Looking into FEC Individual level fundraising for Democratic Party

The FEC requires candidates file their contributions for the 2020 elections with the agency. The data is uploaded as forms which the FEC cleans and adds to CSV files. The data is a treasure-trove of information. Polls get attention from week to week, but the news generally lacks any deep analysis of the FEC filings. Reporting typically covers headline numbers advertising total volume of donations and total amount. The FEC data is rich with self-identifying donor information. People identify their position, industry and location. Money acts as a proxy for electoral success. In the 2020 primary almost all candidates have sworn off contributions from lobbyists. On the surface, this is a noble attempt by democratic candidates to prove they aren’t influenced by corporations. However, digging below the surface, there is plenty of corporate influence in the FEC data.

Below, I look through the FEC data to find which candidates are receiving money from corporate level executives. We can also see if corporate executives are more likely to attempt to capture multiple candidates with donations and which candidates are receiving money from the same executives. I mean it when I say this a is a treasure-trove of unexplored data. I will explain my process for analysis shortly, but first let me address some boring data integrity issues. Feel free to skip tp the section titled “The Plan”, the next several paragraphs are here to outline my assumptions in data collection for anyone who distrusts my analysis.

Data integrity concerns

The FEC mandates quarterly reports to track the individual contributions into democratic candidate’s accounts. As a disclaimer It seems the engineers on the FEC website are slowly going through the individual files on the candidates and adding them to the Big table located on FEC pages here. Originally, I pulled the data from the above link, but the raw data is available on the website in full and some of the candidates q2 FEC filings hadn’t been added to the above database. Therefore, for this report I just grabbed all the individual level csv files and did my own manipulation on the data. It seems I keep several individual level contributors’ (less than .1%) for each candidate that were discarded for whatever reason by the FEC data engineers. The FEC filings aren’t representative of the whole population of donors, as reporting under 200 $ isn’t mandatory. We know that all these candidates have reached 100k contributors, yet most candidates in the database have less than 100k donations listed.

Some other peripheral information, caps are given on individual level contributions at 2800 per year. Throughout the FEC report there are some people who exceed these limits, they are known as conduits. Conduits are individuals that act on behalf of other people to give a donation to a candidate. These conduits aren’t restricted by the $ 2800 level if they exercise no discretion over where the funds go.

Assumptions

Below I make several categorizing assumptions in my data manipulation operations. The regex approach I took to clustering these groups will absolutely include some examples that do not actually represent the categorization of job title that I have made. RegEX is a brute force attempt to categorize the occupations algorithmically. However, as I have tracked the original job occupation, you can see overwhelmingly that these job titles are correctly identified. I display tables for each new clustered position, so that the reader understands that the approach is largely effective. My aggregated DF is at my website, I have categorized all job titles that have been changed with a binomial field to track them. Please feel free to look at the assumptions I made and the false captures in the clusters. if you would like to take this on in your own analysis, or perhaps develop your own clusters, please feel free to do so as well. The original unedited job titles are still present in the dataset.

The Plan

There are many different combinations of the same job title, my solution is to create new clustered occupation columns, which tracks and categorize occupations based upon corporate structures of power. I.E looking at the FEC data you would find examples of the words “C.E.O” “Chief executive officer” “chief officer” “Chief executive”. I bundle all these roles into a singular role of CEO. Other Self-reporting words like “Founder”, “EXECUTIVE DIRECTOR”, and “President” are also representative of those in charge of their organization, so they are also labeled as C.E.O. From there I step down the corporate ladder and attempt to identify all general C level executives as c level executives and place the rest of executive positions into a category of eXECS.

This was all done in careful order to preserve the hierarchy of attempting to capture positions correctly. For instance, Executive secretaries were captured first as secretary, to not show up as executive level job positions. After i completed corporate power structures, I moved onto clustering the rest of the data by the occupation type. The main clusters formed were IT, Academia, Blue-collar workers, Legal field, Science & medicine, law enforcement, and artistic/freelance.

