December 22, 2019

Abstract

The Federal Election Commission is an independent government agency created by Congress in an effort to promote confidence and participation in U.S. elections. The FEC provides a variety of data about candidates, parties, and other committees publicly. I will focus on the contributions made to the 2020 Presidential Campaign. This topic will be increasingly more relevant as we move closer to the 2020 Presidential Election.

In this project I’ll answer the following questions:

1. Which candidate is supported the most?
2. Which candidate received the greatest amount of donations?
3. Who are those donors? What do they do?
4. Do men and women donate the same amount of money?
5. Is Elizabeth Warren supported financially by both the genders equally?

Data

There are 18 variables in the dataset. Some of the information provided in the dataset is ,Contributor name and hometown, Contributor employer, Contributor occupation, Date of contribution.

A detailed description can be obtained by taking a summary of the dataset. Additionally, to map the names of the contributors to their genders I will be using the “gender” libaray provided in R. Library uses millions of American names to create a model and predicts the gender based on the basis of first name of the candidate. To obtain the latitude and longitude information I will use the “zipcode” library.

This dataset contains 82013 contributions and 18 variables. Checking the distributions of the contributions, I found that there are lot of outliers which are either extreme high or low values. Some of these values are also negative. Hence, I tranform to log scale for a better understanding of the contribution. The distribution appears normal and it seems most of the donations are in smaller amount. The minimum donation is -6100.0 and the maximum donation is 2721782.1. The negative contributions which were refunded and the contributions which exceeded $2700 were removed from the dataframe as these amounts will be refunded.

## 'data.frame':    82013 obs. of  18 variables:
##  $ cmte_id          : chr  "C00580100" "C00580100" "C00580100" "C00580100" ...
##  $ cand_id          : chr  "P80001571" "P80001571" "P80001571" "P80001571" ...
##  $ cand_nm          : chr  "Trump, Donald J." "Trump, Donald J." "Trump, Donald J." "Trump, Donald J." ...
##  $ contbr_nm        : chr  "KULIK, HENRY C." "OMALLEY, JOHN J" "MAZZOLA, BARBARA" "WHORISKEY, THOMAS" ...
##  $ contbr_city      : chr  "LEOMINSTER" "MILTON" "WATERTOWN" "SAGAMORE BEACH" ...
##  $ contbr_st        : chr  "MA" "MA" "MA" "MA" ...
##  $ contbr_zip       : int  1453 2186 2472 2562 1020 1510 2155 1106 1545 1226 ...
##  $ contbr_employer  : chr  "HENRY KULIK CPA LLC" "VICUS THERAPEUTICS" "INFORMATION REQUESTED" "RETIRED" ...
##  $ contbr_occupation: chr  "CPA" "BUSINESS" "INFORMATION REQUESTED" "RETIRED" ...
##  $ contb_receipt_amt: num  3.75 750 104.33 75 26.25 ...
##  $ contb_receipt_dt : chr  "13-May-18" "25-Aug-18" "20-Jan-17" "22-Mar-18" ...
##  $ receipt_desc     : chr  NA NA NA NA ...
##  $ memo_cd          : chr  "X" "X" "X" "X" ...
##  $ memo_text        : chr  NA NA NA NA ...
##  $ form_tp          : chr  "SA18" "SA18" "SA18" "SA18" ...
##  $ file_num         : int  1263561 1301594 1174081 1263561 1301594 1248056 1248053 1301594 1301594 1174081 ...
##  $ tran_id          : chr  "SA18.1359591" "SA18.1591907" "SA18.422960" "SA18.1398669" ...
##  $ election_tp      : chr  "P2020" "P2020" "P2020" "P2020" ...

As we can see, contribution amounts seem to be somewhat normally distributed. There does seem to be a few outliers to the far right. One could argue that the distribution is slightly skewed right because of this. The vast majority of donations made seem to be a modest dollar amount, with the outliers being very large donations.

1. Which candidate is supported the most?

The top most supported candidate is Elizabeth Warren, which is expected since she is a Massachusetts Senator. Surprisingly, Donald Trump is a close second with Bernie Sanders trailing closely behind in third place. Despite having a Republican Governor, Massachusetts is known to be a liberal/progressive state so having Donald Trump as a close second is unexpected.

