Rename long column names for readability (e.g., year, slaves_embarked, slaves_disembarked).

trans <- trans %>%
  rename(
    year = voyage_dates__imp_arrival_at_port_of_dis_sparsedate__year,
    slaves_embarked = voyage_slaves_numbers__imp_total_num_slaves_embarked,
    slaves_disembarked = voyage_slaves_numbers__imp_total_num_slaves_disembarked,
    mortality_ratio = voyage_slaves_numbers__imp_mortality_ratio,
    
    dis_broad = voyage_itinerary__imp_broad_region_slave_dis__name,
    dis_region = voyage_itinerary__imp_principal_region_slave_dis__name,
    dis_port = voyage_itinerary__imp_principal_port_slave_dis__name,
    
    purchase_region = voyage_itinerary__imp_principal_region_of_slave_purchase__name,
    purchase_place = voyage_itinerary__imp_principal_place_of_slave_purchase__name,
    
    ship_nationality = voyage_ship__imputed_nationality__name,
    
    outcome_slaves = voyage_outcome__outcome_slaves__name,
    outcome_owner = voyage_outcome__outcome_owner__name,
    outcome_type = voyage_outcome__particular_outcome__name)

intra <- intra %>%
  rename(
    year = voyage_dates__imp_arrival_at_port_of_dis_sparsedate__year,
    slaves_embarked = voyage_slaves_numbers__imp_total_num_slaves_embarked,
    slaves_disembarked = voyage_slaves_numbers__imp_total_num_slaves_disembarked,
    mortality_ratio = voyage_slaves_numbers__imp_mortality_ratio,
    
    dis_broad = voyage_itinerary__imp_broad_region_slave_dis__name,
    dis_region = voyage_itinerary__imp_principal_region_slave_dis__name,
    dis_port = voyage_itinerary__imp_principal_port_slave_dis__name,
    
    purchase_region = voyage_itinerary__imp_principal_region_of_slave_purchase__name,
    purchase_place = voyage_itinerary__imp_principal_place_of_slave_purchase__name,
    
    ship_nationality = voyage_ship__imputed_nationality__name,
    
    outcome_slaves = voyage_outcome__outcome_slaves__name,
    outcome_owner = voyage_outcome__outcome_owner__name,
    outcome_type = voyage_outcome__particular_outcome__name)

Clean the data:

Convert year to integer, slave numbers to numeric.

trans <- trans %>%
  mutate(
    year = as.integer(year),
    slaves_embarked = as.numeric(slaves_embarked),
    slaves_disembarked = as.numeric(slaves_disembarked))

intra <- intra %>%
  mutate(
    year = as.integer(year),
    slaves_embarked = as.numeric(slaves_embarked),
    slaves_disembarked = as.numeric(slaves_disembarked))

Filter out rows where slaves_disembarked is 0 or NA (incomplete voyages).

trans <- trans %>%
  filter(!is.na(slaves_disembarked), slaves_disembarked > 0)

intra <- intra %>%
  filter(!is.na(slaves_disembarked), slaves_disembarked > 0)

Filter for successful outcomes (e.g., “Slaves disembarked”, “Voyage completed”, “Sold slaves”).

success_pattern <- regex(
  "slaves disembarked|voyage completed|sold slaves|disembark|completed|sold|arriv|landed|delivered",
  ignore_case = TRUE)

trans <- trans %>%
  filter(
    str_detect(coalesce(outcome_slaves, ""), success_pattern) |
      str_detect(coalesce(outcome_type, ""), success_pattern))

intra <- intra %>%
  filter(
    str_detect(coalesce(outcome_slaves, ""), success_pattern) |
      str_detect(coalesce(outcome_type, ""), success_pattern))

Add new columns: decade (e.g., floor(year / 10) * 10), estimated_deaths (slaves_embarked - slaves_disembarked), is_us (TRUE if disembark is US-based, using dis_broad == “Mainland North America” or specific US regions/ports like “New Orleans”).