## [1] 7522

## [1] 8552

#sort(table(clustered_df$clustered_jobs),decreasing=TRUE) [1:100]


cluster_jobs <- function(df,reg_express,replacement_val){    
   
    ## Populate new col with imputed value
    df$clustered_jobs[str_detect(df$clustered_jobs,reg_express)] <- replacement_val
    ## Build column to track if change was made to occupation 
    df <- df %>% 
            mutate(job_changed = ifelse(clustered_jobs == job_position,'0','1'))
    return(df)
}

# 
# clustered_df$clustered_jobs[str_detect(clustered_df$clustered_jobs,"ACCOUNT ")]
# 
animated_donation_by_job(job_positions="aTTORNEY",graph_title="Attorneys",filename="Attorneys")
animated_donation_by_job(job_positions="aCADEMIA",graph_title="ACADEMIA",filename="ACADEMIA")

# animated_donation_by_job(job_positions="ItFIELD",graph_title="ItFIELD",filename="ItFIELD")
# animated_donation_by_job(job_positions="rETIRED",graph_title="rETIRED",filename="rETIRED")
# animated_donation_by_job(job_positions="mEDIA/cREATIVE",graph_title="creative white collar",filename="creativewhitecollar")
# animated_donation_by_job(job_positions="sCIENTIST/pHYSICIAN",graph_title="medical and science",filename="medicalscience")

## Function changes occupation based upon a regular expression. 
## Creates column to note if change was made,stores old names, returns DF
cluster_jobs <- function(df,reg_express,replacement_val){    
   
    ## Populate new col with imputed value
    df$clustered_jobs[str_detect(df$clustered_jobs,reg_express)] <- replacement_val
    ## Build column to track if change was made to occupation 
    df <- df %>% 
            mutate(job_changed = ifelse(clustered_jobs == job_position,'0','1'))
    return(df)
}

 ## Build new_jobs_col
final_df <- final_df %>% 
  mutate(job_position= toupper(job_position))  %>% 
  mutate(clustered_jobs = job_position) 

## create Date column

## Job Clustering
clustered_df <- cluster_jobs(final_df,"CEO","CEO")
clustered_df$clustered_jobs <- as.character(clustered_df$clustered_jobs)
clustered_df$clustered_jobs[is.na(clustered_df$clustered_jobs)] <- "Not selected"

## GRAB SOME EXLCUSIONS
clustered_df <- cluster_jobs(clustered_df,"EXECUTIVE RECRUITER|EXECUTIVE SEARCH|RECRUITER","rECRUITER")


## Executive level positions
## WORDS INTENTIONALLY ARE REPLACED WITH SOME LOWERCASE LETTERS

clustered_df <- cluster_jobs(clustered_df,'EXECUTIVE ASSISTANT|EXEC. A|EXECUTIVE ASSITANT|EXEC ASST|EXECUTIVE ADMINISTRATOR|EXECUTIVE COORDINATOR|EXECUTIVEASSISTANT','EX aSSISTANT')
clustered_df <- cluster_jobs(clustered_df,"CHIEF EXECUTIVE OFFICER|MANAGING PARTNER|EXECUTIVE DIRECTOR|^PRESIDENT\\b|^FOUNDER\\b|EXEC.*DIR","CEO")
#clustered_df <- cluster_jobs(clustered_df,'^PRESIDENT\\b|^FOUNDER\\b|EXEC.*DIR',"CEO")
clustered_df <- cluster_jobs(clustered_df,'CHIEF \\w* OFFICER|^C[A-D|F-Z]O\\b|/C[A-D|F-Z]O\\b|^EXECUTIVE OFFICER',"ClEVEL_eXECS")
clustered_df <- cluster_jobs(clustered_df,'EXECUTIVE VICE PRESIDENT|FINANCIAL DIRECTOR|CORPORATE EXECUTIVE|SVP|EXECUTIVE VP|EVP|MANAGING DIRECTOR|CHAIRMAN|CORPORATE OFFICER|COO$','ClEVEL_eXECS')
clustered_df <- cluster_jobs(clustered_df,'^EXECUTIVE$|EXEC|EXECUTIVE$|^PARTNER|^DIRECTOR|MANAGING|CHAIR OF| CHAIR|BOARD |CO-C',"eXECS")
clustered_df <- cluster_jobs(clustered_df,'^VP|VP\\b|VICE PRESIDENT','Vp roles')


