House of Representatives - Preferences
Preferences are a key aspect of Australia’s preferential voting system. auspol includes two functions to interact with the preference distribution data:
- get_house_preferences()
- house_preference_flow_data()
- house_preference_flow_plot()
- get_house_2PF()
- get_house_2PP()
What is this?
If you are unfamiliar with the Australian electoral system and preferential voting, please look at this explainer before proceeding.
Getting the data
get_house_preferences() - provides an interface to access preference flows for a particular division (electorate) on a given year, as published on the AEC’s website. For instance:
get_house_preferences("Wills",2019)
#> # A tibble: 6,120 × 16
#> Year StateAb DivisionId DivisionNm PPId PPNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Sitti…¹ Calcu…² Calcu…³
#> <dbl> <chr> <int> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <chr> <chr> <dbl>
#> 1 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 254
#> 2 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 4.65
#> 3 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 4 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 5 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 238
#> 6 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 4.35
#> 7 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 8 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 9 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 862
#> 10 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 15.8
#> # … with 6,110 more rows, and abbreviated variable names ¹SittingMemberFl, ²CalculationType, ³CalculationValueResults can be also retrieved disaggregated by polling place:
get_house_preferences("Wills",2019,aggregation = FALSE) |> head(10)
#> # A tibble: 10 × 16
#> Year StateAb DivisionId DivisionNm PPId PPNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Sitti…¹ Calcu…² Calcu…³
#> <dbl> <chr> <int> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <chr> <chr> <dbl>
#> 1 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 254
#> 2 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 4.65
#> 3 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 4 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 5 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 238
#> 6 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 4.35
#> 7 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 8 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 9 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 862
#> 10 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 15.8
#> # … with abbreviated variable names ¹SittingMemberFl, ²CalculationType, ³CalculationValueIt is also possible to filter by Polling Place Name (PPNm in the source):
get_house_preferences("Wills",2019, polling_places=c("ABSENT"),aggregation = FALSE) |> head(10)
#> # A tibble: 10 × 16
#> Year StateAb DivisionId DivisionNm PPId PPNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Sitti…¹ Calcu…² Calcu…³
#> <dbl> <chr> <int> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <chr> <chr> <dbl>
#> 1 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 254
#> 2 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Prefer… 4.65
#> 3 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 4 2019 VIC 234 Wills 0 ABSENT 0 1 32688 MILES Christopher James AJP Animal Justice Party N Transf… 0
#> 5 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 238
#> 6 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Prefer… 4.35
#> 7 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 8 2019 VIC 234 Wills 0 ABSENT 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists N Transf… 0
#> 9 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 862
#> 10 2019 VIC 234 Wills 0 ABSENT 0 3 33251 KILLIN Peter John LP Liberal Party of Austr… N Prefer… 15.8
#> # … with abbreviated variable names ¹SittingMemberFl, ²CalculationType, ³CalculationValueBuilt on top of the basic functions, preference_flow_data() - provides an interface to access preference flows for a particular division (electorate) in a given year. This function returns a list, in which each element is a data frame with the results for each counting round. In its most basic use:
house_preference_flow_data(division = "Wills", year=2019)
#> $`Round 0`
#> # A tibble: 6 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Last
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2019 VIC 234 Wills 0 6 32350 KHALIL Peter ALP Australian Labor Party (Victor… 42355 0 1 FALSE
