Summary

This article explores the summary and preMMP_results datasets.

The summary dataset consolidates key electoral information, providing a bird’s-eye view of New Zealand’s elections. It can be used to find the number of seats, votes (n & %), and nominations/on party list by ballot type - Candidate Vote, Party Vote, or Total (Seats column = Candidate + Party; Vote column = Party only). It can also be used to find the number of overhang seats and which party these can be attributed to.

Loading the Data

Begin by loading the summary dataset to understand the structure:

# Load datasets by using the following helper function:
df <- scgUtils::get_data("summary")
# Alternatively, use: data("summary"), then df <- summary

# View the data
head(df)

Election	Party	Ballot	Seats	Votes	Percentage	Nominated	Registered	Successful
2023	ACT Party	Candidate	2	149507	5.45	59	Yes	Yes
2023	Animal Justice Party	Candidate	0	5829	0.21	17	Yes	No
2023	Aotearoa Legalise Cannabis Party	Candidate	0	12566	0.46	14	Yes	No
2023	DemocracyNZ	Candidate	0	12060	0.44	13	Yes	No
2023	Freedoms NZ	Candidate	0	0	0.00	0	Yes	No
2023	NZ Outdoors & Freedom Party	Candidate	0	3030	0.11	3	Yes	No

Party and Candidate Votes

To view the top performing parties, rank the Party and Candidate Vote by Party for the 2023 election by removing the Total ballot type from the dataset:

df %>%
  filter(Ballot != "Total", Election == 2023) %>%
  arrange(-Percentage) %>%
  head(n = 10)

Election	Party	Ballot	Seats	Votes	Percentage	Nominated	Registered	Successful
2023	National Party	Candidate	43	1192251	43.47	67	Yes	Yes
2023	National Party	Party	5	1085851	38.08	74	Yes	Yes
2023	Labour Party	Candidate	17	855963	31.21	72	Yes	Yes
2023	Labour Party	Party	17	767540	26.92	76	Yes	Yes
2023	Green Party	Party	12	330907	11.61	49	Yes	Yes
2023	ACT Party	Party	9	246473	8.64	60	Yes	Yes
2023	Green Party	Candidate	3	226575	8.26	52	Yes	Yes
2023	NZ First	Party	8	173553	6.09	35	Yes	Yes
2023	ACT Party	Candidate	2	149507	5.45	59	Yes	Yes
2023	Maori Party	Candidate	6	106584	3.89	17	Yes	Yes

Overhang Seats

As a reminder, overhang seats occur when the Party Vote entitles that party to fewer seats than the number of electorate seats that it won. To view any overhang seats, remove the Candidate and Party ballot types and unsuccessful parties from the dataset. Then, calculate the overhang seats by multiplying the 120 total parliamentary seats by the proportion of the party vote for successful parties:

df %>%
  filter(Successful == "Yes", Ballot == "Total", Election == 2023) %>%
  group_by(Election) %>%
  mutate(`Successful_%` = Votes / sum(Votes) * 100) %>%
  ungroup() %>%
  mutate(Overhang = Seats - round(120 * `Successful_%` / 100, 0)) %>%
  arrange(-Overhang)

Election	Party	Ballot	Seats	Votes	Percentage	Nominated	Registered	Successful	Successful_%	Overhang
2023	Maori Party	Total	6	87844	3.08	48	Yes	Yes	3.26	2
2023	ACT Party	Total	11	246473	8.64	119	Yes	Yes	9.16	0
2023	Green Party	Total	15	330907	11.61	101	Yes	Yes	12.29	0
2023	Labour Party	Total	34	767540	26.92	148	Yes	Yes	28.51	0
2023	NZ First	Total	8	173553	6.09	69	Yes	Yes	6.45	0
2023	National Party	Total	48	1085851	38.08	141	Yes	Yes	40.33	0