## RANDOM EXCLUSIONS
clustered_df <- cluster_jobs(clustered_df,"CHRISTIAN|^MINISTER|RABBI|PRIEST|CHAPLAIN|PARISH|CLERGY|BISHOP|PASTOR",'rELIOGOUS')

## LEGAL
clustered_df <- cluster_jobs(clustered_df,'LAWYER|ATTORNEY|GENERAL COUNSEL|COUNSEL$|PROSECUTOR|ARBITRAT','aTTORNEY')
clustered_df <- cluster_jobs(clustered_df,'LEGAL|JUDGE|CRIMINAL|COURT','lEGALfIELD')

## IT JOBS
clustered_df <- cluster_jobs(clustered_df,"SOFTWARE|FULL STACK DESIGNER|PROGRAMMER|DATA|TECHNOLOGY|SYSTEMS|DBA|COMPUTER|CYBER|INFORMATION|NETWORK|TECH|UX|GRAPHIC|I\\.T\\.|^IT|^WEB",'ItFIELD')



## Academia

clustered_df <- cluster_jobs(clustered_df,'PROFESSOR|TEACHER|SCHOLAR|LECTURER|DOCTORAL|INSTRUCTOR|ECONOMIST|EDUC|DEAN|TUTOR|RESEARCH','aCADEMIA')
clustered_df <- cluster_jobs(clustered_df,'STUDENT','sTUDENT')
clustered_df <- cluster_jobs(clustered_df,'LIBRAR','lIBRARIAN')
clustered_df <- cluster_jobs(clustered_df,'ACADEMIC|COLLEGE|SCHOOL|HIGHER ED|MATH|^PRINCIPAL$','aCADEMIA') 
## Unemployed self employed
clustered_df <- cluster_jobs(clustered_df,'^SELF','sELF eMPLOYED')
clustered_df <- cluster_jobs(clustered_df,'^NOT-EMPLOYED$','nOT eMPLOYED')


## CREATIVE 
clustered_df <- cluster_jobs(clustered_df,'DESIGNER|ARTIST|MUSICIAN|FILM|CREATIVE|PHOTO|EDITOR|WRITER|PUBLISHER|AUTHOR|GRAPHICS|JOURNALIST|SCULPTOR|ART |CARTOON|^ART|PUBLIC RELATIONS|PUBLICIST|TELEVISION|DESIGN|COMPOSER|MUSIC|DESIGN|DIGITAL MEDIA|DANCE|SINGER|NOVEL|BLOG','mEDIA/cREATIVE')
clustered_df <- cluster_jobs(clustered_df,'^(AUTHOR|ACTOR|ACTRESS)','mEDIA/cREATIVE')
clustered_df <- cluster_jobs(clustered_df,"TALENT|PRODUCER|SOCIAL MEDIA ",'mEDIA/cREATIVE')
clustered_df <- cluster_jobs(clustered_df,"ACTIVIST|ENVIRONMEN",'aCTIVIST')


## FINANCIAL JOBS
clustered_df <- cluster_jobs(clustered_df,'^(FINANCIAL|INVESTOR|INVESTMENT|CAPITAL|ENTREPRENEUR|FINANCE|INVESTOR$|FINANCIAL|INVESTOR|BANKING)','fINANCIAL sECTOR')
clustered_df <- cluster_jobs(clustered_df,'^VENTURE|EQUITY|HEDGE|QUANTITATIVE','fINANCIAL sECTOR')
clustered_df <- cluster_jobs(clustered_df,"ACCOUNT |ACCOUNTS |COMPLIANCE|AUDIT|ACTUARY|FRAUD|ESTIM|SURVEYOR",'aCCOUNT sPECIALISTS')