#> 2 2019 VIC 234 Wills 0 4 32657 PULFORD Adam Ross GVIC The Australian Greens - Victor… 25575 0 2 FALSE
#> 3 2019 VIC 234 Wills 0 3 33251 KILLIN Peter John LP Liberal Party of Australia (Vi… 17241 0 3 FALSE
#> 4 2019 VIC 234 Wills 0 2 32398 BOLTON Susanne Alison VNS Victorian Socialists 4344 0 4 FALSE
#> 5 2019 VIC 234 Wills 0 1 32688 MILES Christopher James AJP Animal Justice Party 3596 0 5 FALSE
#> 6 2019 VIC 234 Wills 0 5 32329 VENKAT Manju UAPP United Australia Party 2979 0 6 TRUE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPosition
#>
#> $`Round 1`
#> # A tibble: 5 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Last
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2019 VIC 234 Wills 1 6 32350 KHALIL Peter ALP Australian Labor Party (Victor… 43301 946 1 FALSE
#> 2 2019 VIC 234 Wills 1 4 32657 PULFORD Adam Ross GVIC The Australian Greens - Victor… 25811 236 2 FALSE
#> 3 2019 VIC 234 Wills 1 3 33251 KILLIN Peter John LP Liberal Party of Australia (Vi… 18521 1280 3 FALSE
#> 4 2019 VIC 234 Wills 1 2 32398 BOLTON Susanne Alison VNS Victorian Socialists 4511 167 4 FALSE
#> 5 2019 VIC 234 Wills 1 1 32688 MILES Christopher James AJP Animal Justice Party 3946 350 5 TRUE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPosition
#>
#> $`Round 2`
#> # A tibble: 4 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Last
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2019 VIC 234 Wills 2 6 32350 KHALIL Peter ALP Australian Labor Party (Victorian… 43982 681 1 FALSE
#> 2 2019 VIC 234 Wills 2 4 32657 PULFORD Adam Ross GVIC The Australian Greens - Victoria 27083 1272 2 FALSE
#> 3 2019 VIC 234 Wills 2 3 33251 KILLIN Peter John LP Liberal Party of Australia (Victo… 19108 587 3 FALSE
#> 4 2019 VIC 234 Wills 2 2 32398 BOLTON Susanne Alison VNS Victorian Socialists 5917 1406 4 TRUE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPosition
#>
#> $`Round 3`
#> # A tibble: 3 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Last
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2019 VIC 234 Wills 3 6 32350 KHALIL Peter ALP Australian Labor Party (Victorian Bra… 45442 1460 1 FALSE
#> 2 2019 VIC 234 Wills 3 4 32657 PULFORD Adam Ross GVIC The Australian Greens - Victoria 30254 3171 2 FALSE
#> 3 2019 VIC 234 Wills 3 3 33251 KILLIN Peter John LP Liberal Party of Australia (Victorian… 20394 1286 3 TRUE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPosition
#>
#> $`Round 4`
#> # A tibble: 2 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Elected
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2019 VIC 234 Wills 4 6 32350 KHALIL Peter ALP Australian Labor Party (Victorian Br… 55898 10456 1 TRUE
#> 2 2019 VIC 234 Wills 4 4 32657 PULFORD Adam Ross GVIC The Australian Greens - Victoria 40192 9938 2 FALSE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPositionThe function also includes other variables, which are better explained through its use as part of the next function, house_preference_flow_plot().
Plotting Preference flows.
Perhaps it is most interesting to see how each round of counting assigns votes until a candidate is chosen. This package comes with the convenience function house_preference_flow_plot(), which calls preference_flow_data() and plots the results in an alluvial format. Using the previous example:
house_preference_flow_plot(division = "Wills",year=2019) +
labs(title="Preference Flow for Wills, 2019")
By default the vote is present as a percentage, which can be changed by changing the value of var:
house_preference_flow_plot(division = "Wills",year=2019,var="Preference Count") +
labs(title="Preference Flow for Wills, 2019")
For analysis, it may be interesting to remove some parties from the initial round - e.g. to see how votes flow to the finalists. This can be done using exclude_parties:
house_preference_flow_plot(division = "Warringah",year=2022) +
labs(title="Preference Flow for Warringah, 2022")
house_preference_flow_plot(division = "Warringah",year=2022, exclude_parties = c("LP","IND-STEGGALL")) +
labs(title="Flow of preference votes to Liberal and Teal in Warringah, 2022")
Like the other plotting functions, it is possible to override the colour scheme by providing an additional named vector:
#adding teal
house_preference_flow_plot(division = "Warringah",year=2022,
exclude_parties = c("LP","IND-STEGGALL"),
extra_colours = c("IND-STEGGALL"="#008080")) +
labs(title="Flow of preference votes to Liberal and Teal in Warringah, 2022")
Similarly, it is possible to get the results of house_preference_flow_plot() directly, via include_data:
p <- house_preference_flow_plot(division = "Warringah",year=2022,
