Mini Project 3

Background

The US Constitution sets the basic rules of electing the President in Section 1 of Article II, which we quote here in part:

Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress: but no Senator or Representative, or Person holding an Office of Trust or Profit under the United States, shall be appointed an Elector.

The Electors shall meet in their respective States, and vote by Ballot for two Persons, of whom one at least shall not be an Inhabitant of the same State with themselves. And they shall make a List of all the Persons voted for, and of the Number of Votes for each; which List they shall sign and certify, and transmit sealed to the Seat of the Government of the United States, directed to the President of the Senate. The President of the Senate shall, in the Presence of the Senate and House of Representatives, open all the Certificates, and the Votes shall then be counted. The Person having the greatest Number of Votes shall be the President, if such Number be a Majority of the whole Number of Electors appointed; and if there be more than one who have such Majority, and have an equal Number of Votes, then the House of Representatives shall immediately chuse by Ballot one of them for President; and if no Person have a Majority, then from the five highest on the List the said House shall in like Manner chuse the President. But in chusing the President, the Votes shall be taken by States, the Representation from each State having one Vote; A quorum for this Purpose shall consist of a Member or Members from two thirds of the States, and a Majority of all the States shall be necessary to a Choice. In every Case, after the Choice of the President, the Person having the greatest Number of Votes of the Electors shall be the Vice President. But if there should remain two or more who have equal Votes, the Senate shall chuse from them by Ballot the Vice President. Though the details have varied over time due to amendment, statue, and technology, this basic outline of this allocation scheme remains unchanged:

Each state gets electoral college votes, where is the number of Representatives that state has in the US House of Representatives. In this mini-project, you can use the number of districts in a state to determine the number of congressional representatives (one per district). States can allocate those votes however they wish The president is the candidate who receives a majority of electoral college votes Notably, the Constitution sets essentially no rules on how the electoral college votes (ECVs) for a particular state are allocated. At different points in history, different states have elected to use each of the following:

Direct allocation of ECVs by state legislature (no vote) Allocation of all ECVs to winner of state-wide popular vote Allocation of all ECVs to winner of nation-wide popular vote Allocation of ECVs to popular vote winner by congressional district + allocation of remaining ECVs to the state-wide popular vote winner Currently, only Maine and Nebraska use the final option; the other 48 states and the District of Columbia award all ECVs to the winner of their state-wide popular vote. We emphasize here that “statewide winner-take-all” is a choice made by the individual states, not dictated by the US constitution, and that states have the power to change it should they wish.1

To my knowledge, no US state uses true proportionate state-wide representation, though I believe such a ECV-allocation scheme would be consistent with the US Constitution. For example, if a state with 5 ECVs had 60,000 votes for Candidate A and 40,000 cast for Candidate B, it could award 3 ECVs to A and 2 to B, regardless of the spatial distribution of those votes within the state.

Set-Up and Initial Exploration

Data I: US House Election Votes from 1976 to 2022

The MIT Election Data Science Lab collects votes from all biennial congressional races in all 50 states here. Download this data as a CSV file using your web browser. Note that you will need to provide your contact info and agree to cite this data set in your final report.

Additionally, download statewide presidential vote counts from 1976 to 2022 here. As before, it will likely be easiest to download this data by hand using your web browser.

Data II: Congressional Boundary Files 1976 to 2012

Jeffrey B. Lewis, Brandon DeVine, Lincoln Pritcher, and Kenneth C. Martis have created shapefiles for all US congressional districts from 1789 to 2012; they generously make these available here.

We are going to download those with the following code:

Code 
# Task 1: import congressional district data from cdmaps from 1976 to 2012
get_cdmaps_file <- function(fname) {
  BASE_URL <- "https://cdmaps.polisci.ucla.edu/shp/"
  fname_ext <- paste0(fname, ".zip")
  if (!file.exists(fname_ext)) {
    FILE_URL <- paste0(BASE_URL, fname_ext)
    download.file(FILE_URL,
                  destfile = fname_ext
    )
  }
}
options(timeout = 180) # keep the downloads from potentially timing out

get_cdmaps_file("districts112") # January 5, 2011 to January 3, 2013
get_cdmaps_file("districts111") # January 6, 2009 to December 22, 2010
get_cdmaps_file("districts110") # January 4, 2007 to January 3, 2009
get_cdmaps_file("districts109") # January 4, 2005 to December 9, 2006
get_cdmaps_file("districts108") # January 7, 2003 to December 8, 2004
get_cdmaps_file("districts107") # January 3, 2001 to November 22, 2002
get_cdmaps_file("districts106") # January 6, 1999 to December 15, 2000
get_cdmaps_file("districts105") # January 7, 1997 to December 19, 1998
get_cdmaps_file("districts104") # January 4, 1995 to October 4, 1996
get_cdmaps_file("districts103") # January 5, 1993 to December 1, 1994 
get_cdmaps_file("districts102") # January 3, 1991 to October 9, 1992
get_cdmaps_file("districts101") # January 3, 1989 to October 28, 1990
get_cdmaps_file("districts100") # January 6, 1987 to October 22, 1988
get_cdmaps_file("districts099")  # January 3, 1985 to October 18, 1986
get_cdmaps_file("districts098")  # January 3, 1983 to October 12, 1984
get_cdmaps_file("districts097")  # January 5, 1981 to December 23, 1982
get_cdmaps_file("districts096")  # January 15, 1979 to December 16, 1980
get_cdmaps_file("districts095")  # January 4, 1977 to October 15, 1978
get_cdmaps_file("districts094")  # January 14, 1975 to October 1, 1976