2.Which candidate received the greatest amount of donations?

This chart measures dollar amount contributed to candidates. Similar to the first chart, Elizabeth Warren is at the top. Donald Trump once again secures second place, this time with a wider margin between him and Warren. Interestingly, Bernie Sanders loses his third place spot to Pete Buttigieg and falls to fourth place instead. This shows that high support from individuals does not always translate to higher financial support. This could be for a number of reasons like income and occupation of contributors.

3. Who are those donors? What do they do?

With this chart I observed that “NOT EMPLOYED” is the occupation that made the highest amount of donations. The pool of money available from this demographic should be among the lowest, so how could this be? Perhaps the individuals who fall under this category have enough money that they do not need to work or receive financial assistance from sources that allow them to make these donations. This may highlight a limitation of surveying, when individuals give inaccurate information or misrepresent information about themselves.

4. Do men and women donate the same amount of money?

In the last elections, there was no statistically significant difference between contributions from male voters and female voters. For the upcoming elections, I analysed this by doing a Welch Two Sample t-test.

Null hypothesis \(H_{0}\) : Men and Women contribute equally on an average to the elections in the state of Massachusetts Alternative hypothesis \(H_{a}\) : Men and Women didn’t contribute equally on an average to the elections in the state of Massachusetts

## 
##  Welch Two Sample t-test
## 
## data:  contb_receipt_amt by gender
## t = -10.911, df = 77147, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -20.69924 -14.39514
## sample estimates:
## mean in group female   mean in group male 
##             85.29705            102.84424

We reject \(H_{0}\) at 95% significance level. The test indicates that contributions are not equal for males and females. Males contribute more towards the elections.

5. Is Elizabeth Warren supported financially by both the genders equally?

Hillary Clinton drew a large number of female voters during the 2016 election. Let’s see if one of the leading female candidates, Elizabeth Warren, has more statistically different contributions on an average from both genders.

Null hypothesis \(H_{0}\) : Men and Women contribute equally on an average towards Elizabeth Warren in the elections in the state of Massachusetts Alternative hypothesis \(H_{a}\) : Men and Women didn’t contribute equally on an average towards Elizabeth Warren in the elections in the state of Massachusetts

## 
##  Welch Two Sample t-test
## 
## data:  contb_receipt_amt by gender
## t = -10.178, df = 24931, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -24.55385 -16.62386
## sample estimates:
## mean in group female   mean in group male 
##             82.25539            102.84424

We reject \(H_{0}\) at 95% significance level. The test indicates that contributions are not equal for males and females for Elizabeth Warren. Though on an average, she does recieve more money from females than other candidates do.

Bonus: How Republican/Democratic is Massachusetts?

Here the map shows the locations of contributors as well as the parties they donated to. Boston residents have overwhelmingly donated to the Republican party while other parts of the state are more of a mix. This raises questions, since Eastern MA is usually seen as the more liberal, urban part of Massachusetts while Istern MA is seen as the rural conservative area of the state. One reason for this result could be the popular view that the Republican Party is more aligned with the elite/corporate class while the Democratic Party represents the working class. With Boston being the wealthiest city in the state with the highest incomes and property values, it indicates why the majority of donations went to the Republican Party.

Bonus: Contribution trend over time.

Clearly, as the election comes closer, people tend to contribute more. Additionally, as soon as previous election is over there is a surprising interest in contributions towards next election. This maybe due to an emotional outburst. This can also be used to create a model for predicting how much contribution a candidate can get at a given point of time, in the campaign cycle.

Conclusion

As a result of our analysis I conclude that (1) Elizabeth Warren is the most supported candidate; (2) Elizabeth Warren received the greatest amount of donations; (3) donors’ occupations ranged from Attorney to Software Engineer, but the unemployed demographic donated the most amount of money; (4) male and female voters did not contribute equally, male voters contributed more towards the upcoming election; (5) male and female voters did not equally contribute to Elizabeth Warren’s Campaign, but Elizabeth Warren did in fact receive more donations from female voters compared to other candidates; (6) Boston may be less liberal than it is thought to be, as seen by its support for the Republican Party; (7) as the 2020 Presidential Election approaches, people tend to donate more in preparation.