How We Investigated California’s Campaign Finance Watchdog

A recent CalMatters analysis found that investigations by the Fair Political Practices Commission, California’s campaign watchdog, can take years, at times only resolved after the investigated politician has won an election or vacated office, leaving some voters in the dark as they fill out their ballots and vet a candidate’s character.

When we started reporting on this story, we began with a straightforward question: How long does it take for the commission to resolve cases? The answer was not easy to find.

Data Collection

Throughout our investigation, we filed nine public records requests, analyzed more than 3,000 pages of documents and spoke to a total of 25 people, including former and current commissioners and staffers, ethics advocates, state and local elected officials and people who have filed complaints with the commission.

Collecting case data

First, we filed a Public Records Act request with the commission. We asked for data about all cases resolved by Jan. 1, 2024, including details of those cases and when they were each opened and closed. We also asked for cases that remained open as of the start of the year.

Meanwhile, to double-check that data, we also compared the results from our records request with a data analysis of open and closed cases that we downloaded directly from the commission’s case search portal.

The data we collected:

• Records request database: A database of open and closed cases we received from the commission in response to our public records request. The earliest record was a case opened in 2002; the next earliest was opened in 2009. We narrowed down the database to only contain cases opened since 2017. (See Incorrect case closure dates and Inconsistencies in post-2016 data.)
• Downloaded database: A database on case information we compiled based on data available on the commission’s own case search portal. We used this information to cross-reference the records request database but found too many flaws to rely on the data. (See Limitations and Commission Responses.)

Due to all the inconsistencies we noticed in both databases, however, we decided not to depend too heavily on them for detailed statistics, instead using them largely as a reference guide. We used the records request database only to calculate the duration of cases since 2017. We cross-checked the records request database with the commission’s search portal, agendas and reports for specific case information.

Using the records request database, we then analyzed the time it took to resolve each case by calculating the number of days between the date the case was opened and the date it was closed.

Gathering documents and statistics

We reviewed thousands of pages of documents and hours of meeting footage to paint a more accurate picture of the commission’s enforcement efforts. We also reviewed the commission’s executive staff reports, case enforcement records, commission meetings and agendas, staff memos, budget change proposals and legislative reports.

The commission publishes caseload and case resolution statistics, including the number of warning letters and penalties issued each year, in annual reports and executive staff reports, which are presented at each commission meeting.

Researching specific cases

We filed six public records requests for documents on specific cases on state and local officials, asking for records from both the commission and the politicians, including details of open probes into Gov. Gavin Newsom, Insurance Commissioner Ricardo Lara and Attorney General Rob Bonta. For example, the commission sent us documents about Newsom’s ongoing investigation for late behested payment reporting that the governor’s office declined to share.

In the case of former Compton City Councilmember Isaac Galvan, we built a timeline of communications between enforcement staff and Galvan’s associates based on documents we received from the commission to assess how the case took six years to resolve. We cross-referenced his bank records and interactions with the commission with his social media posts for certain details, such as his promotion of the film “Daddy’s Home 2” on the day a subpoena server tried to reach him and his posting of a hangout with friends, which the commission investigation concluded was paid for with campaign cash.

Findings

Investigation length

We learned that cases that took longer to resolve skewed the average: A total 15% of cases closed during that period of time took more than two years to resolve, with the longest taking almost seven years, according to our analysis of the records request database.

In some cases, investigations lasted so long that the politician in question had left office or won an election, sometimes even several elections, before the cases were adjudicated.

Upcoming election

Commission data shows, as of Oct. 23, that 20 of the 305 candidates for the state Legislature, U.S. House and U.S. Senate on the Nov. 5 ballot have an open case against them:

1. Anamaria Avila Farias (Democratic) – Assembly District 15
2. George Barks (Republican) – Assembly District 66
3. Jessica Caloza (Democratic) – Assembly District 52
4. Franky Carrillo (Democratic) – Assembly District 52
5. Clarissa Cervantes (Democratic) – Assembly District 58
6. Phillip Chen (Republican) – Assembly District 59
7. Steven “Steve” Choi (Republican) – State Senate District 37
8. Damon Connolly (Democratic) – Assembly District 12
9. Carl DeMaio (Republican) – Assembly District 75
10. David Fennell (Republican) – State Senate District 1
11. Matt Haney (Democratic) – Assembly District 17
12. Robert Paul Howell (Republican) – State Senate District 15
13. Kevin Kiley (Republican) – Congressional District 3
14. Evan Low (Democratic) – Congressional District 16
15. Jessica Martinez (Republican) – Assembly District 56
16. Dave Min (Democratic) – Congressional District 47
17. Stephanie Nguyen (Democratic) – Assembly District 10
18. Tami Nobriga (American Independent) – Assembly District 9
19. David Tangipa (Republican) – Assembly District 8
20. Ken Weir (Republican) – Assembly District 32

