programs in reducing delinquency. Without them, children in under-served areas are more likely to be drawn into criminal activity due to boredom, peer pressure, or economic need. If left unaddressed, these conditions could result in long-term consequences for both the children involved and the community at large.

4 - Policy Relevance and City Action

Baltimore City has acknowledged the need for crime reduction strategies through initiatives like the Baltimore Youth Violence Prevention Plan and other crime-reduction efforts. But these policies often focus on intervention rather than prevention (addressing the consequences rather than the root causes of youth crime). A key element missing in these initiatives is the emphasis on extracurricular and educational opportunities as preventative measures. By analyzing the relationship between crime and the availability of extracurricular programs, this project will highlight the need for city policies that invest in after-school programs, community centers, and educational equity. Baltimore could draw inspiration from cities like Boston, which have successfully implemented after-school programs as part of its Community Learning Initiative, with positive results in reducing youth crime.

5 - Concerns and Stakeholder Impacts: Political Disagreements and Affected Organizations

The issue of youth crime in Baltimore raises several concerns and disagreements among stakeholders. On one hand, law enforcement agencies and policymakers often focus on increasing security measures and punishments to addressing youth crime. On the other hand, educators, community leaders, and nonprofit organizations advocate for increased funding for educational programs, extracurricular activities, and community services. Political disagreements center around the allocation of resources, with some stakeholders prioritizing policing and crime reduction, while others push for more preventive measures, such as education reform and youth programs. The key stakeholders impacted by this issue include the Baltimore City Public Schools, community organizations, local law enforcement, and, most importantly, the children and families in under-served communities. Each stakeholder group has a different perspective on how to address youth crime, which all contribute to the complexity of the issue.

6 - Proposed Solution: Suggested Approach and Implementation Plan

The proposed solution to the problem of youth crime in Baltimore is to develop a citywide initiative that prioritizes the funding and expansion of after-school programs, particularly those focused on skill-building, technology education, and mentorship. By increasing investment in extracurricular activities that expose children to different skills (such as programming languages, or STEM education), Baltimore can provide young people with alternatives to criminal activity. The initiative would require collaboration between the Baltimore City Public Schools, local nonprofit organizations, community centers, and private sector partners who are willing to fund and support these programs. Data on school attendance, crime rates, and program participation can be tracked and analyzed to measure the effectiveness of this initiative. By targeting schools and neighborhoods with high crime rates and low educational funding, the program can focus resources where they are most needed. Additionally, the city should explore opportunities for public-private partnerships to sustain these programs in the long term.

7 - Conclusion

In conclusion, addressing youth crime in Baltimore requires a multifaceted approach that goes beyond law enforcement and punishment. By focusing on preventive measures such as extracurricular programs and educational equity, Baltimore can reduce youth involvement in crime and provide children with the tools they need to succeed. Through data-driven analysis of crime trends, school attendance, and the availability of after-school programs, this project will make the case for a renewed investment in Baltimore’s youth, ensuring that they have the opportunities and resources necessary to thrive.

Assignment 02: Civic Tech Adaptation

1 - How does your project fit within the framework of the seven attributes of civic technology, the types of civic tech tools available, the three main pillars of civic technology, or the four segments of the civic tech sector?

Sync-the-City fits seamlessly within the broader framework of civic technology by enhancing collaboration between nonprofit organizations in Baltimore, which serves the community more efficiently. It aligns with the three main pillars of civic technology: transparency and accountability, facilitating citizen-government interaction, and digital tools to facilitate everyday life. By providing a publicly accessible platform that visualizes nonprofit locations and their partnerships, Sync-the-City improves transparency, showing which services are available and where. This not only benefits nonprofit leaders but also enables local governments to collaborate with nonprofits to optimize service delivery. Lastly, the platform leverages digital tools such as mapping and network graphs, simplifying collaboration between organizations by making it easier to visualize partnerships and coordinate efforts. By enabling a data-driven approach to collaboration, Sync-the-City exemplifies the potential of civic tech to solve community problems.

2 - For the existing civic tech initiative that you have chosen, what was the original purpose? In other words, what is the problem that it was intended to solve?

The original purpose of Sync-the-City was to address the communication gap between nonprofits in Baltimore. Many nonprofits often work in isolation and are unaware of what other organizations are doing in their communities, which leads to missed opportunities for collaboration and duplicated efforts. By centralizing information about nonprofit activities and partnerships, Sync-the-City allows organizations to see who is working on similar initiatives, where they are located, and how they can collaborate more effectively. Additionally, the platform supports funding institutions by helping them identify areas with service gaps, enabling them to make informed decisions about where to invest resources. This collaborative approach amplifies the impact that organizations can have on their communities, fostering efficiency in resource use and improving service coverage.

3 - Describe the type of data that was required for it to function, and/or the type of data it collected or generated.

