Durham P.D. Reputational Changes Following BLM

library(tidyverse)
library(patchwork)

#durham_2018 <- read_csv("projects/durham/data/durham2018.csv") |>
durham_2018 <- read_csv("data/durham2018.csv") |>
  janitor::clean_names() |>
  select(matches("q7_01_1"), "asian", "white", "native_american", "black",
         "latino", "other")

#durham_2022 <- readxl::read_xlsx("projects/durham/data/durham2022.xlsx") |>
durham_2022 <- readxl::read_xlsx("data/durham2022.xlsx") |>
  janitor::clean_names() |>
  select(
    matches("q7_01_1"), 
    matches("q31"), 
    matches("q32_01"), 
    matches("q32_02"), 
    matches("q32_03"), 
    matches("q32_04"), 
    matches("q32_05"), 
    matches("q32_06")
  )

colnames(durham_2018) <- c("police_relationship", "asian", "white",
                           "native_american", "black", "hispanic", "other")
colnames(durham_2022) <- c("police_relationship", "hispanic", "asian", "black",
                           "native_american", "white", "native_hawaiian", 
                           "other")

# 1 = Yes, 2 = no --------------------------------------------------------------
durham_2018 <- durham_2018 |>
  mutate(
    year = 2018,
    across(where(is.numeric), ~ na_if(., 9))
  )

durham_2022 <- durham_2022 |>
  mutate(
    other = if_else(native_hawaiian == 1, 1, other),
    year = 2022,
    across(where(is.numeric), ~ na_if(., 9))
  ) |>
  select(!native_hawaiian)

The nationwide Black Lives Matter (BLM) protests were widely attended in Durham, a disproportionately black city in North Carolina. Although initial responses by Durham PD lacked substantive policy or funding changes, the following years began to show moderate progress towards meeting the demands of the BLM advocates.

Every year, the city of Durham sends a robust satisfaction survey to all its residents (of which I used to be one). The questions range from opinions on transportation, public parks, policing, and more. Demographic questions in the survey allow us to see how Durham PD’s reputation changed within different racial groups before and after widespread BLM protests in Durham.

The primary policy shift brought about by protesters’ demands was establishing the HEART program to support regular policing efforts in Durham. In short, the HEART program seeks to reduce the potential of police violence by adding social workers and mental health experts to the 9-1-1 call centers, giving unarmed response options for calls for which armed police are unlikely to be necessary. Since HEART began, they have handled over 32,000 calls, with only 0.02% of calls ultimately requiring traditional police backup. Although HEART has been widely successful, Durham PD’s reputational shifts remain mixed. Figure 1 explores those changes.

durham_2018_long <- durham_2018 |>
  pivot_longer(
    cols = asian:other, 
    names_to = "race", 
    values_to = "race_value"
  ) |>
  filter(!is.na(race_value))

durham_2022_long <- durham_2022 |>
  pivot_longer(
    cols = hispanic:other, 
    names_to = "race", 
    values_to = "race_value"
  ) |>
  filter(!is.na(race_value))

combined <- bind_rows(durham_2018_long, durham_2022_long) |>
  filter(!is.na(police_relationship))  # remove missing responses

summary <- combined |>
  mutate(
    race = if_else(race %in% c("black", "white", "hispanic"), race, "Other")
  ) |>
  group_by(year, race, police_relationship) |>
  summarise(n = n(), .groups = "drop") |>
  group_by(year, race) |>
  mutate(prop = n / sum(n))

summary$race <- str_to_title(summary$race)

summary <- summary |>
  mutate(
    race = str_to_title(race),
    race = factor(race, levels = c("Black", "White", "Hispanic", "Other")),
    year = factor(year, levels = c(2022, 2018)),
    police_relationship = factor(police_relationship, levels = 5:1)
  )

summary |>
  filter(!is.na(race)) |>
  ggplot(aes(x = year, y = prop, fill = police_relationship)) +
  geom_col(position = "stack", width = 0.6) +
  coord_flip() +
  facet_wrap(~race, ncol = 1, switch = "both") +
  scale_fill_brewer(
    palette = "RdBu",
    direction = -1, 
    labels = rev(c(
      "Very Dissatisfied",
      "Dissatisfied",
      "Neutral",
      "Satisfied",
      "Very Satisfied"
    ))
  ) +
  labs(
    title = paste0(
      "How satisfied are you with the overall police relationship ",
      "with your community?"
    ),
    subtitle = "Durham Resident Satisfaction Surveys (2018, 2022)",
    x = NULL,
    y = element_blank(),
    fill = element_blank()
  ) +
  ggthemes::theme_clean() +
  theme(
    strip.text = element_text(size = 11, face = "bold"),
    strip.placement = "outside",
    legend.position = "bottom",
    legend.background = element_blank(),
    axis.ticks.x = element_blank(),
    axis.text.x = element_blank(),
    axis.line.x = element_blank(),
    plot.background = element_rect(fill = "white"),
    panel.background = element_blank(),
    plot.title = element_text(size = 18),
    plot.subtitle = element_text(size = 14)
  ) +
  guides(fill = guide_legend(reverse = T))

ggsave("plot.png", width = 12, height = 6)

The success of Durham’s attempts to repair its relationship with black communities depends on the goals of Durham’s policy shifts. If the objective was to diminish negative feelings towards police, Durham has seen modest success. Black communities are the only ones in which fewer members report negative feelings towards police in 2022 than in 2018. However, positive sentiment has also declined in the same communities, resulting in a growing indifference towards Durham PD among black communities.

Both increasingly negative and decreasingly positive sentiments were most pronounced in white communities. There are many potential explanations for those shifts, including economic shifts, changes to national mood towards policing, shifts in who responds to the survey, and more. However, for the sake of brevity, we will leave those confounding factors unexplored.

Like many other major American police forces, Durham PD has a long way to go. As it stands, fewer than half of all racial groups in Durham are satisfied with their community’s relationship with its police force. Even with the success of the HEART program, moods towards Durham PD continue to cool on average. A national re-exploration of policing could be to blame, but it is the responsibility of municipal leadership to combat those national currents. Although HEART has handled over 32,000 calls instead of Durham PD, the city has not reduced police funding, a key demand of the BLM movement. As violent crime rates shrink year-over-year in Durham, police funding continues to grow. Durham is fortunate to have a strong statistical justification for a reduction in police budget. Still, city leadership continues to drag its feet, and without continued public pressure, broader policing change may be unlikely in the near future.