us_port_pattern <- regex(
  "New Orleans|Charleston|Savannah|Norfolk|New York|Newport|Baltimore|Philadelphia|Mobile|Pensacola|Galveston|Boston",
  ignore_case = TRUE
)

trans <- trans %>%
  mutate(
    decade = floor(year / 10) * 10,
    estimated_deaths = slaves_embarked - slaves_disembarked,
    is_us =
      dis_broad == "Mainland North America" |
      str_detect(coalesce(dis_port, ""), us_port_pattern) |
      str_detect(coalesce(dis_region, ""), regex("United States|USA", ignore_case = TRUE)))

intra <- intra %>%
  mutate(
    decade = floor(year / 10) * 10,
    estimated_deaths = slaves_embarked - slaves_disembarked,
    is_us =
      dis_broad == "Mainland North America" |
      str_detect(coalesce(dis_port, ""), us_port_pattern) |
      str_detect(coalesce(dis_region, ""), regex("United States|USA", ignore_case = TRUE)))

Combine the datasets with bind_rows(), adding a source_type column (“Trans-Atlantic” or “Intra-American”).

all_trades <- bind_rows(
  trans %>% mutate(source_type = "Trans-Atlantic"),
  intra %>% mutate(source_type = "Intra-American"))

Part 2: Analysis and Questions (5 points)

Total slaves imported to the US: Filter for is_us == TRUE, sum slaves_disembarked from both datasets.

us_total_disembarked <- all_trades %>%
  filter(is_us) %>%
  summarise(total_us_disembarked = sum(slaves_disembarked, na.rm = TRUE)) %>%
  pull(total_us_disembarked)

us_total_disembarked

## [1] 420091

Proportion of all slaves taken from Africa: Calculate US total / total slaves_embarked from Trans-Atlantic dataset (as this represents slaves taken from Africa).

africa_total_embarked <- trans %>%
  summarise(total_africa_embarked = sum(slaves_embarked, na.rm = TRUE)) %>%
  pull(total_africa_embarked)

prop_us_of_africa <- us_total_disembarked / africa_total_embarked

tibble(
  us_total_disembarked = us_total_disembarked,
  africa_total_embarked = africa_total_embarked,
  proportion_us_of_africa_embarked = prop_us_of_africa)

## # A tibble: 1 × 3
##   us_total_disembarked africa_total_embarked proportion_us_of_africa_embarked
##                  <dbl>                 <dbl>                            <dbl>
## 1               420091               8899661                           0.0472

Graph slave imports by decade to the US: Filter for US, group by decade, sum slaves_disembarked, plot as a bar graph with ggplot2.

us_by_decade <- all_trades %>%
  filter(is_us) %>%
  group_by(decade) %>%
  summarise(total_disembarked = sum(slaves_disembarked, na.rm = TRUE), .groups = "drop") %>%
  arrange(decade)

ggplot(us_by_decade, aes(x = factor(decade), y = total_disembarked)) +
  geom_col() +
  labs(
    title = "Slave Imports to the US by Decade (Trans-Atlantic + Intra-American)",
    x = "Decade",
    y = "Total slaves disembarked"
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Imports to the US by decade and region/port/state: Filter for US, group by decade, dis_region, dis_port (approximate state from port/region, e.g., “New Orleans” -> “Louisiana”), sum slaves_disembarked. Use a table and faceted bar plot.

us_by_decade_region_port <- all_trades %>%
  filter(is_us) %>%
  mutate(
    approx_state = case_when(
      str_detect(coalesce(dis_port, ""), regex("New Orleans", TRUE)) ~ "Louisiana",
      str_detect(coalesce(dis_port, ""), regex("Charleston", TRUE)) ~ "South Carolina",
      str_detect(coalesce(dis_port, ""), regex("Savannah", TRUE)) ~ "Georgia",
      str_detect(coalesce(dis_port, ""), regex("Norfolk", TRUE)) ~ "Virginia",
      str_detect(coalesce(dis_port, ""), regex("Baltimore", TRUE)) ~ "Maryland",
      str_detect(coalesce(dis_port, ""), regex("Philadelphia", TRUE)) ~ "Pennsylvania",
      str_detect(coalesce(dis_port, ""), regex("New York", TRUE)) ~ "New York",
      str_detect(coalesce(dis_port, ""), regex("Boston", TRUE)) ~ "Massachusetts",
      str_detect(coalesce(dis_port, ""), regex("Mobile", TRUE)) ~ "Alabama",
      str_detect(coalesce(dis_port, ""), regex("Pensacola", TRUE)) ~ "Florida",
      str_detect(coalesce(dis_port, ""), regex("Galveston", TRUE)) ~ "Texas",
      TRUE ~ "Other/Unknown"
    )
  ) %>%
  group_by(decade, dis_region, dis_port, approx_state) %>%
  summarise(total_disembarked = sum(slaves_disembarked, na.rm = TRUE), .groups = "drop") %>%
  arrange(decade, desc(total_disembarked))