Sync-the-City relies on several key data types to function effectively. The primary dataset consists of nonprofit data—including the organization’s name, location, mission, and partnerships—sourced from publicly available IRS records. In addition, the platform utilizes geospatial data to map the locations of nonprofits in Baltimore, which helps users visualize where services are being provided and identify areas with limited coverage. The platform also gathers collaboration data to show existing partnerships between nonprofits. This data is visualized in a network graph, which helps users see where collaboration is occurring and where further partnerships could be beneficial. The platform generates valuable insights into service coverage, gaps in nonprofit collaboration, and areas for further resource allocation, enabling better decision-making for both nonprofits and funders.

4 - What type of technology is used for the initiative (Python, CSS, D3.js, SQL, ArcGIS, javascript, etc.)?

The technology stack for Sync-the-City consists primarily of R and Shiny. R is used for backend data processing, while Shiny allows users to dynamically engage with nonprofit data. HTML is used to structure the content on the web interface, and Leaflet is used for geospatial data visualization, allowing users to explore nonprofit locations on a map. Lastly, Docker is used to package and deploy the application across different environments, ensuring that the platform is easily accessible and scalable. This combination of technologies creates a powerful and interactive tool that enables nonprofits and funders to visualize nonprofit activity and collaborate based on the shared data.

5 - What are the changes in design that would be necessary for the civic tech initiative to work for your policy problem?

To adapt Sync-the-City for addressing youth crime prevention in Baltimore, several key design changes are necessary. The platform would need to integrate Baltimore crime data, specifically focusing on areas with high juvenile crime rates. This integration would allow nonprofits to better identify neighborhoods that are underserved in terms of after-school programs and other resources for at-risk youth. In addition to helping nonprofits target these high-need areas, the platform could also introduce features that directly connect parents and students with local after-school programs. By expanding the user base to include families, Sync-the-City could serve as a resource discovery tool, enabling parents to search for programs based on location, income level, and their children’s interests.

To maximize impact, Sync-the-City could prioritize programs that operate during peak juvenile crime hours, typically from 3 PM to 6 PM. This would help channel resources into after-school programs designed to engage young people during these high-risk times. The platform could also offer data-driven recommendations that foster collaboration between organizations focused on youth education, mentorship, and development. And this collaboration would allow nonprofits to pool their resources and address the root causes of youth crime more effectively. This type of approach aligns with the idea of using collective intelligence technologies to enable groups of organizations to collaborate intelligently for a common goal (Mačiulienė and Skaržauskienė, 2020).

In addition, the platform’s map functionality could be enhanced to highlight hotspots for youth crime and identify deserts, where little to no nonprofit services are available. This information would guide the strategic deployment of new resources and programs. And by incorporating a parent/student-facing interface, the platform could allow families to create profiles, filter programs based on factors like price, location, and activity type, and even find scholarships or other forms of financial assistance. This personalized approach would ensure that more families are connected to the right resources, contributing to crime prevention by keeping youth engaged in structured, enriching activities.

Reference: Mačiulienė, Monika and Aelita Skaržauskienė. “Building the Capacities of Civic Tech Communities Through Digital Data Analytics.” Journal of Innovation & Knowledge, Volume 5, no. 4, (2020): 244-250.

Assignment 03: Policy Literature Review

Gottfredson, Denise C., Stephanie A. Gerstenblith, David A. Soulé, Shannon C. Womer, and Shaoli Lu. “Do after school programs reduce delinquency?.” Prevention science 5 (2004): 253-266.

This study investigates the impact of After School Programs (ASPs) on delinquent behavior among youth, focusing on Maryland during the 1999-2000 school year. The key concepts in the study are After School Programs and delinquent behavior. ASPs are defined as structured programs offering youth development and skill-building activities, while delinquent behavior is defined as illegal or antisocial behavior typically resulting in involvement with the juvenile justice system. The authors hypothesize that ASP participation reduces delinquent behavior, particularly in programs that emphasize social skills and character development.

The variables used to operationalize these concepts include ASP participation as the independent variable (whether or not students attended ASPs), and delinquent behavior as the dependent variable, measured through self-reports and school records of delinquent acts. Additional variables considered include peer associations, time spent unsupervised, and intentions not to use drugs.

The unit of analysis is the individual students who participated in ASPs, focusing on both middle-school-aged and elementary-aged youth. The study uses both nominal and interval levels of measurement to assess delinquent behavior, such as whether students engaged in specific behaviors (nominal) and how frequently they engaged in them (interval). To establish causality, the study employs a quasi-experimental design with comparison groups and statistical controls for factors like previous behavior, unsupervised time, and peer influence.

The findings indicate that ASPs reduced delinquent behavior among middle-school-aged youth but not elementary-aged youth. The reduction in delinquency was attributed to improved peer associations and increased intentions not to use drugs, rather than decreases in unsupervised time or increased participation in constructive activities. As noted in Chapter 5, the authors demonstrate that “newness does not closely correspond to fitness for purpose” (p. 61), highlighting that the ASPs’ success lay not in innovation, but in addressing youth needs through practical, effective programming. Additionally, Chapter 9 underscores the importance of “policy implementation” as the critical area where technology and strategy can make a difference (p. 101). This ASP study shows that consistent, well-implemented programs drive better results in reducing delinquency.