## Labor 
clustered_df <- cluster_jobs(clustered_df,'^(PIPE|STEAM|ELECTR|UNION|CARPENTER|MECHANIC|WELDER|UNION)','bLUECOLLAR')
clustered_df <- cluster_jobs(clustered_df,'LABOR|CONSTRUCTION|ELECTRICIAN|PLUMBER|PIPEFITTER|MACHINIST|WOOD|WELDERFISHER|PAINT','bLUECOLLAR')
clustered_df <- cluster_jobs(clustered_df,'^(TRUCK|DRIVER)|DRIVER|TRUCK|DELVIERY|UPS|FEDEX|RIDESHARE','bLUECOLLAR')
clustered_df <- cluster_jobs(clustered_df,'CONTRACTOR','cONTRACTOR')

## SCIENCE AND MEDICINE
clustered_df <- cluster_jobs(clustered_df,'PHYSICIAN|MD|M\\.D\\.|CHIROPRACTOR|DOCTOR|VETERI|DENTIST|SURGEON|PEDIATRICIAN|TRIST$|SCIENTIST|PHYSICIST|CHEMIST|LAB|LABORATORY|LABOROATORY|SCIENCE|BIOLOGIST|GEOLOGIST|STATISTICIAN|OLOGIST|ICIAN$','sCIENTIST/pHYSICIAN')
clustered_df <- cluster_jobs(clustered_df,'PSY|SOCIAL WORKER|SOCIAL WORK|FAMILY THERAPIST|CASEWORKER','mENTAL hEALTH')
clustered_df <- cluster_jobs(clustered_df,'^RN$|APRN|CRNA|NURSE|CAREGIVER|DAYCARE|HOMECARE|CHILDCARE|CHILD CARE|HOME CARE|CARETAKER|\\WCARE\\W|CARE PROVIDER','rN/nurse/cAREGIVER')
clustered_df <- cluster_jobs(clustered_df,'MASSAGE|OCCUPATIONAL|PHYSICAL THERAPY|PHYSICAL THERAPIST|ACUPUNCTURIST|LMT','iNJURY_RECOVERY')
clustered_df <- cluster_jobs(clustered_df,'PHARMA','pHARMA')
clustered_df <- cluster_jobs(clustered_df,'MEDICAL|PARAMEDICAL|MEDICINE|DENTAL|PARAMEDIC','mEDICAL')

## SKILLED LABOR
clustered_df <- cluster_jobs(clustered_df,'ANALYST|ENGINEER|ARCHITECT','wHITE_COLLAR')
clustered_df <- cluster_jobs(clustered_df,'CPA|ACCOUNTANT|ACCOUNTING|BOOKKEEPER|TAX','wHITE_COLLAR')
## Real estate
clustered_df <- cluster_jobs(clustered_df,'REALTOR|REAL ESTATE','rEAL ESTATE')

## Secretary
clustered_df <- cluster_jobs(clustered_df,'SECRETARY|RECEPTIONIST','sECRETARY')
clustered_df <- cluster_jobs(clustered_df,'\\w ASSISTANT|ASSISTANT$','sECRETARY')
clustered_df <- cluster_jobs(clustered_df,'RESOURCES|HR|CAREER|VOCATION','hr_cAREER_sERVICES')

## MILITARY AND LAW ENFORCEMENT SECURITY
clustered_df <- cluster_jobs(clustered_df,'MILITARY|ARMY|SOLDIER|LAW ENFORCEMENT|FIRE|SECURITY|POLICE','lAW eNFORCEMENT')