Data III: Congressional Boundary Files 2014 to Present

To get district boundaries for more recent congressional elections, we can turn to the US Census Bureau. Unfortunately, these data - while authoritative and highly detailed - are not in quite the same format as our previous congressional boundary files. We can review the US Census Bureau shape files online and download thm with the following code:

Code 
# download shape files for 113th congress
if (!file.exists("districts113.zip")) {
  download.file("https://www2.census.gov/geo/tiger/TIGER2013/CD/tl_2013_us_cd113.zip",
    destfile = "districts113.zip"
  )
}

# download shape files for 114th congress
if (!file.exists("districts114.zip")) {
  download.file("https://www2.census.gov/geo/tiger/TIGER2014/CD/tl_2014_us_cd114.zip",
    destfile = "districts114.zip"
  )
}

# download shape files for 115th congress
if (!file.exists("districts115.zip")) {
  download.file("https://www2.census.gov/geo/tiger/TIGER2016/CD/tl_2016_us_cd115.zip",
    destfile = "districts115.zip"
  )
}

# download shape files for 116th congress
if (!file.exists("districts116.zip")) {
  download.file("https://www2.census.gov/geo/tiger/TIGER2018/CD/tl_2018_us_cd116.zip",
    destfile = "districts116.zip"
  )
}

Initial Exploration of Vote Count Data

Now that we have our data downloaded, let’s explore our vote counts per election by answering the following questions.

  1. Which states have gained and lost the most seats in the US House of Representatives between 1976 and 2022?
Code 
# find the number of seats each state has in 1976 and 2022
house_seats_1976_2022 <- house_rep_vote_count |>
  filter(year %in% c(1976, 2022)) |>
  group_by(year, state) |>
  summarise(total_seats = n_distinct(district)) |> # count number of seats by number of electoral districts
  select(year, state, total_seats)

# pivot table to find difference easily: state, 1976 seats, 2022 seats
house_seats_1976_2022_wide <- house_seats_1976_2022 |>
  pivot_wider(names_from = year, values_from = total_seats, names_prefix = "total_seats_") |>
  mutate(difference = total_seats_2022 - total_seats_1976)

# find the change in seats from 2022 to 1976
seat_changes <- house_seats_1976_2022_wide |>
  select(state, difference)

# visual representation
seat_changes_filtered <- seat_changes |> # excluding zero for visual aesthetic 
  filter(difference != 0)

ggplot(seat_changes_filtered, aes(x = reorder(state, difference), y = difference, fill = difference > 0)) +
  geom_bar(stat = "identity", show.legend = FALSE) +
  scale_fill_manual(values = c("salmon2", "cornflowerblue")) +  # Blue for increases, red for decreases
  coord_flip() +  # Flip coordinates for horizontal bars
  labs(title = "House Seats Gained/Lost by State (1976-2022)",
       x = "State",
       y = "Change in Seats") +
  theme_minimal()

We can clearly see that Texas is the State that gained the most seats and New York is the State that lost the most seats between 1976 and 2022.

  1. New York State has a unique “fusion” voting system where one candidate can appear on multiple “lines” on the ballot and their vote counts are totaled. For instance, in 2022, Jerrold Nadler appeared on both the Democrat and Working Families party lines for NYS’ 12th Congressional District. He received 200,890 votes total (184,872 as a Democrat and 16,018 as WFP), easily defeating Michael Zumbluskas, who received 44,173 votes across three party lines (Republican, Conservative, and Parent).

Are there any elections in our data where the election would have had a different outcome if the “fusion” system was not used and candidates only received the votes their received from their “major party line” (Democrat or Republican) and not their total number of votes across all lines?

Code 
house_election_winner <- house_rep_vote_count |>
  group_by(year, state, district, candidate) |>
  summarize(total_votes = sum(candidatevotes), .groups = 'drop') |> # aggregate votes across different parties on ticket for fusion ticket candidates
  group_by(year, state, district) |>
  slice_max(order_by = total_votes, n = 1, with_ties = FALSE) |> # find the winner based on who has the most total votes
  rename(historical_winner = candidate) # rename for conventional understanding
  
# find winner without fusion system
primary_party_winner <- house_rep_vote_count |>
  group_by(year, state, district, candidate, party) |>
  summarize(primary_party_votes = sum(candidatevotes), .groups = "drop") |> 
  group_by(year, state, district) |>
  slice_max(order_by = primary_party_votes, n = 1, with_ties = FALSE) |>
  rename(single_party_winner = candidate) |> # rename for conventional understanding
  select(-party) # deselecting since we are only interested in the candidate name and votes

# find any elections where the historical winner is not the same as the single major party winner
potential_election_changes <- house_election_winner |>
  left_join(primary_party_winner, by = c("year", "state", "district")) |>
  mutate(different_outcome = historical_winner != single_party_winner) |> # create a logic column checks if the outcomes were the same or not
  filter(different_outcome == TRUE) |> # filter where the historical winner is not the same as the single party vote winner
  select(-different_outcome)