This article is highly relevant to my policy problem of youth crime prevention, as it demonstrates how targeted, structured ASPs focusing on social skills and character development can have a meaningful impact on reducing delinquency. For my specific focus on Baltimore, adopting a similar approach could provide evidence to policymakers that investing in structured ASPs with a focus on social skills can reduce youth crime, especially during after-school hours, when juvenile crime is most likely to occur.

Goldschmidt, Pete, Denise Huang, and Marjorie Chinen. The long-term effects of after-school programming on educational adjustment and juvenile crime: A study of the LA’s BEST after-school program. Technical report, National Center for Research on Evaluation, Standards, and Student Testing (CRESST), Los Angeles, CA, 2007.

This study examines the long-term effects of the LA’s BEST after-school program on educational attainment and juvenile crime, utilizing a longitudinal dataset of approximately 6,000 students over eight years. The key concepts in the study are after-school programming and juvenile crime. After-school programming is defined as organized, supervised activities designed to engage students beyond regular school hours. Juvenile crime is measured as criminal activities committed by individuals under the age of 18, with specific attention to felonies and misdemeanors.

The variables used to operationalize these concepts include participation rates in LA’s BEST as the independent variable and incidences of juvenile crime as the dependent variable. Additionally, the study uses variables related to educational attainment, such as standardized test scores and school performance, to explore broader outcomes. The study’s hypotheses center on the assumption that higher participation in after-school programs will lead to lower rates of juvenile crime and potentially improved academic outcomes.

The unit of analysis is the individual student, with data collected over an eight-year period. The level of measurement for juvenile crime is nominal, categorized as incidences of criminal behavior such as felonies and misdemeanors. Educational attainment is measured on an interval scale based on test scores and grades. The study establishes causality by employing longitudinal analyses, controlling for confounding variables like prior academic performance, socioeconomic status, and neighborhood crime rates.

The findings suggest that students who participated more frequently in LA’s BEST had significantly lower incidences of juvenile crime. However, the impact on educational achievement was less conclusive. The study also conducted a cost-benefit analysis, finding that every dollar invested in the program resulted in an average saving of $2.50 in juvenile crime costs. As Chapter 9 suggests, policy levers such as funding and incentives play a crucial role in the successful implementation of programs (p. 101). The program’s results align with Chapter 5, where it is noted that “innovation does not need to be cutting-edge, but rather strategically targeted to solve real problems” (p. 61). LA’s BEST shows how a well-structured program can make measurable, long-term impacts without relying on groundbreaking technologies or methods.

The cost-benefit analysis in this article is particularly relevant to my policy focus on youth crime prevention in Baltimore. Policymakers could use the data-driven insights from LA’s BEST to prioritize funding for after-school programs, supported by the findings in Chapter 9, where policy implementation is shown to be crucial for achieving broader societal outcomes (p. 101). Such programs can significantly reduce juvenile crime while providing financial savings to the community.

Gottfredson, Denise C., Amanda Cross, and David A. Soulé. “Distinguishing characteristics of effective and ineffective after‐school programs to prevent delinquency and victimization.” Criminology & Public Policy 6, no. 2 (2007): 289-318.

This study examines the characteristics of 35 After-School Programs (ASPs) that are predictive of reducing delinquency and victimization. The study focuses on how program structure, staffing, and other variables contribute to the effectiveness of ASPs in preventing problem behaviors among participants. The authors employ multi-level modeling techniques to control for individual-level predictors of problem behavior, including substance use, delinquency, and victimization, while also accounting for the composition of the ASPs themselves.

Key concepts in this study include structured programming, staff education, and program size, which are defined as important predictors of behavior outcomes. Delinquency and victimization are the dependent variables, operationalized through self-reported behaviors and incidents of victimization among the youth participants. The authors hypothesize that smaller programs, those with more structured activities, and those with better-educated staff will show reductions in delinquency and victimization.

The unit of analysis is the after-school programs themselves, although individual-level data on participants is also considered. The level of measurement for delinquency and victimization is nominal, based on whether or not youth engage in or experience particular behaviors. The study establishes causality through statistical controls for individual characteristics such as age, gender, and socioeconomic background, as well as program-level factors like staff education and gender composition.

The findings indicate that ASPs with smaller enrollment sizes, highly structured programming, and better-educated staff had the most success in reducing delinquency and victimization. As Chapter 9 points out, effective policy relies heavily on implementation, and the success of these ASPs shows that careful attention to staffing and program design during implementation leads to better outcomes (p. 101). This study also highlights that “fitness for purpose” (p. 61), as discussed in Chapter 5, is more important than pursuing novelty, underscoring how structured and well-planned programs achieve better results.

