"}],"short_description":"Explore how to map U.S. counties' predominant disadvantage categories using Justice40 data and ArcGIS API for Python.","flexible_content":[{"acf_fc_layout":"content","content":"The Justice40 Initiative is a federal effort to ensure that 40 percent of the benefits from federal investments reach disadvantaged communities across the United States. It targets inequities in eight key areas: Climate Change, Clean Energy, Clean Transit, Affordable Housing, Workforce Development, Legacy Pollution, Health Burdens, and Clean Water Infrastructure.<\/span>\u00a0<\/span><\/p>\n Esri has integrated the Justice40 Tracts from Council on Environmental Quality (CEQ) into ArcGIS Living Atlas of the World, enabling detailed mapping and analysis of these disadvantaged communities across the U.S. and its territories. By leveraging this data in ArcGIS, users can create maps that identify these communities and support deeper analysis and targeted interventions.<\/span>\u00a0<\/span><\/p>\n Because many policy decisions and resource allocations occur at the county level, aggregating data from census tracts to counties ensures that analysis is directly applicable to the administrative units responsible for implementing policies. In this article, we will use the ArcGIS API for Python and Justice40 data from Esri to map the predominant disadvantage categories at the county level across the U.S. <\/span>\u00a0<\/span><\/p>\n Importance of Identifying Predominant Disadvantage Category<\/strong><\/p>\n Identifying the predominant disadvantage category for each county involves determining which issue\u2014such as climate change, clean energy, housing, or health\u2014most affects the census tracts within that county. This is crucial because it highlights the primary challenge impacting the largest portion of the population, enabling more focused and effective policy interventions.<\/span>\u00a0<\/span><\/p>\n Additionally, aggregating racial and ethnic demographic data to the county level is essential for recognizing and addressing disparities. By analyzing the share of each racial and ethnic group within a county, policymakers can ensure that their efforts are inclusive and equitable, targeting the needs of all community members, particularly those who have been historically marginalized.<\/span>\u00a0<\/span><\/p>\n Overall, understanding the predominant disadvantage category in a county helps policymakers:<\/span>\u00a0<\/span><\/p>\n To create a map that visualizes the predominant disadvantage categories in U.S. counties using Justice40 data, we will follow these five steps:<\/span>\u00a0<\/span><\/p>\n"},{"acf_fc_layout":"sidebar","content":" Step 1: Import Required Libraries and Access Justice40 Data\u00a0<\/span><\/strong><\/p>\n First, we import the necessary libraries and connect to ArcGIS Online to access the Justice40 data.<\/span>\u00a0<\/span><\/p>\n"},{"acf_fc_layout":"content","content":" \u00a0Step 2: Query and Prepare Tract-Level Data\u00a0<\/strong><\/p>\n Next, we query the Justice40 data layer to retrieve data fields including population, disadvantage categories, and racial demographics. We then prepare this data for aggregation to the county level by adding essential columns, such as the county FIPS code and calculated fields like the number of people affected in each demographic group.\u00a0<\/span><\/p>\n"},{"acf_fc_layout":"content","content":" \u00a0Step 3: Aggregate Data to County Level\u00a0<\/strong><\/p>\n In this step, we aggregate the tract-level data to the county level to identify the predominant disadvantage category for each county. This is crucial for understanding which issues are most prevalent in each county in the U.S., allowing for targeted interventions.<\/span>\u00a0<\/span>We also calculate the percentage of different racial and ethnic groups within each county. This helps ensure that the analysis considers the diverse populations affected by the disadvantages.<\/p>\n <\/p>\n"},{"acf_fc_layout":"content","content":" Step 4: Identify the Predominant Disadvantage Category\u00a0<\/strong><\/p>\n Here, we identify the predominant disadvantage category for each county by determining which category is most prevalent.<\/span> The function,\n
\n
\n
\n