kable(
  potential_election_changes,
  col.names = c(
    "Year", "State", "District", 
    "Historical Winner", "Votes (with Fusion)", 
    "Single Party Winner", "Votes (Single Party)"
  ),
  caption = "Influence of Fusion Voting on House Elections"
)
Influence of Fusion Voting on House Elections
Year State District Historical Winner Votes (with Fusion) Single Party Winner Votes (Single Party)
1976 NEW YORK 29 EDWARD W PATTISON 100663 JOSEPH A MARTINO 96476
1980 NEW YORK 3 GREGORY W CARMAN 87952 JEROME A AMBRO JR 75389
1980 NEW YORK 6 JOHN LEBOUTILLIER 89762 LESTER L WOLFF 74319
1984 NEW YORK 20 JOSEPH J DIOGUARDI 106958 OREN J TEICHER 102842
1986 NEW YORK 27 GEORGE C WORTLEY 83430 ROSEMARY S POOLER 81133
1992 CONNECTICUT 2 SAM GEJDENSON 123291 EDWARD W MUNSTER 119416
1992 NEW YORK 3 PETER T KING 124727 STEVE A ORLINS 116915
1994 NEW YORK 1 MICHAEL P FORBES 90491 GEORGE J HOCHBRUECKNER 78692
1996 NEW YORK 1 MICHAEL P FORBES 116620 NORA L BREDES 93816
1996 NEW YORK 30 JACK QUINN 121369 FRANCIS J PORDUM 97686
2000 CONNECTICUT 2 ROB SIMMONS 114380 SAM GEJDENSON 111520
2006 NEW YORK 25 JAMES T WALSH 110525 DAN MAFFEI 100605
2006 NEW YORK 29 JOHN R “RANDY” KUHL JR 106077 ERIC J MASSA 94609
2010 NEW YORK 13 MICHAEL G GRIMM 65024 MICHAEL E MCMAHON 60773
2010 NEW YORK 19 NAN HAYMORTH 109956 JOHN J HALL 98766
2010 NEW YORK 24 RICHARD L HANNA 101599 MICHAEL A ARCURI 89809
2010 NEW YORK 25 ANN MARIE BUERKLE 104602 DANIEL B MAFFEI 103954
2012 NEW YORK 27 CHRIS COLLINS 161220 KATHLEEN C HOCHUL 140008
2018 NEW YORK 1 LEE M ZELDIN 139027 PERRY GERSHON 124213
2018 NEW YORK 24 JOHN M KATKO 136920 DANA BALTER 115902
2018 NEW YORK 27 CHRIS COLLINS 140146 NATHAN D MCMURRAY 128167
2022 NEW YORK 4 ANTHONY P D’ESPOSITO 140622 LAURA A GILLEN 130871
2022 NEW YORK 17 MICHAEL V LAWLER 143550 SEAN PATRICK MALONEY 133457
2022 NEW YORK 22 BRANDON M WILLIAMS 135544 FRANCIS CONOLE 132913

We can see in this table all of the elections that would have had a different single party winner if a fusion of parties would have not happened. We also noticed that the fusion system was mostly use in New York State.

  1. Do presidential candidates tend to run ahead of or run behind congressional candidates in the same state? That is, does a Democratic candidate for president tend to get more votes in a given state than all Democratic congressional candidates in the same state?

Does this trend differ over time? Does it differ across states or across parties? Are any presidents particularly more or less popular than their co-partisans?

Code 
#Aggregate Votes for House Candidates by Year, State, and Party
congressional_party_votes <- house_rep_vote_count |>
  group_by(year, state, party) |>
  summarize(total_congressional_votes = sum(candidatevotes, na.rm = TRUE), .groups = "drop")

#Aggregate Votes for Presidential Candidates by Year, State, and Party
presidential_party_votes <- presidential_vote_count |>
  group_by(year, state, party_detailed) |>
  summarize(total_presidential_votes = sum(candidatevotes, na.rm = TRUE), .groups = "drop") |>
  rename(party = party_detailed) # Renaming for easier joining

#Calculate Vote Disparity Between Presidential and Congressional Votes
vote_disparity <- presidential_party_votes |>
  inner_join(congressional_party_votes, by = c("year", "state", "party")) |>
  mutate(vote_difference = total_presidential_votes - total_congressional_votes) |>
  select(-total_presidential_votes, -total_congressional_votes)

#Filter for Democrat and Republican Parties
vote_disparity_year <- vote_disparity |>
  filter(party %in% c("DEMOCRAT", "REPUBLICAN")) |>
  group_by(year, party) |>
  summarize(total_vote_difference = sum(vote_difference, na.rm = TRUE), .groups = "drop")

#Visualize Vote Disparity Trends Over Time by Party
ggplot(vote_disparity_year, aes(x = year, y = total_vote_difference, color = party, group = party)) +
  geom_line(size = 1) +
  geom_point() +
  scale_color_manual(values = c("DEMOCRAT" = "red", "REPUBLICAN" = "darkblue")) + # Set color for each party
  labs(
    title = "Presidential vs Congressional Vote Disparity (Democrats & Republicans)",
    subtitle = "Total Vote Difference by Year",
    x = "Year",
    y = "Total Vote Difference",
    color = "Party"
  ) +
  theme_minimal()

Code 
#Display Results in a Table
kable(
  vote_disparity_year,
  caption = "Vote Disparity Between Presidential and Congressional Candidates by Party and Year",
  col.names = c("Year", "Party", "Total Vote Difference")
)
Vote Disparity Between Presidential and Congressional Candidates by Party and Year
Year Party Total Vote Difference
1976 DEMOCRAT -944354
1976 REPUBLICAN 7648467
1980 DEMOCRAT -3850568
1980 REPUBLICAN 6550533
1984 DEMOCRAT -5529685
1984 REPUBLICAN 15597052
1988 DEMOCRAT -1876590
1988 REPUBLICAN 11599224
1992 DEMOCRAT -3728896
1992 REPUBLICAN -4543819
1996 DEMOCRAT 3538546
1996 REPUBLICAN -4441137
2000 DEMOCRAT 4313035
2000 REPUBLICAN 3543123
2004 DEMOCRAT 5900998
2004 REPUBLICAN 6128252
2008 DEMOCRAT 4247430
2008 REPUBLICAN 7545117
2012 DEMOCRAT 6180158
2012 REPUBLICAN 2670821
2016 DEMOCRAT 4436567
2016 REPUBLICAN -188055
2020 DEMOCRAT 4124029
2020 REPUBLICAN 1525747

Over time, Democratic presidential candidates have garnered more votes nationwide compared to their party’s House candidates. Meanwhile, Republican presidential nominees have seen a gradual decline in popularity, often receiving vote totals closer to those of their House counterparts.

