U.S. Border Patrol Encounters

Author

Tony Galvan

Introduction

Recently, I saw this graph on X (formerly Twitter) and discovered that the data is available on the U.S. Customs and Border Patrol website. I thought that this data is timely considering the election in a few days, so I decided to analyze the data.

About the Data

Encounter data includes U.S. Border Patrol Title 8 apprehensions, Office of Field Operations Title 8 inadmissibles, and all Title 42 expulsions for fiscal years 2020 to date. Data is available for the Northern Land Border, Southwest Land Border, and Nationwide (i.e., air, land, and sea modes of transportation) encounters.

Nationwide Encounters by Area of Responsibility

Here is a glimpse of the data:

Code 
glimpse(cbp_resp_raw)
Rows: 112,603
Columns: 12
$ `Fiscal Year`            <dbl> 2020, 2020, 2020, 2020, 2020, 2020, 2020, 202…
$ `Month Grouping`         <chr> "FYTD", "FYTD", "FYTD", "FYTD", "FYTD", "FYTD…
$ `Month (abbv)`           <chr> "APR", "APR", "APR", "APR", "APR", "APR", "AP…
$ Component                <chr> "Office of Field Operations", "Office of Fiel…
$ `Land Border Region`     <chr> "Northern Land Border", "Northern Land Border…
$ `Area of Responsibility` <chr> "Boston Field Office", "Boston Field Office",…
$ `AOR (Abbv)`             <chr> "Boston", "Boston", "Boston", "Boston", "Bost…
$ Demographic              <chr> "FMUA", "FMUA", "Single Adults", "Single Adul…
$ Citizenship              <chr> "BRAZIL", "CANADA", "CANADA", "CANADA", "CHIN…
$ `Title of Authority`     <chr> "Title 8", "Title 8", "Title 42", "Title 8", …
$ `Encounter Type`         <chr> "Inadmissibles", "Inadmissibles", "Expulsions…
$ `Encounter Count`        <dbl> 3, 1, 2, 239, 1, 0, 1, 6, 1, 1, 1, 18, 52, 1,…

Reproduce the plot

Let’s try to reproduce the “Inadmissible Aliens” plot by @fentasyl. It is not called out in the Tweet, but “CHNV” refers to citizens of Cuba, Haiti, Nicaragua, and Venezuela. Also, “Interior” includes Northern Land Border encounters.

Code 
cbp_resp |>
  filter(
    encounter_type == "Inadmissibles",
    chnv
  ) |>
  mutate(
    chnv_region = if_else(
      land_border_region == "Southwest Land Border",
      "CHNV\nSouthern\nBorder",
      "CHNV\nInterior\n"
    )
  ) |>
  count(
    bom_month_date, chnv_region,
    wt = encounter_count,
    name = "total_encounters"
  ) |>
  ggplot(
    aes(
      x = bom_month_date,
      y = total_encounters,
      fill = chnv_region
    )
  ) +
  geom_col(
    color = "black",
    lwd = 0.25,
    # make the bars wider
    width = 28
  ) +
  scale_fill_manual(
    values = c("#C10000", "#0F4962")
  ) +
  scale_x_date(
    date_breaks = "1 year", 
    date_labels = "%Y"
  ) +
  scale_y_continuous(
    minor_breaks = seq(0, 65000, 2500),
    breaks = seq(0, 65000, 10000)
  ) +
  labs(
    x = NULL,
    y = NULL,
    fill = NULL,
    title = '"Inadmissible" Aliens',
    subtitle = "Per Month, Jan 2020 - Sep 2024",
    caption = "Source: U.S. Customs and Border Patrol  |  Inspired by: @fentasyl"
  ) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.ticks = element_blank(),
    panel.border = element_blank(),
    plot.title = element_text(
      size = 20,
      face = "bold"
    ),
    # add more horizontal grid lines
    panel.grid.major.y = element_line(
      color = "gray60", 
      linewidth = 0.25
    ),
    panel.grid.minor.y = element_line(
      color = "gray90", 
      linewidth = 0.25
    ),
    # move the legend inside the panel/plot
    legend.position = c(0.3, 0.5),
    # add a border to the legend
    legend.background = element_rect(
      color = "black",
      fill = "white"
    ),
    # make the legend key a small box instead of a tall rectangle
    legend.key.width = unit(0.25, "cm"),
    legend.key.height = unit(0.25, "cm"),
    # move up the x axis text
    axis.text.x = element_text(
      margin = margin(t = -10)
    )
  )

Reproduction of @fentasyl‘s ’Inadmissible Aliens’ plot

The data in the plot above does not seem to match the original. Especially the end of 2023. Also, I couldn’t quite figure out how to make the legend key a small box instead of a tall rectangle.

Are “Admissible” border crossings down?

