nc-campaign-finance/reports/2023-11-14_repub-donors/republican-donors-2020-2022.qmd

---
title: "Republican Donors, 2020-2022"
author: Garrick Aden-Buie
format: html

execute:
  echo: true
---

## Task

> I’m working on some donor data and I’m wondering if we may have it already in
> what we’re doing… I’m trying to pull all disclosed donors ($50+) that gave to
> Republican campaigns in ’20 and ’22. Even if we couldn’t specify party of
> campaign but I could sort it, that would even be helpful. Is that something we
> have already?

## Setup 

```{r setup}
library(tidyverse)
library(fs)
library(zoomerjoin)
pkgload::load_all(here::here("process"))

report_dates <- prep_open_dataset_db("report_dates")
officers <- prep_open_dataset_db("officers")
receipts <- prep_open_dataset_db("receipts")
candidate_listing <- prep_open_dataset_db("candidate_listing")
cover <- prep_open_dataset_db("cover")

report_list_local <- targets::tar_read(report_list, store = here::here("process", "_targets"))
copy_to(duckdb_global_con(), report_list_local, "report_list")
report_list <- tbl(duckdb_global_con(), "report_list")

con <- DBI::dbConnect(RSQLite::SQLite(), here::here("data-prep", "address_lookup.sqlite"))
address_db <- tbl(con, "resolved")
```

## Reports

1. Reports from 2020 and 2022
1. Where an individual donated more than $50
1. Receipt types that match individual donors and not clerical records

```{r receipt-types}
receipts |> 
  count(receipt_type_desc, receipt_type_code, sort = TRUE) |>
  print(n = 20)
```

```{r reports-first-pass}
report_list_2020_2022 <- 
  report_list |>
  filter(year %in% c(2020, 2022))

receipts_2020_2022_src <-
  receipts |>
  semi_join(report_list_2020_2022, by = "report_id") |>
  mutate(donor_name = toupper(org_name)) |>
  filter(
    # Keep individual/party donors; drop record keeping things, like refunds
    receipt_type_code %in% c("IND", "CPCM", "GEN", "PPTY", "OUTS", "NFPC"),
    !donor_name %in% c("AGGREGATED INDIVIDUAL CONTRIBUTION", "VARIOUS VARIOUS")
  ) |>
  add_address_lookup(name = "donor_address") |>
  select(sboe_id, report_id, donor_name, donor_address, amount, profession, employers_name, form_of_payment_desc, everything())

total_donations_per_committee <-
  receipts_2020_2022_src |> 
  group_by(sboe_id, donor_name, donor_address) |>
  summarize(total = sum(amount), .groups = "drop") |>
  filter(total >= 50)

receipts_2020_2022 <- 
  receipts_2020_2022_src |>
  semi_join(total_donations_per_committee, by = c("donor_name", "donor_address"))
```

## Donors

```{r donors}
donors <- 
  receipts_2020_2022 |>
  distinct(donor_name, donor_address) |>
  semi_join(total_donations_per_committee, by = c("donor_name", "donor_address")) |>
  dbplyr::window_order(donor_name) |>
  mutate(donor_id = row_number(), .before = 1)

donor_receipt_types <- 
  receipts_2020_2022 |>
  left_join(donors, by = c("donor_name", "donor_address")) |>
  group_by(donor_id, receipt_type_desc) |>
  summarize(n_records = n(), total = sum(amount), .groups = "drop")

donor_type <-
  donor_receipt_types |> 
  slice_max(total, by = donor_id, n = 1, with_ties = FALSE) |> 
  select(donor_id, donor_type = receipt_type_desc)
```

## Add Committee Information

```{r committee-info}
cover_2020_2022 <- 
  cover |>
  semi_join(report_list_2020_2022, by = "report_id") |>
  left_join(report_dates |> select(report_id, sboe_end_date))

committee_names <-
  cover_2020_2022 |> 
  mutate(committee_name = toupper(committee_name)) |>
  filter(!is.na(committee_name)) |>
  add_address_lookup(street_1, city, state, zip_code, name = "committee_address") |>
  group_by(sboe_id) |>
  slice_max(sboe_end_date, n = 1, with_ties = FALSE) |>
  select(sboe_id, committee_name, committee_address)

committee_types <-
  cover_2020_2022 |>
  filter(!is.na(committee_type)) |>
  group_by(sboe_id) |>
  slice_max(sboe_end_date, n = 1, with_ties = FALSE) |>
  ungroup() |>
  select(sboe_id, committee_type)

committees <- 
  committee_names |>
  left_join(committee_types, by = "sboe_id")
```