The 1984 and 1988 elections stand out as exceptions when Republican presidential candidates outperformed their House colleagues by over 10 million votes. Ronald Reagan in 1984 and George H.W. Bush in 1988 achieved this milestone, with both securing victories in their respective presidential races.

Importing and Plotting Shape File Data

Now let’s import and plot the shape files data to explore further the Presidential Election State Results.

We can unzip the shape files with the following code:

Code 
read_shp_from_zip <- function(zip_file) {
  temp_dir <- tempdir() # Create a temporary directory
  zip_contents <- unzip(zip_file, exdir = temp_dir) # Unzip the contents and
  shp_file <- zip_contents[grepl("\\.shp$", zip_contents)] # filter for .shp files
  sf_object <- read_sf(shp_file) # Read the .shp file into an sf object
  return(sf_object) # Return the sf object
}

Using the data we downloaded earlier, we are going to create a chloropleth visualization of the electoral college results for the 2000 presidential election (Bush vs. Gore), coloring each state by the party that won the most votes in that state.

Code 
# This represents which election year we want to analyze
create_election_year_data <- function(year_to_look_for) {
  
  # Filter presidential data for the specified year and the two main parties
  president_data_for_given_year <- read_csv("1976-2020-president.csv") |>
    filter(
      year == year_to_look_for,
      office == "US PRESIDENT",
      party_simplified %in% c("DEMOCRAT", "REPUBLICAN")
    ) |>
    # Grouping the data by state
    group_by(state) |>
    # Select candidate with max vote
    slice_max(candidatevotes, n = 1, with_ties = FALSE) |>
    # Select the relevant columns
    select(state, state_po, candidate, party_simplified, candidatevotes)

  # Filter congressional vote count data for the specified year and summarize Electoral College Votes (ECV)
  ecv_data_for_given_year <- read_csv("1976-2022-house.csv") |>
    filter(year == year_to_look_for, office == "US HOUSE") |>
    group_by(state) |>
    summarize(
      num_representatives = n_distinct(district),
      # Count 2 senators for each state
      ecv = num_representatives + 2,
      .groups = "drop"
    ) |>
    # Add DC manually with its 3 electoral votes
    bind_rows(data.frame(state = "DISTRICT OF COLUMBIA", ecv = 3))

  # Combine presidential data with ECV data and format state names
  combined_data <- president_data_for_given_year |>
    # Left_join will merge 
    left_join(ecv_data_for_given_year, by = "state") |>
    # Formats state names to title case
    mutate(state = str_to_title(state))

  return(combined_data)
}
Code 
# Let's create function to take an argument(year)
create_state_shp_data <- function(year) {
  # Find file location for a given year
  file_location_for_given_year <- "districts106.zip"
  # This calls the previously defined function (read_shp_from_zip) to extract and load the shape file data for the specified year
  shapefile_us <- read_shp_from_zip(file_location_for_given_year)

  # Standardize STATENAME to title case
  congressional_districts_for_given_year <- shapefile_us |>
    mutate(STATENAME = str_to_title(trimws(STATENAME))) |> # Convert to title case (e.g., "new york" becomes "New York")
    rename(state = STATENAME) |> # Rename column to 'state' for consistency
    select(state, geometry) |> #select the given column
    st_make_valid()

# Let's aggregate the individual congressional districts into a single geometry for each state
  state_level_shape_for_given_year <- congressional_districts_for_given_year |>
    group_by(state) |> # groupby state
    reframe(geometry = st_union(geometry)) |> # Use reframe to handle ungrouping automatically
    distinct(state, .keep_all = TRUE) # Ensure unique entries remain for each state 

  return(state_level_shape_for_given_year)
}
Code 
create_election_map <- function(election_year_data, shp_data_for_year, year_num, file_name = NULL) {
  yearly_data <- shp_data_for_year |>
    inner_join(election_year_data, by = "state")

  # Define the small northeastern states to display abbreviations outside
  yearly_data <- yearly_data |>
    mutate(
      state_abbr = ifelse(state %in% c(
        "Connecticut", "Delaware", "Maryland", "Massachusetts",
        "New Jersey", "New Hampshire", "Rhode Island", "Vermont"
      ), state_po, NA), # Use abbreviations only for small northeastern states
      label_text = ifelse(!is.na(state_abbr), paste0(state_abbr, " ", ecv), NA)
    )

  northeastern_states <- c(
    "Connecticut", "Delaware", "Maryland", "Massachusetts",
    "New Jersey", "New Hampshire", "Rhode Island", "Vermont"
  )

  if (!inherits(yearly_data, "sf")) {
    yearly_data <- st_as_sf(yearly_data)
  }

  # Prepare the main map excluding Alaska and Hawaii
  mainland <- yearly_data |> filter(!state %in% c("Alaska", "Hawaii"))
  alaska <- yearly_data |> filter(state == "Alaska")
  hawaii <- yearly_data |> filter(state == "Hawaii")


  mainland_plot <- ggplot(mainland) +
    geom_sf(aes(fill = party_simplified)) +

    # Only show electoral votes for non-northeastern states
    geom_sf_text(
      data = mainland |> filter(!state %in% northeastern_states),
      aes(label = ecv), size = 6, color = "black", fontface = "bold"
    ) +

