In ranawg/Div_iii_fec: Data Package for the 2016 Election

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 6, fig.height = 4.5
)
library(fec16)
library(scales)
library(dplyr)
library(tidyr)
library(ggplot2)
library(ggrepel)

The fec16 package provides tamed datasets of the Federal Election Commission (FEC)'s 2015-2016 election cycle.

Inspiration

We wanted to create a data package that is easy to use for people just beginning to learn R, for example, in introductory statistics and data science classes. For this purpose, this package is nice because students and instructors do not need to worry about the unnecessary data wrangling and can immediately use the data for analysis.

Our package is inspired by Hadley Wickham's nycflights13 package. We used the data taming principles from Albert Kim's fivethirtyeight package, which are explained in detail here.

Basics of tame data

We used the tame data principles to produce uniform data frames that are easy to navigate and link together. The following are the guidelines we used to tame our data:

1. Variable names:
2. Lower case and with underscores instead of spaces using clean_names()
3. 20 characters or less
4. The same names were used for the same variables in different data frames
5. Names to be meaningful
6. Variable types:
7. Encode variable with dates with as.Date, unless it is only a year then it would be numeric type
8. Encode categorical variables that a have a limited number of values with as.factor otherwise it would be of character type
9. Tidy or long data format instead of wide
10. Missing data has value NA
11. Removed signs such as % and $ attached to numeric values

Other notes

• We dropped variables that are not useful for analysis from the original FEC datasets
• We summarized large datasets with relevant data in order to compress its size

Data frames included:

All the datasets are taken from FEC

candidates: all candidates registered with the FEC during the 2015-2016 election cycle

committees: all committees registered with the FEC during the 2015-2016 election cycle

results: the results of the 2016 general presidential election

individuals: a sample of 5000 individual contributions to candidates/committees during the primary and general 2016 elections

committee_contributions: total contributions, aggregated by candidate, from committees

Who should use this package?

Mainly, anyone interested in US politics and elections and wants to use actual data to think critically and make inferences about it should use this package. Additionally, since we made this package with students and their instructors in mind to make working with the data itself as smooth as possible, it is a good set of data for teaching. Moreover, this package can provide inspiration for questions and projects that are data-driven and based on the real world.

What does the data look like?

The first six rows of the results dataset look like:

head(results)

What can we do with this data?

We can use this package to address the (non-exhaustive) list of questions:

1. Which presidential candidate won majority in more states?
2. What is the relationship between contributions of candidates and total votes they get?
3. Which candidate got the most popular vote and how many?
4. Is there a trend in the contributions to a candidate depending on their party?

To answer our questions we can make use of some data wrangling and data visualization techniques. Some examples (addressing the questions above) are shown below:

Example 1:

Which presidential candidate won a majority in more states?

Using the results dataset, we can also see what actually happened in the 2016 elections.

Here is how we can summarize the number of wins by candidate:

wins <- results %>%
  filter(winner_indicator == "W") %>%
  group_by(last_name) %>%
  summarise(num_wins = n())
wins

We can show the results using a simple bar chart:

Each win is for a single state. There are 51 total wins for the 50 States and Washington, D.C.

ggplot(wins, aes(x = last_name, y = num_wins, fill = last_name)) + 
  geom_col() +
  scale_fill_manual(values = c("blue", "red")) + 
  labs(
    title = "Number of States Won: Clinton vs. Trump",
    x = "Candidate", y = "Count", fill = "Candidate"
  )

We can see that Trump had majority in more states than Clinton.

Example 2:

What is the relationship between contributions of candidates and total votes they get?

Here we investigate what kind of relationship, if any, the candidates with over a 1000 votes got with the total contributions they made. We are interested in candidates with an ID number so we can join them with the contribution data set by using inner_join.

results_by_cand <- results %>%
  drop_na(general_results, cand_id) %>%
  group_by(cand_id, last_name) %>%
  summarise(sum_votes = sum(general_results)) %>%
  filter(sum_votes > 1000) %>%
  inner_join(committee_contributions, by = "cand_id")
results_by_cand

Next, we plot the contributions and votes on a scatter plot and plot a trend line that would make it easy for us to see the relationship. Since there are outliers in the data, it is best to not use a continuous axis scale in order to see all of the points.

ggplot(results_by_cand, mapping = aes(x = total_contributions, y = sum_votes)) +
  geom_point() +
  scale_x_log10(labels = comma) +
  scale_y_sqrt(labels = comma) +
  geom_smooth(method = "auto") +
  labs(title = "Contributions vs. Votes in 2016", 
       x = "Contributions in US Dollars", y = "Total Votes") + 
  geom_text_repel(aes(label = last_name))

As we can see, the highest contributors got the highest amount of votes so it has a positive correlation.

Example 3:

Which candidate got the most popular vote and how many?

Visualize the results of the popular vote in the elections and see how many people voted:

results_by_cand <- results %>%
  drop_na(general_results, cand_id) %>%
  group_by(cand_id, last_name) %>%
  summarise(sum_votes = sum(general_results)) %>%
  filter(sum_votes > 100000)

ggplot(results_by_cand, mapping = aes(x = last_name, y = sum_votes, fill = last_name)) +
  geom_col() +
  xlab("Candidates") +
  ylab("Number of Votes") + labs(title = "How Many People Voted?", fill = "Candidate") +
  scale_y_continuous(labels = comma)

We can see that Clinton got more popular votes than Trump! It looks like Clinton has about 65 million votes.

Example 4:

Is there a trend in the contributions to a candidate depending on their party?

We can create a scatter-plot of total contributions per candidate, colored by party (only including Democratic and Republican parties).

joined_data <- candidates %>%
  full_join(committee_contributions, by = "cand_id") %>%
  filter(!is.na(total_contributions)) %>%
  filter(cand_pty_aff == "REP" | cand_pty_aff == "DEM") %>%
  filter(total_contributions <= 1e+07)

ggplot(joined_data, aes(x = cand_id, y = total_contributions, color = cand_pty_aff)) +
  geom_point() +
  geom_jitter() +
  theme(axis.text.x = element_blank()) +
  labs(
    title = "Contributions per Candidate by Party",
    x = "Candidates", y = "Total Contributions ($)",
    color = "Candidate Party Affiliation"
  ) +
  scale_y_continuous(labels = comma)

It doesn't look like there are any specific patterns in the contributions (for contributions under $ 10 million).

ranawg/Div_iii_fec documentation built on Dec. 23, 2019, 10:29 p.m.