import pandas as pd\r\nfrom arcgis.gis import GIS\r\n\r\n# Connect to ArcGIS Online account\r\ngis = GIS(\"https:\/\/www.arcgis.com\", \"username\", \"password\")\r\n\r\n# Retrieve the Justice40 data layer from ArcGIS Online\r\nitem_id = 'f95344889cab44bd84207052f44cb940'\r\njustice = gis.content.get(item_id)\r\njustice_layer = justice.layers[0]\r\n\r\n# Display available layers\r\nprint(\"\\n\".join([f\"{lyr.properties.id}: {lyr.properties.name}\" for lyr in justice.layers]))\r\n<\/code><\/pre>\n"},{"acf_fc_layout":"content","content":"# Query the layer and convert to Spatially Enabled DataFrame \r\nlyr_df = justice_layer.query(out_fields=['GEOID10', 'SF', 'CF', 'TPF', 'N_WTR', 'N_WKFC', 'N_CLT', 'N_ENY', 'N_TRN', 'N_HSG', 'N_PLN', 'N_HLTH', 'CC', 'SN_C', 'DM_B', 'DM_AI', 'DM_A', 'DM_HI', 'DM_W', 'DM_H', 'DM_T'], out_sr=4326).sdf \r\n\r\n# First, ensure GEOID10 is a string\r\nlyr_df1['GEOID10'] = lyr_df1['GEOID10'].astype(str)\r\n\r\n# Extract the first 5 characters to get the state+county FIPS code\r\nlyr_df1['FIPS'] = lyr_df1['GEOID10'].str.slice(start=0, stop=5)\r\n\r\n# Create new columns for racial and demographic calculations \r\nlyr_df['African American'] = lyr_df['TPF'] * lyr_df['DM_B'] \r\nlyr_df['White'] = lyr_df['TPF'] * lyr_df['DM_W'] \r\nlyr_df['Asian'] = lyr_df['TPF'] * lyr_df['DM_A'] \r\nlyr_df['Hispanic'] = lyr_df['TPF'] * lyr_df['DM_H'] \r\nlyr_df['American Indian'] = lyr_df['TPF'] * lyr_df['DM_AI'] \r\nlyr_df['Native Hawaiian or Pacific'] = lyr_df['TPF'] * lyr_df['DM_HI'] \r\nlyr_df['Two or more races'] = lyr_df['TPF'] * lyr_df['DM_T'] \r\nlyr_df['Pop_in_Disadv_Tracts'] = lyr_df['TPF'] * lyr_df['SN_C'] \r\nlyr_df['FIPS'] = lyr_df['GEOID10'].str.slice(start=0, stop=5) \r\nlyr_df['Total_Tracts_Per_County'] = lyr_df.groupby('FIPS')['FIPS'].transform('size')\r\n<\/code><\/pre>\n"},{"acf_fc_layout":"content","content":"# Aggregate data at the county level \r\ncounty_level_df = lyr_df.groupby(['FIPS', 'SF', 'CF']).agg({ \r\n 'TPF': 'sum', \r\n 'N_WTR': 'sum', \r\n 'N_WKFC': 'sum', \r\n 'N_CLT': 'sum', \r\n 'N_ENY': 'sum', \r\n 'N_TRN': 'sum', \r\n 'N_HSG': 'sum', \r\n 'N_PLN': 'sum', \r\n 'N_HLTH': 'sum', \r\n 'SN_C': 'sum', \r\n 'African American': 'sum', \r\n 'White': 'sum', \r\n 'Asian': 'sum', \r\n 'Hispanic': 'sum', \r\n 'American Indian': 'sum', \r\n 'Native Hawaiian or Pacific': 'sum', \r\n 'Two or more races': 'sum', \r\n 'Pop_in_Disadv_Tracts': 'sum', \r\n 'Total_Tracts_Per_County': 'first' \r\n}).reset_index()\r\n<\/code><\/pre>\n"},{"acf_fc_layout":"content","content":"# Calculate share of population in disadvantaged tracts and demographic percentages \r\ncounty_level_df['Share_of_Pop_in_Disadv_Tracts'] = county_level_df['Pop_in_Disadv_Tracts'] \/ county_level_df['TPF'] \r\ncounty_level_df['Share_of_African Americans'] = county_level_df['African American'] \/ county_level_df['TPF']\r\ncounty_level_df['Share_of_Whites'] = county_level_df['White'] \/ county_level_df['TPF'] \r\ncounty_level_df['Share_of_Hispanics'] = county_level_df['Hispanic'] \/ county_level_df['TPF'] \r\ncounty_level_df['Share_of_Asians'] = county_level_df['Asian'] \/ county_level_df['TPF'] \r\ncounty_level_df['Share_of_American_Indians'] = county_level_df['American Indian'] \/ county_level_df['TPF']\r\ncounty_level_df['Share_of_Native_Hawaiians'] = county_level_df['Native Hawaiian or Pacific'] \/county_level_df['TPF'] \r\ncounty_level_df['Share_of_Two_or_more races'] = county_level_df['Two or more races'] \/ county_level_df['TPF']\r\n<\/code><\/pre>\n"},{"acf_fc_layout":"content","content":"find_predominant_category_explicit<\/code> , determines the predominant disadvantage category for each county. It checks whether the county has any disadvantaged tracts using the Has_Disadv_Tract<\/code> flag. If there are no disadvantaged tracts, it returns “No Disadvantaged Tracts.” Otherwise, it iterates through each disadvantage category to find the one with the highest impact (i.e., the highest value). The function then returns the full name of the predominant category.<\/span><\/p>\n"},{"acf_fc_layout":"content","content":"# Create a flag for counties with disadvantaged tracts \r\ncounty_level_df['Has_Disadv_Tract'] = county_level_df['SN_C'] > 0 \r\n\r\n# Define disadvantage categories and their names \r\ndisadvantage_categories = ['N_WTR', 'N_WKFC', 'N_CLT', 'N_ENY', 'N_TRN', 'N_HSG', 'N_PLN', 'N_HLTH'] \r\n\r\ncategory_names = { \r\n 'N_WTR': 'Water and Wastewater Disadvantaged', \r\n 'N_WKFC': 'Workforce Development Disadvantaged', \r\n 'N_CLT': 'Climate Change Disadvantaged', \r\n 'N_ENY': 'Energy Disadvantaged', \r\n 'N_TRN': 'Transportation Disadvantaged', \r\n 'N_HSG': 'Housing Disadvantaged', \r\n 'N_PLN': 'Legacy Pollution Disadvantaged', \r\n 'N_HLTH': 'Health Disadvantaged'} \r\n\r\n# Function to find predominant disadvantage category \r\ndef find_predominant_category_explicit(row): \r\n if not row['Has_Disadv_Tract']: \r\n return \"No Disadvantaged Tracts\" \r\n max_value = -1 \r\n max_category = None \r\n for category in disadvantage_categories: \r\n category_value = row[category] if pd.notnull(row[category]) else 0 \r\n if category_value > max_value: \r\n max_value = category_value \r\n max_category = category \r\n return category_names[max_category] if max_category else \"No Disadvantaged Tracts\" \r\n\r\n# Apply the function to determine predominant category \r\ncounty_level_df['Predominant_Disadvantage'] = county_level_df.apply(find_predominant_category_explicit, axis=1)\r\n<\/code><\/pre>\n"},{"acf_fc_layout":"content","content":"