    # Add labels for small northeastern states with electoral counts outside
    geom_text_repel(
      data = yearly_data |> filter(!is.na(state_abbr)), # Only for small northeastern states
      aes(
        geometry = geometry,
        label = label_text
      ),
      stat = "sf_coordinates", # Use spatial coordinates for label placement
      size = 4,
      color = "black",
      fontface = "bold", # Make labels bold
      nudge_x = 5, # Set a strong rightward nudge to move all labels to the right
      nudge_y = 0, # Keep y-nudging minimal for alignment
      hjust = 0, # Left-align labels
      direction = "y", # Keep lines vertically aligned
      lineheight = 0.9 # Adjust line height for clarity if needed
    ) +
    scale_fill_manual(values = c("DEMOCRAT" = "darkblue", "REPUBLICAN" = "red", "Other" = "gray")) +
    labs(
      title = paste0(year_num, " Presidential Election Electoral College Results"),
      fill = "Winning Party"
    ) +
    theme_minimal() +
    theme(
      plot.title = element_text(hjust = 0.5, size = 20, face = "bold"), # Bold title
      legend.position = "bottom"
    ) +
    coord_sf(expand = FALSE)

  # Alaska plot without legend
  alaska_plot <- ggplot(alaska) +
    geom_sf(aes(fill = party_simplified)) +
    geom_sf_text(aes(label = ecv), size = 6, color = "black", fontface = "bold") +
    scale_fill_manual(values = c("DEMOCRAT" = "darkblue", "REPUBLICAN" = "red", "Other" = "gray")) +
    theme_void() +
    theme(legend.position = "none") + # Hide legend in Alaska inset
    coord_sf(xlim = c(-180, -130), ylim = c(50, 72), expand = FALSE)

  # Hawaii plot without legend
  hawaii_plot <- ggplot(hawaii) +
    geom_sf(aes(fill = party_simplified)) +
    geom_sf_text(aes(label = ecv), size = 6, color = "black", fontface = "bold") +
    scale_fill_manual(values = c("DEMOCRAT" = "darkblue", "REPUBLICAN" = "red", "Other" = "gray")) +
    theme_void() +
    theme(legend.position = "none") + # Hide legend in Hawaii inset
    coord_sf(xlim = c(-161, -154), ylim = c(18, 23), expand = FALSE)

  # Combine plots using patchwork with adjusted insets on bottom left
  combined_plot <- mainland_plot +
    inset_element(alaska_plot, left = 0.05, bottom = 0.1, right = 0.25, top = 0.3) + # Alaska on bottom left
    inset_element(hawaii_plot, left = 0.05, bottom = 0.05, right = 0.25, top = 0.15) # Hawaii below Alaska on bottom left
  # Save plot
  
   if (is.null(file_name)) {
      ggsave(filename = paste0("mp03data/election_maps/election_map_", year_num, ".png"), plot = combined_plot, width = 16, height = 12, dpi = 300)
   }else{
     ggsave(filename = file_name, plot = combined_plot, width = 16, height = 12, dpi = 300)
   }
  
}
Code 
library(ggrepel) 
library(patchwork)

# Calls the function with 2000 as the argument to generate data for the 2000 presidential election

election_year_data <- create_election_year_data(2000) # The argument to create spatial data for each state in 2000 by combining congressional district geometries into state-level boundaries. 
election_year_shp_data <- create_state_shp_data(2000) # To create and save a map of the 2000 presidential 
create_election_map(election_year_data, election_year_shp_data, 2000, "election_map_2000.png")  

knitr::include_graphics("election_map_2000.png")

Now let’s try another code to make our map animated and show the election results over time per state.

Code 
# Write a function that gets the presidential election winners in each desired election year
election_years <- c(1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020) # election years in the data

# Include the year
get_winner_election_year <- function(input_year) {
  presidential_vote_count |>
    filter(year == input_year) |> # Filter for the specific year
    group_by(state, party_simplified) |>
    summarize(total_votes = sum(candidatevotes), .groups = "drop") |>
    group_by(state) |>
    slice_max(total_votes, n = 1) |>
    ungroup() |>
    select(state, party_simplified) |>
    rename(winning_party = party_simplified) |>
    mutate(year = input_year) # add year to the table
}

# Bind the results into one table for time frame animation
election_winner_by_year <- bind_rows(lapply(election_years, get_winner_election_year)) # lapply() will take the arguments in election_years and input it into the get_winner_election_year function

shapefile_states <- read_shp_from_zip("tl_2018_us_state.zip")
# Join the US States shape file to the list of election winners
shapefile_states <- shapefile_states|>
  mutate(NAME = toupper(trimws(NAME))) |> 
  left_join(election_winner_by_year,
    join_by(NAME == state),
    relationship = "many-to-many"
  ) |>
  filter(!is.na(year))

# Animated plot
animate_election <- ggplot(shapefile_states,
  aes(
    geometry = geometry,
    fill = winning_party
  ),
  color = "black"
) +
  geom_sf() +
  scale_fill_manual(values = c("DEMOCRAT" = "darkblue", "REPUBLICAN" = "firebrick1")) +
  theme_minimal() +
  labs(
    title = "Presidential Election State Results in {closest_state}",
    fill = "Winning Party"
  ) +
  theme(legend.position = "bottom") +
  transition_states(year, transition_length = 0, state_length = 1) +
  coord_sf(xlim = c(-175, -60), expand = FALSE)

animate(animate_election, renderer = gifski_renderer(file = paste0(save_directory, "/election_results_animation.gif"))) # save as a GIF

Comparing the Effects of ECV Allocation Rules

To compare the effecrs of ECV allocation rules, we are going to go through the historical voting data and assign each state’s ECVs according to various strategies:

  • State-Wide Winner-Take-All
  • District-Wide Winner-Take-All + State-Wide “At Large” Votes
  • State-Wide Proportional
  • National Proportional

We wil answer the following question:what patterns do you see? Are the results generally consistent or are one or more methods systematically more favorable to one party?