This article is particularly relevant to my policy focus because it emphasizes the critical importance of program structure and staff qualifications in achieving successful outcomes. By incorporating the recommendations from this study into Baltimore’s after-school programs, policymakers can ensure that programs are not only available but also effective in preventing delinquency. As discussed in Chapter 9, understanding the policy levers that can be applied (such as: improved staffing and structured programming) is essential to improving outcomes (p. 101).

Glisson, Charles, and Sonja K. Schoenwald. “The ARC organizational and community intervention strategy for implementing evidence-based children’s mental health treatments.” Mental health services research 7 (2005): 243-259.

This article discusses the ARC model (Availability, Responsiveness, and Continuity), a framework designed to implement effective mental health treatments for children in community practice settings, particularly for delinquent youth. The key concepts in the study include organizational development, interorganizational collaboration, and evidence-based mental health treatment, specifically targeting the factors that influence the successful implementation of children’s services. The authors focus on how ARC integrates components from organizational theory, diffusion of innovation, and technology transfer to address systemic issues in community-based mental health services.

The variables used to operationalize these concepts include the implementation of Multisystemic Therapy (MST), an evidence-based intervention, as the independent variable, while the dependent variable includes youth outcomes such as delinquency reduction and access to mental health services. The study hypothesizes that the implementation of the ARC model, particularly in underserved and rural areas, will enhance the availability and quality of mental health services for youth at risk of delinquency.

The unit of analysis is the community organizations implementing MST through the ARC model, focusing on extremely rural, impoverished areas in Tennessee. The level of measurement for the outcomes is nominal for delinquent behavior (whether youth were involved in delinquency) and interval for service accessibility and engagement. The study establishes causality by assessing the effects of the ARC model on the implementation fidelity of MST, utilizing statistical controls for external variables like community resources and organizational capacity.

The findings indicate that the ARC model significantly improved the implementation of MST in rural communities, leading to better mental health outcomes and reductions in delinquency among high-risk youth. As Chapter 9 emphasizes, policy implementation is a key area where technology can play a supportive role (p. 101). Additionally, Chapter 5 reinforces the idea that “innovation isn’t necessarily about what’s new, but what works” (p. 61). The ARC model exemplifies how structured, community-based interventions can improve both mental health services and reduce youth crime by focusing on practical, achievable goals rather than technological advancements.

Assignment 04: Part I-APIs and Open Data and Smart Data

1 - Describe your proposed initiative. Use the concepts from the readings to fully describe the stage of the policy process for the initiative, the methods, and the change that would be produced with a successful intervention. Describe the ways in which the data you select for this assignment will contribute to the assessment of conditions, and effective implementation of your solution.

My proposed initiative aims to reduce youth crime rates in Baltimore by expanding after-school programs targeted at at-risk populations. The initiative is currently in the agenda-setting stage of the policy process, as discussed in The Civic Technologists’ Practice Guide, where the focus is on identifying the problem, gathering evidence, and proposing feasible solutions (Bouchet, 2022, 100). This stage is crucial for framing the issue of youth crime as a community challenge and leveraging open data to develop targeted, effective interventions.

The methods for implementation will involve collaboration between community organizations, schools, and local governments to broaden access to after-school programs in high-crime areas. As emphasized in Chapter 5 of The Civic Technologists’ Practice Guide, the focus is on effectiveness over innovation, ensuring that the interventions address the root causes of youth crime, such as unsupervised time after school and limited access to community resources (Bouchet, 2022, 61). By using real-time data on youth crime, school performance, and poverty from Baltimore’s open data portal, the initiative can tailor its efforts to the neighborhoods most in need of support.

The data collected through APIs will serve as an essential tool for assessing current conditions and monitoring the success of the intervention. For example, crime data will identify hotspots, while school enrollment trends will highlight areas where students are disengaging. As Bouchet notes, data-driven decision-making allows civic technologists to continuously adjust their approach, ensuring that interventions remain effective over time (Bouchet, 2022, 92). This dynamic, responsive model is critical for producing long-term change, as the initiative can evolve based on the data, ensuring that programs are scalable and adaptable to meet the needs of Baltimore’s youth.

2 - Select at least five different indicators from your city’s open data portal and use the API to import the data into R. Include the code in your R Markdown document. Use the either the kable or datatable function (from Lander, Lesson 2.2) to show the data that you have imported. Write a description of the dataset or datasets. And describe each of the five indicators and discuss what they show about the conditions that are relevant for your issue and policy initiative.

