Connecticut elections

What it covers

The Connecticut scraper returns election results from the CT Election Management System (CTEMS) at ctemspublic.tgstg.net. It covers all elections available in the CTEMS dropdown — general elections, primaries, runoffs, and special elections — from 2016 to the present.

Results include an election_level column classifying each race as "Federal", "State", or "Local", and a contest_outcome column ("Won" / "Lost") on the statewide totals.

Note: The CTEMS site is a JavaScript-rendered AngularJS SPA. All scraping uses a headless Chromium browser via Playwright. Scraping all 169 towns across all elections in a year typically takes 30–60 minutes per election year. Use level = "state" for a much faster statewide-only result.

Arguments

Argument	Default	Description
`state`	—	`"CT"`, `"connecticut"`, or `"Connecticut"`
`year_from`	`NULL`	Start year, inclusive; `NULL` = no lower bound
`year_to`	`NULL`	End year, inclusive; `NULL` = no upper bound
`level`	`"all"`	`"all"` — statewide + town; `"state"` — statewide only; `"town"` — town-level only
`max_workers`	`2L`	Parallel Chromium browsers for town scraping (one per county)

Examples

Statewide + town results for a single year

res <- scrape_elections(state = "CT", year_from = 2024, year_to = 2024)

# Statewide totals — one row per candidate per office
res$state %>%
  filter(election_level == "Federal") %>%
  select(election_name, office, candidate, party, votes, vote_pct, contest_outcome)

# Town-level results
res$town %>%
  filter(office == "President of the United States") %>%
  select(town, county, candidate, party, votes)

Statewide totals only (faster — skips town scraping)

state_df <- scrape_elections(
  state     = "CT",
  year_from = 2024,
  year_to   = 2024,
  level     = "state"
)

state_df %>%
  filter(contest_outcome == "Won") %>%
  select(election_name, election_level, office, candidate, party, votes, vote_pct)

Town-level only

town_df <- scrape_elections(
  state     = "CT",
  year_from = 2022,
  year_to   = 2022,
  level     = "town"
)

# Summarise votes by county for a specific race
town_df %>%
  filter(office == "United States Senator") %>%
  group_by(county, candidate, party) %>%
  summarise(votes = sum(votes, na.rm = TRUE), .groups = "drop") %>%
  arrange(county, desc(votes))

Multi-year scrape with more parallel workers

res <- scrape_elections(
  state             = "CT",
  year_from         = 2020,
  year_to           = 2024,
  level             = "all",
  max_workers = 4L
)

Filter by election level

res <- scrape_elections(state = "CT", year_from = 2023, year_to = 2023)

# Local races only (municipal elections)
res$state %>%
  filter(election_level == "Local") %>%
  select(election_name, office, candidate, party, votes, contest_outcome)

Output columns

`$state` (when `level = "all"` or `"state"`)

Column	Description
`election_name`	Human-readable election name (e.g. `"11/05/2024 -- November General Election"`)
`election_year`	Calendar year
`election_date`	ISO date string when parseable; `NA` otherwise
`election_level`	`"Federal"`, `"State"`, or `"Local"`
`office`	Office name (includes district when present)
`candidate`	Candidate name
`party`	Party name (e.g. `"Democratic"`, `"Republican"`)
`votes`	Total votes (integer)
`vote_pct`	Percentage of votes in contest (recomputed from town totals for State/Local)
`winner`	`TRUE` if the candidate won the statewide contest

`$town` (when `level = "all"` or `"town"`)

Column	Description
`election_name`	Human-readable election name
`election_year`	Calendar year
`election_date`	Date of the election
`county`	County name
`town`	Town name
`office_level`	`"Federal"`, `"State"`, or `"Local"`
`office`	Office name
`candidate`	Candidate name
`party`	Party name
`votes`	Votes in this town (integer)
`vote_pct`	Percentage as reported by CTEMS for this town
`town_winner`	`TRUE` if the candidate won within this town

Performance notes

level = "state" is much faster. Town scraping opens one Chromium browser per county and iterates all 169 towns. A single election year with full town scraping takes roughly 30–60 minutes. Use level = "state" when you only need statewide totals.
Multiple elections per year. CTEMS often has several elections in a single year (primary, general, special elections). All are scraped and combined in the output.
max_workers. CT has 8 counties. The default of 2 parallel browsers means 4 batches. Increasing to 4 roughly halves town-scraping time. Keep at or below 4 to avoid memory pressure.