First, we identify the number of electoral votes each state has in each election year with the following code:

Code 
# find number of electoral votes each state has each year
electoral_votes_by_state <- house_rep_vote_count |>
  group_by(year, state) |>
  summarize(total_reps = n_distinct(district)) |>
  mutate(electoral_votes = total_reps + 2) |> # R+2 votes
  select(year, state, electoral_votes)

# add DC votes
electoral_votes_by_state <- electoral_votes_by_state |>
  ungroup() |>
  add_row(year = 1976, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 1980, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 1984, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 1988, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 1992, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 1996, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2000, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2004, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2008, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2012, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2016, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  add_row(year = 2020, state = "DISTRICT OF COLUMBIA", electoral_votes = 3) |>
  distinct() # avoid any duplicate entries

Now, let’s see the effects of State-Wide Winner-Take-All scheme with the following code:

Code 
# Find the candidate with the most votes each year in each state
state_wide_winner_take_all <- presidential_vote_count |>
  group_by(year, state, candidate) |>
  summarize(total_votes = sum(candidatevotes), .groups = "drop") |>
  group_by(year, state) |>
  slice_max(order_by = total_votes, n = 1, with_ties = FALSE) |> # Find the winner of each state
  rename(winner = candidate) # Rename for clarity

# Join state, winner, and number of electoral votes & sum electoral votes for the winning candidate
state_wide_winner_take_all <- state_wide_winner_take_all |>
  left_join(electoral_votes_by_state,
    by = c("year", "state")
  ) |>
  group_by(year, winner) |>
  summarize(total_electoral_votes = sum(electoral_votes)) |> # Sum electoral votes for each year
  slice_max(order_by = total_electoral_votes, n = 1, with_ties = FALSE)

# Create a knitr table
state_wide_winner_take_all |>
  setNames(c("Year", "Winning Candidate", "Electoral Votes")) |>
  kable(
    caption = "Table 5: State-Wide Winner-Take-All: Presidential Winning Candidate",
    format = "html",
    col.names = c("Year", "Winning Candidate", "Electoral Votes"),
    align = "c"
  )
Table 5: State-Wide Winner-Take-All: Presidential Winning Candidate
Year Winning Candidate Electoral Votes
1976 CARTER, JIMMY 297
1980 REAGAN, RONALD 489
1984 REAGAN, RONALD 525
1988 BUSH, GEORGE H.W. 426
1992 CLINTON, BILL 370
1996 CLINTON, BILL 379
2000 BUSH, GEORGE W. 271
2004 BUSH, GEORGE W. 286
2008 OBAMA, BARACK H. 364
2012 OBAMA, BARACK H. 332
2016 TRUMP, DONALD J. 305
2020 BIDEN, JOSEPH R. JR 306

From the table summarizing historical presidential election results using the state-wide winner-take-all system, several observations can be made:

  1. Landslide Victories: Ronald Reagan’s elections in 1980 (489 electoral votes) and 1984 (525 electoral votes) showcase some of the largest margins of victory, indicating strong bipartisan appeal or a dominant Republican electorate during those years. Other significant victories include George H.W. Bush in 1988 (426 electoral votes) and Barack Obama in 2008 (364 electoral votes), suggesting periods of broad public support for the winning candidate.
  2. Close Contests: The elections of 2000 (George W. Bush with 271 votes) and 2020 (Joe Biden with 306 votes) highlight narrow wins, reflecting closely divided electorates. These results demonstrate the pivotal role of swing states in determining outcomes under the winner-take-all rule.
  3. Partisan Shifts: Democratic candidates like Jimmy Carter (1976), Bill Clinton (1992, 1996), and Barack Obama (2008, 2012) consistently achieved strong showings, suggesting periods of Democratic dominance in national politics. Republican successes in the 1980s (Reagan and Bush) and early 2000s (George W. Bush) illustrate periods of conservative momentum.
  4. Electoral Votes Over Time: The total number of electoral votes for the winning candidate has fluctuated significantly, reflecting changes in political landscapes and voter behavior in various states over time. These trends provide insights into the dynamics of the U.S. Electoral College system and the impact of the winner-take-all rule on presidential election outcomes.

We continue by analyzing the District-Wide Winner-Take-All + State-Wide “At Large” Votes scheme, assuming the winning party of the house district is the same for the president.