Table view (top 25 rows by total)

us_by_decade_region_port %>%
  arrange(desc(total_disembarked)) %>%
  head(25) %>%
  knitr::kable()

decade	dis_region	dis_port	approx_state	total_disembarked
1800	South Carolina	Charleston	South Carolina	47743
1760	South Carolina	Charleston	South Carolina	28070
1730	South Carolina	Charleston	South Carolina	23306
1770	South Carolina	Charleston	South Carolina	23086
1750	South Carolina	Charleston	South Carolina	18901
1830	Gulf coast	New Orleans	Louisiana	18792
1840	Gulf coast	New Orleans	Louisiana	18493
1820	Gulf coast	New Orleans	Louisiana	18265
1780	South Carolina	Charleston	South Carolina	10061
1720	Virginia	York River	Other/Unknown	9058
1730	Virginia	York River	Other/Unknown	8088
1790	Georgia	Savannah	Georgia	7725
1850	Gulf coast	New Orleans	Louisiana	7235
1720	South Carolina	Charleston	South Carolina	5346
1700	Virginia	Virginia, port unspecified	Other/Unknown	5238
1740	Virginia	York River	Other/Unknown	5087
1780	Gulf coast	New Orleans	Louisiana	4644
1710	Virginia	York River	Other/Unknown	4404
1760	Virginia	Upper James River	Other/Unknown	4149
1770	Georgia	Savannah	Georgia	3797
1730	Maryland	Maryland, port unspecified	Other/Unknown	3604
1740	Virginia	Upper James River	Other/Unknown	3563
1760	Georgia	Savannah	Georgia	3557
1800	Gulf coast	New Orleans	Louisiana	3313
1700	Virginia	York River	Other/Unknown	3041

Faceted bar plot by approximate state

us_state_plot <- us_by_decade_region_port %>%
  filter(approx_state != "Other/Unknown") %>%
  group_by(decade, approx_state) %>%
  summarise(total_disembarked = sum(total_disembarked), .groups = "drop")

ggplot(us_state_plot, aes(x = factor(decade), y = total_disembarked)) +
  geom_col() +
  facet_wrap(~ approx_state, scales = "free_y") +
  labs(
    title = "US Slave Imports by Decade and Approximate State (from Port)",
    x = "Decade",
    y = "Total slaves disembarked"
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Countries participating in export from Africa, by decade: From Trans-Atlantic dataset, group by decade and voyage_ship__imputed_nationality__name (as “country”), count unique voyages or sum slaves_embarked. Display in a table.

countries_by_decade <- trans %>%
  group_by(decade, ship_nationality) %>%
  summarise(
    voyages = n_distinct(voyage_id),
    total_embarked = sum(slaves_embarked, na.rm = TRUE),
    .groups = "drop"
  ) %>%
  arrange(decade, desc(total_embarked))

show top 10 per decade

countries_by_decade %>%
  group_by(decade) %>%
  slice_max(total_embarked, n = 10, with_ties = FALSE) %>%
  ungroup()

## # A tibble: 215 × 4
##    decade ship_nationality  voyages total_embarked
##     <dbl> <chr>               <int>          <dbl>
##  1   1510 Portugal / Brazil       2            624
##  2   1510 0                       9            223
##  3   1510 Spain / Uruguay         8            144
##  4   1520 Spain / Uruguay         3           1043
##  5   1520 0                       3            597
##  6   1530 0                       9           1777
##  7   1530 Portugal / Brazil       2            560
##  8   1530 Spain / Uruguay         1            224
##  9   1540 0                      23           7750
## 10   1540 Portugal / Brazil       1            160
## # ℹ 205 more rows