Create Plot
To view overhang seats over time, visualise the number of seats won by party at each election:

df %>%
  # remove parties which did not win a seat
  filter(Successful == "Yes", Ballot == "Total") %>%
  ggplot(aes(x = Election, y = Seats,
             fill = factor(Party, levels = c("National Party", "ACT Party", "NZ First",
                                             "United Future", "Jim Anderton's Progressive",
                                             "MANA", "Maori Party", "Alliance", "Green Party",
                                             "Labour Party")))) +
  geom_bar(stat = "identity", colour = "white", linewidth = 0.15, alpha = 0.8) +
  # indicate majority required for a 120 seat parliament
  geom_hline(yintercept = 60, colour = "white", linewidth = 0.5) +
  # indicate overhang seats
  geom_hline(yintercept = 120, colour = scgUtils::colour_pal("French Grey")) +
  annotate("text", x = 2025, y = 120, label = "Overhang",
           colour = scgUtils::colour_pal("Regent Grey"),
           size = 3.5, fontface = 2) +
  coord_flip(clip = "off", xlim = c(1994.5, 2024.5)) +
  labs(title = "No. of Seats by Election and Party (1996 - 2023)",
       fill = "Party") +
  scale_fill_manual(values = scgUtils::colour_pal("polNZ")) +
  scale_y_continuous(expand = c(0, 0)) +
  scale_x_continuous(expand = c(0, 0), "Election",
                     labels = as.character(df$Election), breaks = df$Election) +
  scgUtils::theme_scg() +
  theme(panel.grid.major.y = element_blank(),
        panel.grid.minor.y = element_blank())

preMMP Results

The summary dataset can be combined with the preMMP_results dataset. The preMMP_results dataset contains the overall results between 1890 and 1993 by party at the national-level.

Loading the Data

View the shape of data frame.

df1 <- scgUtils::get_data("preMMP_results")
head(df1)

Election	Party	Seats	Percentage
1890	Liberal	38	56.10
1890	Conservative Party	25	28.90
1890	Independents	11	15.00
1893	Liberal	51	57.80
1893	Conservative Party	13	24.49
1893	Independents	10	17.71

Merge datasets

To match the preMMP_results dataset, remove the summary dataset Party and Candidate ballot types.

df <- df %>%
  filter(Ballot == "Total") %>%
  mutate(Party = ifelse(Successful == "Yes", Party, "Other")) %>%
  group_by(Election, Party) %>%
  summarise(Seats = sum(Seats), Votes = sum(Votes), .groups = 'drop') %>%
  ungroup() %>%
  group_by(Election) %>%
  mutate(Percentage = round(Votes / sum(Votes) * 100, 2)) %>%
  ungroup() %>%
  select(Election, Party, Seats, Percentage) %>%
  arrange(-Election)

# Merge
df2 <- rbind(df, df1)

# view dataset
head(df2)

Election	Party	Seats	Percentage
2023	ACT Party	11	8.64
2023	Green Party	15	11.61
2023	Labour Party	34	26.92
2023	Maori Party	6	3.08
2023	NZ First	8	6.09
2023	National Party	48	38.08

Visualise the Historic Party Vote

Visualise the party/popular vote at each election:

df2 <- df2 %>%
  # filter 1935 until current time (since the beginning of the National and Labour parties)
  filter(Election >= 1935) %>%
  mutate(Party = ifelse(Party %in% c("National Party", "Labour Party"), Party,
                        ifelse(Party == "United-Reform ('National')", "National Party",
                               "Other")),
         Party = factor(Party, levels = c("National Party", "Other", "Labour Party"))) %>%
  group_by(Election, Party) %>%
  summarise(Percentage = round(sum(Percentage), 2), .groups = 'drop') %>%
  ungroup()