Code 
# Find number of districts each party won to represent electoral votes won in each state
district_winner <- house_rep_vote_count |>
  group_by(year, state, district) |>
  slice_max(order_by = candidatevotes, n = 1, with_ties = FALSE) |>
  select(year, state, district, party) |>
  group_by(year, state, party) |>
  summarize(districts_won = n()) # number of electoral votes received by each party

# Find popular vote winner in the state
at_large_winner <- house_rep_vote_count |>
  group_by(year, state) |>
  slice_max(order_by = candidatevotes, n = 1, with_ties = FALSE) |>
  select(year, state, party) |>
  add_column(at_large_votes = 2) # designating the vote count

# Join tables together to find total electoral votes the presidential party receives in each state
district_wide_winner_take_all <- district_winner |>
  left_join(at_large_winner, by = c("year", "state", "party")) |>
  mutate(across(where(is.numeric), ~ ifelse(is.na(.), 0, .))) |> # Set NA to 0 for rows with no resulting joins
  mutate(total_electoral_votes = districts_won + at_large_votes) |>
  select(-districts_won, -at_large_votes) |>
  rename(party_simplified = party) |> # Rename for easier joining convention
  left_join(presidential_vote_count, by = c("year", "state", "party_simplified")) |> # Join to presidential candidate
  select(year, state, total_electoral_votes, candidate) |>
  group_by(year, candidate) |>
  summarize(electoral_votes = sum(total_electoral_votes)) |>
  slice_max(order_by = electoral_votes, n = 1, with_ties = FALSE) |>
  drop_na() # Get rid of the non-presidential election years

# Render table with knitr::kable
district_wide_winner_take_all |>
  setNames(c("Year", "Winning Candidate", "Electoral Votes")) |> # Rename columns
  kable(
    caption = "District-Wide Winner-Take-All: Presidential Winning Candidate",
    col.names = c("Year", "Winning Candidate", "Electoral Votes"),
    align = "c"
  )
District-Wide Winner-Take-All: Presidential Winning Candidate
Year Winning Candidate Electoral Votes
1976 CARTER, JIMMY 353
1980 CARTER, JIMMY 280
1984 MONDALE, WALTER 300
1988 DUKAKIS, MICHAEL 299
1992 CLINTON, BILL 296
1996 DOLE, ROBERT 287
2000 BUSH, GEORGE W. 294
2004 BUSH, GEORGE W. 296
2008 OBAMA, BARACK H. 312
2012 ROMNEY, MITT 275
2016 TRUMP, DONALD J. 300
2020 BIDEN, JOSEPH R. JR 269

From the District-Wide Winner-Take-All + State-Wide “At Large” Votes table, we can conclude that:

  1. Electoral Votes Vary: The total electoral votes for the winning candidate fluctuate by year, showing how electoral vote distribution changes with time.
  2. Impact of At-Large Votes: Winning the popular vote in a state grants an additional 2 at-large electoral votes, which can influence the outcome when combined with district wins.
  3. Electoral Trends: While some years show a large margin of electoral votes (e.g., Carter in 1976), others, like Trump in 2016, reflect more competitive outcomes with around 300 electoral votes.
  4. State-Wide and District Effects: District-wide wins combined with state-wide “at large” votes affect the final presidential electoral count, often reflecting party dominance in the state.

Now, we will at the State-Wide Proportional scheme effects on presidential elections.

Code 
# Find the percentage of the votes received in each state
state_wide_proportional <- presidential_vote_count |>
  select(year, state, candidate, candidatevotes, totalvotes) |>
  mutate(percentage_state_votes = (candidatevotes / totalvotes)) |>
  select(-candidatevotes, -totalvotes)

# Find the number of electoral votes received by each candidate
state_wide_proportional <- state_wide_proportional |>
  left_join(electoral_votes_by_state,
    by = c("year", "state")
  ) |>
  mutate(votes_received = round(percentage_state_votes * electoral_votes, digits = 0)) |>
  select(-percentage_state_votes, -electoral_votes)

# Sum total votes and find presidential winner
state_wide_proportional <- state_wide_proportional |>
  group_by(year, candidate) |>
  summarize(total_electoral_votes = sum(votes_received)) |>
  slice_max(order_by = total_electoral_votes, n = 1, with_ties = FALSE) |>
  rename(winner = candidate)

# Create knitr table
knitr::kable(state_wide_proportional, col.names = c("Year", "Winning Candidate", "Electoral Votes"),
             caption = "State-Wide Proportional: Presidential Winning Candidate", format = "html")
State-Wide Proportional: Presidential Winning Candidate
Year Winning Candidate Electoral Votes
1976 CARTER, JIMMY 267
1980 REAGAN, RONALD 268
1984 REAGAN, RONALD 317
1988 BUSH, GEORGE H.W. 288
1992 CLINTON, BILL 229
1996 CLINTON, BILL 263
2000 BUSH, GEORGE W. 259
2004 BUSH, GEORGE W. 276
2008 OBAMA, BARACK H. 282
2012 OBAMA, BARACK H. 272
2016 CLINTON, HILLARY 258
2020 BIDEN, JOSEPH R. JR 272

The State-Wide Proportional table shows that electoral votes were distributed based on each candidate’s share of the popular vote in each state. This method leads to smaller, more proportional victories compared to winner-take-all systems. For example, Clinton and Reagan won with varying numbers of electoral votes due to differing vote shares. More recent elections, like 2016 and 2020, show closer contests and lower electoral vote totals, reflecting the nuanced nature of proportional allocation.

Finally, we look at National Proportional scheme effects on presidential elections.