clustered_df <- cluster_jobs(clustered_df,"^BAKER|COOK|CHEF|DISH|WAITER|WAITRESS|BARTENDER|RESTAURANT|BARISTA|BAR\\W|BARBACK",'rESTURANT')
clustered_df <- cluster_jobs(clustered_df,"RETIRE",'rETIRED')
clustered_df <- cluster_jobs(clustered_df,"BANKER|MORTG",'bANKER')
clustered_df <- cluster_jobs(clustered_df,"CONSULTANT",'cONSULTANT')
clustered_df <- cluster_jobs(clustered_df,"SALES",'sALES')
clustered_df <- cluster_jobs(clustered_df,"OWNER",'bUSINESSoWNER')
clustered_df <- cluster_jobs(clustered_df,"MANAGER",'mANAGERS')
clustered_df <- cluster_jobs(clustered_df,"ADMINISTRATOR",'aDMINISTRATORS')
clustered_df <- cluster_jobs(clustered_df,"HEALTH",'hEALTHcARE')
clustered_df <- cluster_jobs(clustered_df,"FARM",'fARMER')
clustered_df <- cluster_jobs(clustered_df,"PILOT",'pILOTS')
clustered_df <- cluster_jobs(clustered_df,"MARKETING",'mARKETING')
clustered_df <- cluster_jobs(clustered_df,"MARKETING",'mARKETING')
clustered_df <- cluster_jobs(clustered_df,"PROFIT|FOUNDATION|CHARITY",'nONPROFIT')
clustered_df <- cluster_jobs(clustered_df,"MANAGEMENT",'mANAGEMENT')
clustered_df <- cluster_jobs(clustered_df,"NONE",'NOT EMPLOYED')
clustered_df <- cluster_jobs(clustered_df,"RETAIL|CASHIER",'rETAILER')
clustered_df <- cluster_jobs(clustered_df,"RESEARCHER",'aCADEMIA')
clustered_df <- cluster_jobs(clustered_df,"ADVERTISING|RAISING|FUNDRAISE",'aDVERTISING')
clustered_df <- cluster_jobs(clustered_df,"VOLUNTEER",'VOLUNTEER')
clustered_df <- cluster_jobs(clustered_df,"CUSTOMER SERVICE|CUSTOMER CARE",'CUSTOMER SERVICE')
clustered_df <- cluster_jobs(clustered_df,"BUSINESS|ENTREPRENEUR",'bUSINESS_sERVICES')
clustered_df <- cluster_jobs(clustered_df,"ENVIRONMEN|ORGANIZER",'aCTIVISTS/oRGANIZER')
clustered_df <- cluster_jobs(clustered_df,"POLICY|PUBLIC|GOVERNMENT|GOVT|FEDERAL EMPLOYEE|COUNCIL|LEGISLAT|DIPLO",'pUBLICpOLICY')
clustered_df <- cluster_jobs(clustered_df,'FOUNDER','CEO')
clustered_df <- cluster_jobs(clustered_df,'VEND','vENDOR')
clustered_df <- cluster_jobs(clustered_df,'DOG|ANIMAL|HORSE|PET CARE|PET-CARE','aNIMALS_sERVICES')
clustered_df <- cluster_jobs(clustered_df,'FITNESS|PERSONAL TRAINER|STRENGTH|ATHLETIC','fITNESS_tRAINER')
clustered_df <- cluster_jobs(clustered_df,'DISPATCH','dISPATCHER')
clustered_df <- cluster_jobs(clustered_df,'SUPERVISOR','sUPERVISOR')
clustered_df <- cluster_jobs(clustered_df,'INSURANCE','iNSURANCEfIELD')
clustered_df <- cluster_jobs(clustered_df,'DRESSER|^HAIR|COLORIST','sTYLIST')
clustered_df <- cluster_jobs(clustered_df,'COMMUNICATIONS','cOMMUNICATIONS')
clustered_df <- cluster_jobs(clustered_df,'CLERK','cLERKS')
clustered_df <- cluster_jobs(clustered_df,'LINGUIST|INTERPRET|TRANSLAT','lANGUAGE')
clustered_df <- cluster_jobs(clustered_df,'INVESTIGATOR','iNVESTIGATOR')
clustered_df <- cluster_jobs(clustered_df,'JANITOR|CUSTODIAN','jANITOR')
clustered_df <- cluster_jobs(clustered_df,'OFFICE','oFFICEjOBS')
clustered_df <- cluster_jobs(clustered_df,'[A-Z]*MAN$','bLUECOLLAR')