## Candidates

```{r candidate}
candidates <- 
  officers |>
  filter(type == "Candidate") |>
  semi_join(report_list_2020_2022, by = "report_id") |>
  left_join(report_dates |> select(report_id, sboe_end_date)) |>
  group_by(sboe_id) |>
  slice_max(sboe_end_date, n = 1, with_ties = FALSE) |>
  ungroup() |>
  select(sboe_id, candidate_name = name, candidate_address = address, candidate_phone = phone)
```

## Candidate Affiliation

```{r candidate-affilitaions}
candidate_listing <- targets::tar_read(candidate_listing, store = here::here("process", "_targets"))

candidate_2020_2022 <-
  candidate_listing |>
  filter(
    between(year(election_dt), 2020, 2022),
    name_on_ballot != "No Preference",
    party_candidate %in% c("REP", "DEM")
  )

candidate_affiliation <- 
  candidate_2020_2022 |>
  mutate(
    across(c(first_name, middle_name, last_name), ~ if_else(is.na(.), "", .)),
    candidate_name = paste(first_name, middle_name, last_name),
    candidate_name = gsub(" +", " ", candidate_name)
  ) |>
  distinct(election_dt, candidate_name, party_candidate) |>
  group_by(candidate_name, party_candidate) |>
  slice_max(election_dt, n = 1, with_ties = FALSE) |> 
  ungroup() |>
  select(-election_dt)
```


## Summary

```{r collect-data}
total_donations_per_committee_lcl <- total_donations_per_committee |> collect()

donors_lcl <- 
  donors |>
  left_join(donor_type, by = "donor_id") |>
  collect()

donor_type_amounts_lcl <-
  donor_receipt_types |> 
  collect() |> 
  left_join(donors_lcl, by = "donor_id") |>
  select(-donor_id, -n_records) |>
  relocate(donor_name, donor_address, donor_type) |>
  rename(name = receipt_type_desc, value = total) |>
  mutate(name = fct_reorder(name, value, .fun = sum, .desc = TRUE)) |>
  pivot_wider(values_fill = 0)

committees_lcl <- committees |> collect()

candidates_lcl <- 
  candidates |>
  collect() |>
  replace_na(list(candidate_name = "")) |>
  mutate(
    candidate_name = gsub('"[^"]+"', "", candidate_name),
    candidate_name = gsub("\\(.+\\)?$", "", candidate_name),
    candidate_name = stringr::str_squish(candidate_name),
    candidate_name = toupper(candidate_name)
  ) |> 
  jaccard_left_join(candidate_affiliation, by = "candidate_name", threshold = 0.8) |>
  rename(
    candidate_name_sboe = candidate_name.x,
    candidate_name_list = candidate_name.y,
    candidate_party = party_candidate
  ) |>
  group_by(sboe_id, candidate_name_sboe) |>
  slice(1) |>
  ungroup()

total_donations <-
  total_donations_per_committee_lcl |>
  left_join(
    donors_lcl |> select(-donor_id),
    by = c("donor_name", "donor_address")
  ) |>
  left_join(committees_lcl, by = "sboe_id") |>
  left_join(candidates_lcl, by = "sboe_id") |> 
  group_by(sboe_id) |>
  mutate(total_committee = sum(total)) |>
  ungroup() |>
  arrange(desc(total_committee), desc(total))
```

## Output

```{r prepare-receipts}
report_list_lcl <- 
  report_list_2020_2022 |> 
  select(year:amended, sboe_start_date, sboe_end_date, contains("received_")) |>
  collect()

receipts_2020_2022_lcl <- 
  receipts_2020_2022 |> 
  collect() |>
  semi_join(total_donations, by = c("sboe_id", "donor_name", "donor_address")) |>
  inner_join(
    report_list_lcl |> select(year:report_id),
    y = _, 
    by = c("sboe_id", "report_id")
  )

receipts_split <-
  receipts_2020_2022_lcl |>
  arrange(year, doc_name, report_id) |>
  (\(x) split(x, paste("Receipts -", x$year)))()
```

```{r create-sheets}
sheets <- list(
  "Donation Summary" = total_donations,
  "Donors" = donor_type_amounts_lcl,
  "Committees" = committees_lcl,
  "Candidates" = candidates_lcl,
  "Report List" = report_list_lcl
)

sheets <- c(sheets, receipts_split)
```

```{r preview-sheets}
sheets
```

```{r write-sheets}
writexl::write_xlsx(sheets, "donors_2020_2022.xlsx")
```