# Installing and loading necessary libraries
library(httr)
library(jsonlite)
## Warning: package 'jsonlite' was built under R version 4.3.3
library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
library(sf)
## Warning: package 'sf' was built under R version 4.3.3
## Linking to GEOS 3.11.0, GDAL 3.5.3, PROJ 9.1.0; sf_use_s2() is TRUE
library(DT)
## Warning: package 'DT' was built under R version 4.3.2
library(tidyr)
library(shiny)
## 
## Attaching package: 'shiny'
## The following objects are masked from 'package:DT':
## 
##     dataTableOutput, renderDataTable
## The following object is masked from 'package:jsonlite':
## 
##     validate
library(ggplot2)
library(leaflet)
## Warning: package 'leaflet' was built under R version 4.3.2

Indicator 2: Demographic Data

The dataset represents the percentage of African American (non-Hispanic) students enrolled in Baltimore City public schools over several academic years. It provides a breakdown of student demographics over time, specifically focusing on African American students. The data spans from the 2009-2010 school year to the 2021-2022 school year and tracks how the composition of the student population has changed.

This dataset, sourced from Baltimore’s open data portal, highlights important trends in the public school system. These demographic shifts are significant for understanding resource distribution and educational equity in the city. By examining how the racial composition of students has evolved, insights can be gained regarding which communities may need additional educational support or intervention.

url <- "https://services1.arcgis.com/mVFRs7NF4iFitgbY/ArcGIS/rest/services/Aastud/FeatureServer/1/query?where=1%3D1&outFields=*&outSR=4326&f=json"
response <- GET(url)
data <- content(response, as = "text", encoding = "UTF-8")
dataset <- fromJSON(data, flatten = TRUE)

education_data <- dataset$features %>%
  dplyr::select(
    attributes.City, 
    attributes.aastud10, 
    attributes.aastud11, 
    attributes.aastud12, 
    attributes.aastud13, 
    attributes.aastud14, 
    attributes.aastud15, 
    attributes.aastud16, 
    attributes.aastud17, 
    attributes.aastud19, 
    attributes.aastud20, 
    attributes.aastud22
  )

colnames(education_data) <- c("City", "2009-2010", "2010-2011", "2011-2012", "2012-2013", 
                              "2013-2014", "2014-2015", "2015-2016", "2016-2017", 
                              "2018-2019", "2019-2020", "2021-2022")

datatable(education_data)

The data shows a gradual decline in the percentage of African American students in Baltimore City public schools from the 2009-2010 school year (87.3%) to the 2021-2022 school year (75.3%). This trend could reflect broader demographic shifts in the city, potentially influenced by factors such as migration, birth rates, economic changes, or changes in school enrollment patterns (e.g., shifts to charter or private schools).

This indicator is crucial for assessing racial diversity in the school system and understanding resource allocation needs that might arise from changes in student demographics. This data can be used to ensure equitable access to educational resources for African American students, who historically face educational disparities compared to other groups. Monitoring these percentages also helps evaluate the effectiveness of diversity and inclusion initiatives within the public school system.

This data’s relevance to the project lies in its ability to provide context to discussions around education equity and the distribution of educational resources across racial demographics.

Indicator 3: School Performance Data

The dataset provides the number of students officially enrolled in grades 9 through 12 across various Community Statistical Areas (CSAs) in Baltimore. Spanning from 2011 to 2021, it captures yearly enrollment figures for each CSA, focusing only on public school students. The source of this data is the Baltimore City Public School System, reporting on student enrollment as of September 30th each year.

This dataset offers insights into student engagement at the high school level and allows for tracking trends in school enrollment across different neighborhoods. By analyzing the fluctuations in enrollment, we can better understand the performance of schools in terms of student retention and engagement, particularly in areas where there may be economic or social challenges.

query_url <- "https://services1.arcgis.com/mVFRs7NF4iFitgbY/arcgis/rest/services/Hsenrol/FeatureServer/0/query?where=1%3D1&outFields=*&outSR=4326&f=json"
response <- GET(query_url)

data <- content(response, "text", encoding = "UTF-8")
dataset <- fromJSON(data, flatten = TRUE)


enrollment_data <- dataset$features %>%
            dplyr::select(
                attributes.CSA2010, 
                attributes.hsenrol11, 
                attributes.hsenrol12, 
                attributes.hsenrol13, 
                attributes.hsenrol14, 
                attributes.hsenrol15, 
                attributes.hsenrol16, 
                attributes.hsenrol17, 
                attributes.hsenrol19, 
                attributes.hsenrol20, 
                attributes.hsenrol21
            )

datatable(enrollment_data)

This dataset is valuable for understanding trends in student engagement across Baltimore’s public high schools. A decline in student enrollment in certain areas may suggest that additional educational resources, such as after-school programs, are necessary to help reengage students. This data can help pinpoint neighborhoods where schools are struggling to retain students, allowing targeted interventions in these areas.

For the civic tech solution, this school enrollment data helps identify regions where after-school programs may be especially impactful. Communities with declining or lower student enrollment figures might benefit from these programs as a means of improving engagement and reducing unsupervised time, which could otherwise lead to negative behaviors. By targeting CSAs where enrollment numbers are declining, this solution can ensure that resources are directed to the areas that need them most, promoting both educational success and reducing the risk of juvenile crime.