df2 %>%
  ggplot(aes(x = reorder(Election, -Election), y = Percentage,
             fill = Party)) +
  geom_bar(stat = "identity", alpha = 0.8, width = 1, size = 0) +
  geom_bar(data = df2 %>% filter(Election == 2023), aes(x = 1, y = Percentage, fill = Party),
           stat = "identity", alpha = 1, width = 1) +
  # Add 50% line
  geom_hline(yintercept = 50, colour = "white") +
  annotate("text", x = 31.2, y = 50, label = "50%", size = 3.5, fontface = 2,
           colour = scgUtils::colour_pal("Regent Grey")) +
  annotate("text", x = 31.2, y = 0, label = "Party Vote >", size = 3, fontface = 2,
           colour = scgUtils::colour_pal("Regent Grey"), hjust = 0) +
  annotate("text", x = 31.2, y = 100, label = "< Party Vote", size = 3, fontface = 2,
           colour = scgUtils::colour_pal("Regent Grey"), hjust = 1) +
  # Add dashed lines for 2023 result
  geom_hline(yintercept = df2$Percentage[df2$Election == 2023 & df2$Party == "Labour Party"],
             colour = "white", linewidth = 0.5, linetype = "dashed") +
  geom_hline(yintercept = 100 - df2$Percentage[df2$Election == 2023 & df2$Party == "National Party"],
             colour = "white", linewidth = 0.5, linetype = "dashed") +
  # Add introduction of MMP line
  geom_vline(xintercept = 10.5, colour = scgUtils::colour_pal("Black80"),
             linetype = "dashed", linewidth = 0.25) +
  annotate("text", x = 10.5, y = 100.75, label = "MMP introduced", size = 3.5, fontface = 2,
           colour = scgUtils::colour_pal("Black80"), hjust = 0) +
  # Add text to 2023 result
  geom_text(x = 1, y = 1,
            label = paste0(df2$Percentage[df2$Election == 2023 & df2$Party == "Labour Party"], "%"),
            hjust = 0, size = 3, colour = "white") +
  annotate("text", x = 0, y = 0, label = "Labour", size = 3.5, fontface = 2,
           colour = "#D82A20", hjust = 0) +
  geom_text(x = 1, y = df2$Percentage[df2$Election == 2023 & df2$Party == "Labour Party"] +
    (df2$Percentage[df2$Election == 2023 & df2$Party == "Other"] / 2),
            label = paste0(format(df2$Percentage[df2$Election == 2023 & df2$Party == "Other"],
                                  nsmall = 2), "%"),
            hjust = 0.5, size = 3, colour = scgUtils::colour_pal("Black80")) +
  annotate("text", x = 0, y = df2$Percentage[df2$Election == 2023 & df2$Party == "Labour Party"] +
    (df2$Percentage[df2$Election == 2023 & df2$Party == "Other"] / 2),
           label = "Minor Parties", size = 3.5, fontface = 2,
           colour = scgUtils::colour_pal("Regent Grey"), hjust = 0.5) +
  geom_text(x = 1, y = 99,
            label = paste0(df2$Percentage[df2$Election == 2023 & df2$Party == "National Party"], "%"),
            hjust = 1, size = 3, colour = "white") +
  annotate("text", x = 0, y = 100,
           label = "National", size = 3.5, fontface = 2,
           colour = "#00529F", hjust = 1) +
  labs(title = "% Party Vote by Election between 1935 and 2023") +
  scale_y_continuous(expand = c(0, 0), position = "right",
                     breaks = seq(0, 100, by = 5)) +
  scale_fill_manual(values = scgUtils::colour_pal("polNZ")) +
  coord_flip(clip = "off", xlim = c(1, 30), ylim = c(0, 100.25)) +
  scgUtils::theme_scg() +
  theme(legend.position = "none",
        plot.margin = unit(c(1, 3, 1, 1), "cm"),
        plot.title = element_text(vjust = 5, colour = scgUtils::colour_pal("Black80")),
        panel.grid.major = element_blank(),
        panel.grid.minor = element_blank(),
        axis.line = element_blank(),
        axis.ticks.y = element_blank(),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        axis.text.x = element_blank()
  )
#> Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
#> ℹ Please use `linewidth` instead.
#> This warning is displayed once every 8 hours.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.