Constitutional officers

• Two of the state’s eight constitutional officers are currently under investigation. Both won re-election as their cases were pending.
  • Gov. Gavin Newsom is under an investigation that began in 2021 for late disclosure of behested payments.
  • Insurance Commissioner Ricardo Lara is being investigated for allegations of “laundered campaign contributions” in 2022.
• Seven of the eight top constitutional officers have had past violations, ranging from improper disclosures to illegal campaign contributions, according to commission enforcement records.
  • Gov. Gavin Newsom was issued a warning letter for failing to include proper disclosures in text messages produced by his 2021 anti-recall campaign.
  • Attorney General Rob Bonta was fined for late reporting of behested payments in 2021.
  • Secretary of State Shirley Weber was similarly fined $600 for late behested payment reporting.
  • Insurance Commissioner Ricardo Lara was also sent a warning letter for not filing behested payment disclosures on time.
  • Treasurer Fiona Ma was issued a warning letter for not submitting a financial disclosure report on time and separately fined $11,500 for improper recordkeeping.
  • Controller Malia Cohen was sent a warning letter for not disclosing a behested payment on time.
  • Superintendent of Public Instruction Tony Thurmond was sent a warning letter this year for violating behested payment-related regulations.

See the Appendix to view and download the data acquired from the records request.

Limitations and Commission Response

Although the commission publishes a public database of past and ongoing investigations, we found enough inconsistencies and flaws with it that we relied on the records request database we received for much of our analysis. Here are some of the issues we discovered.

Missing fields

There is no way to download the entire database from the commission’s website at once. Instead, you can only search for cases by case number, county, city, violation type or involved parties.

At first, we searched by county, then downloaded cases from each county along with statewide cases. But that did not capture all cases, because some cases in the database did not contain any jurisdictional information.

Next, we noted that the nine-digit case ID numbers appeared to be formatted to include the year the investigation was opened. We searched for “2024-,” which seemed to return all cases opened this year. However, we found that cases opened before 2016 were not consistently formatted in this way: a search for “2016-” yielded investigations that opened both in 2016 and before 2016.

Incorrect case closure date

Some cases on the commission’s case search portal contain inaccurate information, CalMatters found. For example, the website shows that a San Diego County case regarding a local tax ballot measure was concluded in 2016 after it was opened in 1996. But the case was actually resolved in 2000, according to the commission’s summary of enforcement decisions.

Jay Wierenga, spokesperson for the commission, told CalMatters in a May interview that pre-2016 data “is not as reliable.” That’s in part because the commission that year adopted a new recordkeeping system, where pre-2016 cases looked as if they had been closed in 2016, which is not necessarily the case, he said in a May email.

“In essence, much of the data from pre-2016 was not carried over to the new system as we would’ve liked, due to budgetary constraints,” Wierenga wrote.

“We do the best we can with the tools we are given,” he said in the interview.

Inconsistencies in post-2016 data

But even just looking at cases opened after 2016, we found inconsistencies in the downloaded database:

• Some cases appear to be associated with two case numbers because cases can be resolved together, Wierenga said in an email from May. We submitted another records request for cases opened since 2017 that were resolved together. Cinthya Bernabe, commission counsel for the enforcement division, responded that the agency had conducted a “thorough search” and did not have any materials responsive to our request. In an email this week, Wierenga said that the FPPC does not keep track of cases that are resolved together.
• Cases might also have two case numbers if the case existed before the new system was implemented in 2016, because it would have been assigned a new case number starting with “2016” and would have also retained its original number, Wierenga wrote in an email.
• Some case numbers are assigned to multiple cases. There are two cases labeled as 2019-01565, one in Los Angeles County and one in Sacramento County.
• Not all warning letters issued by the commission can be found on its search portal, even though the website says it includes all such letters since July 1, 2009. For example, for the case numbered 2018-01209, which was concluded in November 2018, we could not find the warning letter when searching by keyword, month or year it was resolved.

The records request database included a fine amount for each case, which we intended to use to calculate fines the commission issued by year. But there is a risk of double-counting some fines, because the commission told us some cases were resolved together, yet the data does not indicate which ones. For example, the data shows two $1.35 million fines levied against Los Angeles County, each linked to a separate case, when in fact the amount represented the total amount the county was fined when the cases were resolved together.

Disclosure of “confidential” cases

When CalMatters obtained the records request database in March, it included cases that did not appear in the commission’s public search portal.

Upon a CalMatters inquiry regarding the discrepancies, enforcement chief James Lindsay said in a May email the cases were “confidential” and should not have been released. He asked us to remove them from our analysis, even though some of them are published on the commission’s website. He and Wierenga declined to explain why some of those cases, along with penalties associated with them, were still publicly accessible.

As of publication, case materials for some of those cases are still available on the commission’s website.

The data CalMatters analyzed does not include those cases, which make up less than half of one percent of the total volume.

Mismatched data

We also noticed that the number of cases resolved some years in the records request database did not match statistics in the annual reports the commission publishes on its own website. For example, the database includes about 1,050 cases that were closed in 2021, but the annual report showed 1,590 — a difference of more than 500 cases.
After we reached out for an explanation of this discrepancy, Wierenga told us that 1,590 was a “typographical error.” He added that the correct number of cases closed in 2021 was 1,058 and that the commission would update the report with the correct number (we verified that it did) and that it instituted “stricter review protocols to ensure that such a discrepancy does not arise in the future.”

Appendix