Indicator 4: Summer Camps and Youth Programs

This dataset contains information on various summer programs offered by the Enoch Pratt Free Library in Baltimore for 2023. The dataset includes details such as the program name, the organization running the program, age groups served, start dates, and geographic coordinates for each camp’s location. The programs range from arts and crafts to gaming, dance, and movie screenings, targeting different age groups like kids, teens, and families. The data was retrieved from Baltimore’s open data portal using ArcGIS API and includes 420 unique entries representing youth engagement initiatives across multiple locations.

url <- "https://services1.arcgis.com/UWYHeuuJISiGmgXx/arcgis/rest/services/SummerSite_2023/FeatureServer/0/query?where=1%3D1&outFields=*&outSR=4326&f=json"
response <- GET(url)
data <- content(response, as = "text", encoding = "UTF-8")
geojson_data <- fromJSON(data, flatten = TRUE)

locations <- geojson_data$features$geometry
site_names <- geojson_data$features$attributes.Summer_Program_Name
organizations <- geojson_data$features$attributes.Organization
ages_served <- geojson_data$features$attributes.Ages_Served
start_date <- geojson_data$features$attributes.Start_Date

start_date <- as.POSIXct(start_date / 1000, origin = "1970-01-01")

camp_data <- data.frame(
  name = site_names,
  organization = organizations,
  ages_served = ages_served,
  start_date = start_date,
  lng = geojson_data$features$geometry.x,
  lat = geojson_data$features$geometry.y
)

datatable(camp_data)

This data is relevant for promoting after-school and summer programs in areas where they can make the greatest difference. Access to structured activities can provide youth with alternatives to unsupervised time, potentially reducing crime and improving educational outcomes. By overlaying this program data with indicators like juvenile crime rates, poverty levels, and school performance, we can develop targeted efforts to ensure that these programs are accessible to youth in the high-need areas.

For the civic tech solution, mapping the availability of summer camps and after-school programs helps ensure that resources are allocated where they are most needed. By identifying gaps in service provision, this data can guide the development and expansion of youth programs, especially in neighborhoods where at-risk youth may not have access to positive, structured activities. This approach aligns with the policy goal of reducing juvenile crime by providing safer, more engaging environments for young people.

Indicator 5: Households Living Below the Poverty Line

The dataset “Percent of Family Households Living Below the Poverty Line” provides an overview of the poverty levels across various Community Statistical Areas (CSA) in Baltimore, measured as the percentage of family households whose income falls below the federal poverty threshold. The data covers several time ranges, from 2011-2015 up to 2018-2022, offering a longitudinal perspective on poverty dynamics.

The dataset contains attributes for different Community Statistical Areas. CSA2010 and CSA2020 identifies the specific CSAs in Baltimore, better known as neighborhoods. Specifically, hhpov15 to hhpov22, which represents the percentage of family households living below the poverty line for the years 2011-2022.

The source data is derived from the American Community Survey and is used by federal and local governments to allocate resources and identify eligible communities for various assistance programs. The dataset also includes spatial references, allowing the visualization of poverty trends across Baltimore neighborhoods over time.

query_url <- "https://services1.arcgis.com/mVFRs7NF4iFitgbY/arcgis/rest/services/Hhpov/FeatureServer/0/query?where=1%3D1&outFields=*&f=pjson"
response <- GET(query_url)

data <- content(response, "text", encoding = "UTF-8")
dataset <- fromJSON(data, flatten = TRUE)


poverty_data <- dataset$features %>%
                dplyr::select(
                  attributes.CSA2010, 
                  attributes.hhpov21, 
                  attributes.hhpov22, 
                  attributes.hhpov15, 
                  attributes.hhpov16, 
                  attributes.hhpov17, 
                  attributes.hhpov18, 
                  attributes.hhpov19
                )

datatable(poverty_data)

This dataset is critical for assessing socio-economic disparities across different regions of Baltimore, particularly in identifying neighborhoods where family households are most affected by poverty. For instance, areas like Cherry Hill and Poppleton/The Terraces/Hollins Market show consistently high percentages of households below the poverty line, with rates above 40% in some years. In contrast, areas such as Greater Roland Park/Poplar Hill and South Baltimore report significantly lower poverty rates. These extremes provide an important context for understanding how poverty is distributed across Baltimore and highlight where the city might focus its resources.

This dataset provides invaluable insight into the areas that may benefit most from targeted interventions. High-poverty neighborhoods such as Cherry Hill and Greater Rosemont can be prioritized for after-school programs and educational support, given that poverty often correlates with limited access to resources. This alignment between economic disadvantage and social outcomes can help justify reallocating funds from reactive measures, like policing, to proactive educational initiatives. After-school programs can provide structured environments that reduce unsupervised time, potentially lowering juvenile crime rates while improving educational outcomes.