Code 
# Find total number of electoral votes available
electoral_votes_available <- electoral_votes_by_state |>
  group_by(year) |>
  summarize(electoral_college_votes = sum(electoral_votes))

# Find percentage of popular vote each candidate received
national_proportional <- presidential_vote_count |>
  select(year, state, candidate, candidatevotes) |>
  group_by(year, candidate) |>
  summarize(total_electoral_votes = sum(candidatevotes)) |>
  group_by(year) |>
  mutate(population_vote_count = sum(total_electoral_votes)) |> # Find total number of votes cast in election year
  ungroup() |>
  mutate(percentage_population_vote = (total_electoral_votes / population_vote_count)) |>
  select(-total_electoral_votes, -population_vote_count) |>
  # Find the proportion of the electoral votes received based on the popular vote percentage
  left_join(
    electoral_votes_available,
    join_by(year == year)
  ) |>
  mutate(electoral_votes_received = round(percentage_population_vote * electoral_college_votes, digits = 0)) |>
  select(-percentage_population_vote, -electoral_college_votes) |>
  group_by(year) |>
  slice_max(order_by = electoral_votes_received, n = 1, with_ties = FALSE) |>
  rename(winner = candidate)

# Create knitr table
knitr::kable(
  setNames(national_proportional, c("Year", "Winning Candidate", "Electoral Votes")),
  caption = "National Proportional: Presidential Winning Candidate"
)
National Proportional: Presidential Winning Candidate
Year Winning Candidate Electoral Votes
1976 CARTER, JIMMY 269
1980 REAGAN, RONALD 273
1984 REAGAN, RONALD 316
1988 BUSH, GEORGE H.W. 287
1992 CLINTON, BILL 231
1996 CLINTON, BILL 265
2000 GORE, AL 260
2004 BUSH, GEORGE W. 273
2008 OBAMA, BARACK H. 285
2012 OBAMA, BARACK H. 275
2016 CLINTON, HILLARY 259
2020 BIDEN, JOSEPH R. JR 276

From the “National Proportional” table, we can conclude that in each election year, the winning candidate is awarded electoral votes proportional to the percentage of the national popular vote they received. However, the key observation here is that the actual winner of the election in terms of total electoral votes does not always align with the real-world outcome. For example:

In 2000, Al Gore won the popular vote but received fewer electoral votes (260) than George W. Bush, who had 273 electoral votes in the real outcome. In 2016, Hillary Clinton received more popular votes but had fewer electoral votes (259) than Donald Trump (who had 304 in the actual result). This demonstrates the difference between the proportional allocation of electoral votes based on popular vote percentages and the winner-take-all systems used in most U.S. states. The table also shows a tendency for the winner to generally reflect the national sentiment, but there are notable exceptions where the electoral votes may not fully mirror the popular vote distribution.

Analysis of Electoral College Vote (ECV) Allocation Schemes

To determine the “fairest” Electoral College Vote (ECV) allocation scheme, we consider how well each method aligns with the principle of one person, one vote, ensuring each citizen’s vote has roughly equal weight in determining election outcomes.

Fairness of ECV Allocation Schemes

State-Wide Winner-Take-All

  • Description: The candidate with the most votes in a state receives all its electoral votes.
  • Fairness Impact: This system amplifies wins in states where a candidate narrowly leads, disregarding the votes of the losing candidate’s supporters. It distorts representation and often results in a “winner-takes-all” effect.

District-Wide Winner-Take-All

  • Description: Electoral votes are allocated by congressional district, with the state’s two “at-large” votes going to the statewide winner.
  • Fairness Impact: Offers more localized representation and is slightly more proportional than the state-wide system. However, it still disproportionately favors the majority party in a state.

State-Wide Proportional

  • Description: Electoral votes are distributed based on the percentage of the popular vote each candidate receives in a state.
  • Fairness Impact: This is the most proportional system, closely aligning electoral votes with voter preferences. However, due to the structure of the Electoral College, disparities may persist.

National Proportional

  • Description: Electoral votes are allocated based on each candidate’s share of the national popular vote.
  • Fairness Impact: Eliminates state-level disparities but retains the limitations of the Electoral College, where the distribution of votes across states can affect outcomes.

Schemes That Produce Different Results

State-Wide Winner-Take-All and Proportional Allocation often lead to different outcomes, particularly in tight races. For example, in the 2000 election, George W. Bush secured 271 electoral votes and won the presidency despite losing the national popular vote to Al Gore by over 500,000 votes. A proportional system would have redistributed electoral votes more equitably, potentially resulting in a Gore victory.

Largest Impact of ECV Scheme: 2000 Election

Under the State-Wide Winner-Take-All System

  • Outcome: Bush won Florida’s electoral votes by a narrow margin, securing 271 total votes to win the presidency despite losing the national popular vote.

Under the Proportional Allocation System

  • Outcome: Electoral votes would have been distributed more in line with the popular vote. Bush would have received fewer electoral votes, and Gore would likely have gained enough to win the presidency, highlighting the distortion caused by the winner-take-all system.

Conclusion

The State-Wide Proportional Allocation System is the fairest as it aligns electoral votes more closely with voter preferences. The 2000 election underscores how the winner-take-all method can distort outcomes, particularly in close elections. A proportional system would better reflect the popular vote, likely changing the results in contested cases.

Citations

1976-2022-house.csv dataset: MIT Election Data and Science Lab, 2017, “U.S. House 1976–2022”, https://doi.org/10.7910/DVN/IG0UN2, Harvard Dataverse, V13, UNF:6:Ky5FkettbvohjTSN/IVldA== [fileUNF]

1976-2020-president.csv dataset: MIT Election Data and Science Lab, 2017, “U.S. President 1976–2020”, https://doi.org/10.7910/DVN/42MVDX, Harvard Dataverse, V8, UNF:6:F0opd1IRbeYI9QyVfzglUw== [fileUNF]

United States Congressional District Shapefiles datasets: Jeffrey B. Lewis, Brandon DeVine, Lincoln Pitcher, and Kenneth C. Martis. (2013) Digital Boundary Definitions of United States Congressional Districts, 1789-2012. [Data file and code book]. Retrieved from https://cdmaps.polisci.ucla.edu on [date of download].