San Diego County’s political parties play a huge role in local elections by spending thousands of dollars to support their candidates. But the details of their spending are hard to know because of the way political parties file their campaign finance reports.
To uncover how much money political parties are spending in this year’s elections, inewsource compiled data from more than 120 financial reports — kept by the state in PDFs — dating to 2011. We then determined which races and candidates have received the most financial support from the county Republican and Democratic parties.
It may sound straightforward, but this was a convoluted process that took hours of manual labor to pull off. Here’s how we did it.
Where did you get the data?
Political parties receive and spend money for many reasons. One way they can spend money is to support candidates running for office. When a party files a financial report, it uses something called a Schedule D to record any money spent supporting or opposing candidates and ballot measures, along with details on how the money was spent.
In our local elections, the San Diego County Democratic and Republican parties are the ones that get involved. (Rarely do state or national political parties fund local elections.) Under state law, the parties are required to file financial reports with the California Secretary of State’s Office, which doesn’t provide detailed spreadsheets of the filings. The basic data downloads the state produces don’t include the exact descriptions of how political parties are spending money.
To get the data we were looking for, inewsource had to download the PDF filings of the financial reports for the county’s Democratic and Republican parties from 2011 to 2018. This totaled more than 120 reports that can each be dozens of pages long with as many as seven sections.
We wanted to narrow down each financial report to the one section we needed for analysis, the Schedule D. Using a PDF manager program, we exported only the Schedule D sections of each report into text files, which were sometimes still more than 10 pages each.
We imported the text files into Excel, even though the conversion process left in many blank rows, misspelled names and put data in the wrong spreadsheet columns. The data was cleaned and standardized using OpenRefine so candidates, offices and jurisdictions were all spelled correctly and listed in the same way to make for easy analysis.
Two of the fields from the PDFs did not transfer into the spreadsheets because they were checkboxes on the forms rather than written descriptions or numbers. These fields listed whether the transactions were monetary or non-monetary and whether the money went to support or oppose a candidate. These data were added in by hand for more than 3,000 transactions.
How did you categorize types of party spending?
When a political party wants to support a candidate, it can spend money in a few different ways, including through contributions or member communications. Though San Diego County has strict limits on how much money parties can contribute to political campaigns, no spending limits exist for member communications. That’s why parties rely on these to support candidates.
Even though member communications are very different from a direct contribution to a candidate, they are not easily distinguished on the parties’ financial reports.
inewsource classified each Schedule D transaction into a type of spending by looking at the descriptions and details of the transactions. If the description included “MBR” or “Member Communication,” we listed it as a member communication. If a box on the form was checked for “Monetary Contribution,” we listed it as a contribution. If neither applied, the transaction type was listed as “Other.”
What analysis did you do on your data?
By categorizing the data, we were able to analyze how much money the parties spent through these different means. We calculated how much money every candidate received from each political party from 2011 to 2018 through member communications.
To compare different types of party spending, we aggregated data from each election cycle (these are two-year periods — 2017-18, for example), through the June 5 primary. Since the 2017-18 election cycle is ongoing, we decided to only use data through the primary for each election cycle for a fair comparison. We then calculated how much money was spent by each political party in each cycle through member communications, contributions and other types of spending.
What other data did you use for your political parties story?
inewsource also compiled data from the secretary of state’s website listing every entity that donated money to each of the county’s political parties going back to 2013, and how much they gave. We calculated who gave the most money to each party in each election cycle.
We’ll let you know when big things happen.