This data also facilitates long-term evaluation of policy impacts, as we can track changes in poverty rates over time in relation to program implementation. For example, if after-school programs are introduced in a high-poverty area, tracking poverty and crime rates from that point could offer evidence of the program’s effectiveness. This kind of evidence-based policymaking is crucial for ensuring that resources are allocated where they can have the greatest positive effect, directly supporting the broader goal of reducing juvenile crime and improving education equity in Baltimore.

Conclusion

The analysis of these five indicators from Baltimore’s open data portal highlights critical factors affecting youth crime, educational equity, and community resource distribution. By leveraging crime data, demographic trends, school performance statistics, and socio-economic indicators, a clear picture emerges of where after-school programs and other interventions can be most effective. The data-driven approach enables targeted solutions, ensuring that resources are directed to high-need areas, particularly those with high juvenile crime rates, high poverty levels, and declining school engagement. This strategic alignment not only promotes safer communities but also supports the development of a more equitable and supportive environment for Baltimore’s youth, reinforcing the broader policy initiative of reducing juvenile crime through proactive community engagement.

Assignment 04: Part II-APIs and Open Data and Smart Data

Smart City Introduction

Smart cities leverage technology and data to improve urban living by optimizing efficiency, sustainability, and citizen engagement. By collecting and analyzing data through various platforms, cities are able to enhance infrastructure, public services, and policy-making, making them more responsive to the needs of the community.

For this assignment, I focus on two key smart data indicators for Baltimore: Public Transit Usage and 311 Service Requests. These indicators provide insight into how residents move through the city and engage with city services, which can be used to inform public safety, and even urban planning initiatives. Understanding these patterns can help create a more equitable and safe environment, especially in neighborhoods with high crime rates or insufficient resources for youth.

The goal is to explore how these data sets can inform new policies that can enhance community safety and engagement, especially around after-school programming for youth in at-risk neighborhoods. By studying trends in public transportation and community participation, more targeted and efficient policy can be developed.

Indicator 1: Public Transit Usage

The Public Transit Usage indicator measures the percentage of commuters aged 16 and above who rely on public transportation in Baltimore, based on data from the American Community Survey. The dataset spans from 2007 to 2022 and tracks the use of public transportation in various neighborhoods across the city, represented as Community Statistical Areas. Understanding the usage patterns of public transit is important for identifying areas where infrastructure may need improvement, and for designing sustainable urban transit solutions. This indicator fits into the technical efficiency framework discussed in “IoT in Smart Cities: A Survey of Technologies, Practices, and Challenges.” By examining transit data, the efficiency of public transport systems can be analyzed, and patterns in commuter behavior can be identified (Zanella et al., 2014). Additionally, in line with the citizen-centric approach from “The many faces of the smart city: Differing value propositions in the activity portfolios of nine cities,” this indicator reflects how well Baltimore’s public transit system actually meets the needs of its residents, contributing to urban accessibility and equity (Csukás & Szabó, 2021).

The Public Transit Usage indicator aligns with the smart city concepts of both efficiency and sustainability. Smart cities leverage technology and data to optimize infrastructure, and public transit data allows city planners to observe where transit services are most needed, and where improvements would be benefited the most from. This indicator supports urban mobility efforts, reducing dependence on personal vehicles, which contributes to broader smart city goals like lowering emissions and fostering sustainable development (Zanella et al., 2014). By identifying neighborhoods with lower transit usage, Baltimore can focus its efforts on ensuring that public transportation is accessible to all, promoting environmental sustainability and social equity.

This data can also play an important role in furthering the policy initiative to increase access to after-school programs and reducing juvenile crime. By identifying underserved neighborhoods with low public transit usage, the policy can target these areas for improved routes, or alternative transportation options. Interventions like this would ensure that youth have reliable, safe transportation to after-school programs, addressing possible gaps in accessibility, and reducing barriers to participation. This aligns with the smart city principles of citizen engagement and social inclusion by focusing on public needs and ensuring that transit improvements support educational and social outcomes (Csukás & Szabó, 2021).

## Simple feature collection with 6 features and 2 fields
## Geometry type: GEOMETRY
## Dimension:     XY
## Bounding box:  xmin: -76.7112 ymin: 39.19724 xmax: -76.52969 ymax: 39.35562
## Geodetic CRS:  WGS 84
## # A tibble: 6 × 3
##   CSA2010                           avg_percent_use                     geometry
##   <chr>                                       <dbl>               <GEOMETRY [°]>
## 1 Allendale/Irvington/S. Hilton               21.6  POLYGON ((-76.65749 39.2761…
## 2 Beechfield/Ten Hills/West Hills             11.6  POLYGON ((-76.69481 39.302,…
## 3 Belair-Edison                               20.5  POLYGON ((-76.56776 39.3264…
## 4 Brooklyn/Curtis Bay/Hawkins Point           19.2  MULTIPOLYGON (((-76.58861 3…
## 5 Canton                                       4.09 POLYGON ((-76.57166 39.2844…
## 6 Cedonia/Frankford                           15.8  POLYGON ((-76.52972 39.3353…

The results from the data table show that neighborhoods such as Allendale/Irvington/S. Hilton have an average public transit usage rate of 21.6%, while Canton has a lower rate of 4.09%. These insights can be directly applied to the policy initiative aimed at improving transportation access for underserved communities. Targeting neighborhoods with lower transit usage for infrastructure improvements can help ensure equitable access to education and employment opportunities, which would address disparities in service and contribute to the overall smart city goals of accessibility, sustainability, and inclusivity.