exclude_parties = c("LP","IND-STEGGALL"),
extra_colours = c("IND-STEGGALL"="#008080"),
include_data=TRUE
)
p$source_data[[1]]
#> # A tibble: 5 × 15
#> Year StateAb DivisionId DivisionNm CountNum BallotPosition CandidateId Surname GivenNm PartyAb PartyNm Prefe…¹ Trans…² Round…³ Last
#> <dbl> <chr> <int> <chr> <int> <int> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <int> <lgl>
#> 1 2022 NSW 151 Warringah 0 6 36808 MICKLEBURGH David Edward ALP Australian Labor Party (N.S.W.… 7806 0 1 FALSE
#> 2 2022 NSW 151 Warringah 0 4 36675 GLANVILLE Kristyn Maree GRN The Greens NSW 6910 0 2 FALSE
#> 3 2022 NSW 151 Warringah 0 3 36271 ROBERTSON Andrew Colin UAPP United Australia Party 2202 0 3 FALSE
#> 4 2022 NSW 151 Warringah 0 1 37516 TRIPP Steven Leslie ON Pauline Hanson's One Nation 1980 0 4 FALSE
#> 5 2022 NSW 151 Warringah 0 7 36396 PATERSON Kate Anne AJP Animal Justice Party 1475 0 5 TRUE
#> # … with abbreviated variable names ¹`Preference Count`, ²`Transfer Count`, ³RoundPositionFlow from primary vote to last round
Although the last plot can be informative, The AEC also records how o understand how the votes have flowed from primary votes to the two last candidates. This can be accessed through get_house_2PF()
get_house_2PF(division = "Lingiari",year=2013,aggregation = TRUE)
#> # A tibble: 16 × 17
#> Year StateAb DivisionId DivisionNm FromCandidateId FromCandidatePartyAb FromCandida…¹ FromC…² FromC…³ FromC…⁴ ToCan…⁵ ToCan…⁶ ToCan…⁷ ToCan…⁸ ToCan…⁹ ToCan…˟ Trans…˟
#> <dbl> <chr> <int> <chr> <int> <chr> <chr> <chr> <chr> <int> <int> <chr> <chr> <chr> <chr> <int> <int>
#> 1 2013 NT 306 Lingiari 24971 IND Independent GOULD Alfred 3 28735 ALP Austra… SNOWDON Warren 2 257
#> 2 2013 NT 306 Lingiari 24971 IND Independent GOULD Alfred 3 28827 CLP Countr… MacFAR… Tina 9 170
#> 3 2013 NT 306 Lingiari 28006 CEC Citizens Ele… FLYNN Peter 7 28735 ALP Austra… SNOWDON Warren 2 138
#> 4 2013 NT 306 Lingiari 28006 CEC Citizens Ele… FLYNN Peter 7 28827 CLP Countr… MacFAR… Tina 9 123
#> 5 2013 NT 306 Lingiari 28016 IND Independent EARLEY Braedon 5 28735 ALP Austra… SNOWDON Warren 2 986
#> 6 2013 NT 306 Lingiari 28016 IND Independent EARLEY Braedon 5 28827 CLP Countr… MacFAR… Tina 9 822
#> 7 2013 NT 306 Lingiari 28555 GRN Australian G… HOAD Rob 8 28735 ALP Austra… SNOWDON Warren 2 2716
#> 8 2013 NT 306 Lingiari 28555 GRN Australian G… HOAD Rob 8 28827 CLP Countr… MacFAR… Tina 9 589
#> 9 2013 NT 306 Lingiari 28561 RUA Rise Up Aust… McCART… Regina 6 28735 ALP Austra… SNOWDON Warren 2 756
#> 10 2013 NT 306 Lingiari 28561 RUA Rise Up Aust… McCART… Regina 6 28827 CLP Countr… MacFAR… Tina 9 742
#> 11 2013 NT 306 Lingiari 28735 ALP Australian L… SNOWDON Warren 2 28735 ALP Austra… SNOWDON Warren 2 17056
#> 12 2013 NT 306 Lingiari 28827 CLP Country Libe… MacFAR… Tina 9 28827 CLP Countr… MacFAR… Tina 9 13605
#> 13 2013 NT 306 Lingiari 28849 ASP Shooters, Fi… RIGHTON Chris 1 28735 ALP Austra… SNOWDON Warren 2 1864
#> 14 2013 NT 306 Lingiari 28849 ASP Shooters, Fi… RIGHTON Chris 1 28827 CLP Countr… MacFAR… Tina 9 1197
#> 15 2013 NT 306 Lingiari 29433 IND Independent GUYULA Yingiy… 4 28735 ALP Austra… SNOWDON Warren 2 1275
#> 16 2013 NT 306 Lingiari 29433 IND Independent GUYULA Yingiy… 4 28827 CLP Countr… MacFAR… Tina 9 579
#> # … with abbreviated variable names ¹FromCandidatePartyNm, ²FromCandidateSurname, ³FromCandidateGivenNm, ⁴FromCandidateBallotPosition, ⁵ToCandidateId,
#> # ⁶ToCandidatePartyAb, ⁷ToCandidatePartyNm, ⁸ToCandidateSurname, ⁹ToCandidateGivenNm, ˟ToCandidateBallotPosition, ˟TransferCountThese results can be charted using house_2PF_plot(). This function can present the results either in an alluvial or bar chart format.
Two-party preferred
The AEC also publishes Two-Party-Preferred summaries (Coalition against Labor). These are available through * get_house_2PP()*, which allows filtering by division and year, and can be displayed by division or polling place
get_house_2PP(division = "Fraser",year=2004,aggregation = TRUE)
#> # A tibble: 1 × 9
#> Year StateAb DivisionID DivisionNm `Liberal/National Coalition Votes` `Australian Labor Party Votes` TotalVotes `Liberal/National Coalition Percentage` Australian L…¹
#> <dbl> <chr> <int> <chr> <int> <int> <int> <dbl> <dbl>
#> 1 2004 ACT 102 Fraser 30546 53134 83680 0.365 0.635
#> # … with abbreviated variable name ¹`Australian Labor Party Percentage`The package includes two plots to visualise this data. The first (house_2PP_comparison_plot()), displays the results for a given election, for one or many electorates.
The second function - house_2PP_historical_plot() - shows the historical results for a given division.