Part 3: Visualizations + Summary

library(tidyverse)
library(scales)

Plot 1: US imports by decade (bar chart)

plot_us_by_decade <- ggplot(us_by_decade, aes(x = factor(decade), y = total_disembarked)) +
  geom_col() +
  labs(
    title = "Enslaved People Disembarked in the US by Decade",
    subtitle = "Combined Trans-Atlantic and Intra-American datasets (successful + complete voyages only)",
    x = "Decade",
    y = "Total disembarked"
  ) +
  scale_y_continuous(labels = comma) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

plot_us_by_decade

Plot 2: US imports by decade and approximate state (faceted bar chart)

us_by_decade_source <- all_trades %>%
  filter(is_us) %>%
  group_by(decade, source_type) %>%
  summarise(total_disembarked = sum(slaves_disembarked, na.rm = TRUE), .groups = "drop") %>%
  arrange(decade)

ggplot(us_by_decade_source, aes(x = factor(decade), y = total_disembarked, fill = source_type)) +
  geom_col() +
  labs(
    title = "US Slave Imports by Decade, Split by Dataset",
    subtitle = "Stacked totals show the relative contribution of Trans-Atlantic vs Intra-American arrivals",
    x = "Decade",
    y = "Total disembarked",
    fill = "Source type"
  ) +
  scale_y_continuous(labels = comma) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Summary of findings

After cleaning the data, the total number of enslaved people disembarked in the United States across both datasets was 420,091.
The total number embarked from Africa in the Trans-Atlantic dataset was 8,899,661. This means the US disembarkations in our cleaned data represent about 4.7% of the total embarked from Africa.
The decade-level bar chart shows that US arrivals were concentrated in specific decades rather than evenly spread over time.
The destination breakdown by inferred state (based on major ports) suggests that a small number of locations, most specifically South Carolina (Charleston) and Louisiana (New Orleans), account for large shares of arrivals in multiple decades. ```

Analyzing Trans-Atlantic and Intra-American Slave Trade with Tidyverse

Calum Jessett & Asher Rosen

2026-03-04

Rename long column names for readability (e.g., year, slaves_embarked, slaves_disembarked).

Clean the data:

Convert year to integer, slave numbers to numeric.

Filter out rows where slaves_disembarked is 0 or NA (incomplete voyages).

Filter for successful outcomes (e.g., “Slaves disembarked”, “Voyage completed”, “Sold slaves”).

Add new columns: decade (e.g., floor(year / 10) * 10), estimated_deaths (slaves_embarked - slaves_disembarked), is_us (TRUE if disembark is US-based, using dis_broad == “Mainland North America” or specific US regions/ports like “New Orleans”).

Combine the datasets with bind_rows(), adding a source_type column (“Trans-Atlantic” or “Intra-American”).

Part 2: Analysis and Questions (5 points)

Total slaves imported to the US: Filter for is_us == TRUE, sum slaves_disembarked from both datasets.

Proportion of all slaves taken from Africa: Calculate US total / total slaves_embarked from Trans-Atlantic dataset (as this represents slaves taken from Africa).

Graph slave imports by decade to the US: Filter for US, group by decade, sum slaves_disembarked, plot as a bar graph with ggplot2.

Imports to the US by decade and region/port/state: Filter for US, group by decade, dis_region, dis_port (approximate state from port/region, e.g., “New Orleans” -> “Louisiana”), sum slaves_disembarked. Use a table and faceted bar plot.

Table view (top 25 rows by total)

Faceted bar plot by approximate state

Countries participating in export from Africa, by decade: From Trans-Atlantic dataset, group by decade and voyage_ship__imputed_nationality__name (as “country”), count unique voyages or sum slaves_embarked. Display in a table.

show top 10 per decade

Part 3: Visualizations + Summary

Plot 1: US imports by decade (bar chart)

Plot 2: US imports by decade and approximate state (faceted bar chart)

Summary of findings