Below is an interactive Shiny app that allows us to view the public transit data as a time-series for each individual CSA.

ui <- fluidPage(
  titlePanel("Public Transit Usage Over Time by CSA"),
  
  sidebarLayout(
    sidebarPanel(
      selectInput("selected_csa", 
                  "Select CSA:", 
                  choices = unique(transport_time_series$CSA2010),
                  selected = unique(transport_time_series$CSA2010)[1])
    ),
    
    mainPanel(
      plotOutput("time_series_plot")
    )
  )
)

server <- function(input, output) {
  
  filtered_data <- reactive({
    transport_time_series %>% filter(CSA2010 == input$selected_csa)
  })
  
  output$time_series_plot <- renderPlot({
    ggplot(filtered_data(), aes(x = year, y = percent_use)) +
      geom_point() +
      labs(title = paste("Public Transit Usage Over Time for", input$selected_csa),
           x = "Year",
           y = "Percentage of Public Transit Users") +
      theme_minimal()
  })
}

shinyApp(ui = ui, server = server)
## 
## Listening on http://127.0.0.1:8723

Indicator 2: Baltimore 311 Participation

The Baltimore 311 Participation indicator provides critical insights into citizen engagement within the city. By analyzing data from the 311 service request source, we can understand which neighborhoods are actively using city services to report issues or request assistance. This indicator highlights where residents are most engaged with the city’s public services and where additional efforts might be needed to encourage higher levels of participation. By exploring the data, patterns of non-emergency requests can be identified, offering a comprehensive view of community involvement across Baltimore’s neighborhoods.

This indicator aligns with the citizen-centric approach outlined in “The many faces of the smart city: Differing value propositions in the activity portfolios of nine cities” (Csukás & Szabó, 2021). The 311 system allows residents to play an active role in identifying and resolving local issues, making it a valuable tool for fostering civic participation. The ability to map and analyze service requests helps city officials understand which neighborhoods are more engaged and where efforts may need to be focused to boost participation. This makes it an essential component of a smart city, emphasizing the importance of citizen involvement in shaping the urban environment.

From the results, we can identify the neighborhoods in Baltimore that have the highest number of 311 service requests for the year, indicating active community engagement in utilizing non-emergency services. However, there is a notable entry with missing neighborhood data (NA), which accounts for 145 service requests—the highest in the dataset. This may suggest that some requests are being submitted without specific neighborhood information, which could impact the ability to allocate resources effectively in certain areas. Beyond this, neighborhoods like Lauraville, Ednor Gardens-Lakeside, and Cedmont appear to have the highest engagement with the 311 system, each logging between 70 and 85 requests, which reflects a strong connection between residents and city services.

In the context of my policy initiative, these results show where non-emergency needs are most frequently reported. This indicator highlights areas of active citizen participation, which may be linked to stronger community engagement or increased city service utilization. For instance, targeting neighborhoods with high 311 usage, such as Lauraville and Ednor Gardens-Lakeside, can help allocate resources more efficiently and improve the effectiveness of community outreach efforts. And to go further, identifying gaps, such as the NA category, suggests there might be neighborhoods where either fewer residents are engaging with the 311 service or data collection needs improvement.

Below, I was able to plot every 311 record from the data source (excluded ones with missing values) on a granular map of Baltimore, MD using leaflet!

ui <- fluidPage(
  h2("Baltimore 311 Service Requests Mapped"),
  leafletOutput("baltimore_map")
)

server <- function(input, output, session) {
  # Convert Latitude and Longitude to numeric, and filter out invalid rows
  data_sf <- data_sf %>%
    mutate(
      Latitude = as.numeric(Latitude),
      Longitude = as.numeric(Longitude)
    ) %>%
    filter(!is.na(Latitude) & !is.na(Longitude))
  
  output$baltimore_map <- renderLeaflet({
    leaflet(data = data_sf) %>%
      addTiles() %>%
      addCircleMarkers(~Longitude, ~Latitude, radius = 3, color = "blue",
                       label = ~paste("Neighborhood:", Neighborhood, "<br>",
                                      "Service Type:", SRType),
                       popup = ~paste("Address:", Address, "<br>",
                                      "Type:", SRType, "<br>",
                                      "Status:", SRStatus)) %>%
      setView(lng = -76.6122, lat = 39.2904, zoom = 12)
  })
}

shinyApp(ui = ui, server = server)
## 
## Listening on http://127.0.0.1:7842