The media is claiming that border crossings are down, but they do not count “inadmissible” border crossings as “border crossings”. Let’s see if the “admissible” border crossings are down.

Code 
cbp_resp |>
  filter(
    encounter_type != "Inadmissibles"
  ) |>
  count(
    bom_month_date,
    wt = encounter_count,
    name = "total_encounters"
  ) |>
  ggplot(
    aes(
      x = bom_month_date,
      y = total_encounters
    )
  ) +
  geom_point(
    color = "#C10000",
    alpha = 0.5
  ) +
  geom_smooth(
    method = "loess",
    color = "#0F4962",
    se = FALSE
  ) +
  scale_x_date(
    date_breaks = "1 year", 
    date_labels = "%Y"
  ) +
  scale_y_continuous(
    labels = scales::comma_format()
  ) +
  labs(
    x = NULL,
    y = NULL,
    fill = NULL,
    title = '"Admissible" Aliens',
    subtitle = "Per Month, Jan 20200 - Sep 2024",
    caption = "Source: U.S. Customs and Border Patrol"
  ) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.ticks = element_blank(),
    panel.border = element_blank(),
    plot.title = element_text(
      size = 20,
      face = "bold"
    ),
    # add more horizontal grid lines
    panel.grid.major.y = element_line(
      color = "gray60", 
      linewidth = 0.25
    ),
    panel.grid.minor.y = element_line(
      color = "gray90", 
      linewidth = 0.25
    ),
    axis.text.x = element_text(
      margin = margin(t = -10)
    )
  )

Admissible Aliens

It seems that “admissible” border crossings are down since the beginning of 2024. Let’s take a look at a combined plot that includes citizens of all countries.

Combined Plot

Code 
cbp_resp |>
  count(
    bom_month_date,
    wt = encounter_count,
    name = "total_encounters"
  ) |>
  ggplot(
    aes(
      x = bom_month_date,
      y = total_encounters
    )
  ) +
  geom_point(
    color = "#C10000",
    alpha = 0.5
  ) +
  geom_smooth(
    method = "loess",
    color = "#0F4962",
    se = FALSE
  ) +
  scale_x_date(
    date_breaks = "1 year", 
    date_labels = "%Y"
  ) +
  scale_y_continuous(
    labels = scales::comma_format()
  ) +
  labs(
    x = NULL,
    y = NULL,
    title = "All Encounters",
    subtitle = "Per Month, Jan 20200 - Sep 2024",
    caption = "Source: U.S. Customs and Border Patrol"
  ) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.ticks = element_blank(),
    panel.border = element_blank(),
    plot.title = element_text(
      size = 20,
      face = "bold"
    ),
    # add more horizontal grid lines
    panel.grid.major.y = element_line(
      color = "gray60", 
      linewidth = 0.25
    ),
    panel.grid.minor.y = element_line(
      color = "gray90", 
      linewidth = 0.25
    ),
    axis.text.x = element_text(
      margin = margin(t = -10)
    )
  )

All Aliens

It looks like border crossings are down since 2024 whether you count “admissible” or “inadmissible” border crossings. Let’s see if the story is different for citizens of Cuba, Haiti, Nicaragua, and Venezuela.

CHNV Plot

Code 
cbp_resp |>
  filter(chnv) |>
  count(
    bom_month_date,
    wt = encounter_count,
    name = "total_encounters"
  ) |>
  ggplot(
    aes(
      x = bom_month_date,
      y = total_encounters
    )
  ) +
  geom_point(
    color = "#C10000",
    alpha = 0.5
  ) +
  geom_smooth(
    method = "loess",
    color = "#0F4962",
    se = FALSE
  ) +
  scale_x_date(
    date_breaks = "1 year", 
    date_labels = "%Y"
  ) +
  scale_y_continuous(
    labels = scales::comma_format()
  ) +
  labs(
    x = NULL,
    y = NULL,
    title = "All CHNV Encounters",
    subtitle = "Per Month, Jan 20200 - Sep 2024",
    caption = "Source: U.S. Customs and Border Patrol"
  ) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.ticks = element_blank(),
    panel.border = element_blank(),
    plot.title = element_text(
      size = 20,
      face = "bold"
    ),
    # add more horizontal grid lines
    panel.grid.major.y = element_line(
      color = "gray60", 
      linewidth = 0.25
    ),
    panel.grid.minor.y = element_line(
      color = "gray90", 
      linewidth = 0.25
    ),
    axis.text.x = element_text(
      margin = margin(t = -10)
    )
  )

CHNV

Conclusion

It seems that border crossings are down since 2024 whether you count “admissible” or “inadmissible” border crossings. The story is the same for citizens of Cuba, Haiti, Nicaragua, and Venezuela.

The data does not seem to match the original plot by @fentasyl. I am not sure if the data is different or if I made a mistake in my analysis.