About 95 percent of Covered California's 2014 ACA premium tax credit subsidy users filed tax returns for 2014. (Image: Covered California)

Most of the people who used the Affordable Care Act advance premium tax credit to pay for exchange plan coverage in 2014 filed tax returns, but California was much worse than other states at verifying the nonfilers’ identities.

Covered California started out letting applicants verify their own identities, by swearing, under penalty of perjury, that they were who they said they were.

A HealthCare.gov identity-proofing system, which used personal financial information to verify applicants’ identities, was much more effective at verifying identities than the warnings about perjury.

The office of the Treasury Inspector General for Tax Administration, a federal watchdog agency, has published those findings in a new report on the Affordable Care Act .

Related: 7 must-know new ACA tax administration facts

TIGTA keeps tabs on the Internal Revenue Service.

Consumers get APTC subsidy help by giving income forecasts to an ACA exchange. Consumers are supposed to report on the APTC help they get, using Form 8962, when they file their tax returns, to see if the IRS owes them extra subsidy help, or if they need to pay excess subsidy help back to the IRS.

The Senate Finance Committee asked TIGTA to look at APTC users who had failed to go through the APTC reconciliation process.

TIGTA analysts started by getting APTC user date from the HealthCare.gov states, and from the managers of the state-based exchanges in California, New York and Vermont.

Those exchanges served about 2.9 million of the households that used APTC help in 2014.

TIGTA has estimated, in earlier reports, that 2.9 million households reported $9.8 billion in APTC help on tax returns for 2014.

About 14 percent of all 2014 APTCusers failed to file tax returns.

In the states TIGTA studied for the new report, only about 1 percent of the APTC users were nonfilers with unverified identities.

In California, only about 5 percent of the APTC users failed to file tax returns, but Covered California had verified the identities of only about 7 percent of the 39,284 nonfilers. About 4 percent of the Covered California APTC users were nonfilers with unverified identities.

California accounted for about one-quarter of the families included in the study data but about 93 percent of the 35,276 nonfilers with unverified identities.

California exchange told managers its identity verification numbers look bad because, up until Aug. 4, 2014, the Centers for Medicare & Medicaid Services let it use an alternative process for verifying identities.

Until Aug. 4, 2014, applicants could qualify for APTC help in California by promising that they were themselves.

In the other states studied, exchange programs verified applicants’ identities by using a service that drew from the personal information stored in credit bureau records and other databases.

In those states, only about 0.1 percent of the f APTC users were nonfilers with unverified identities.

The TIGTA analysts also found signs that HealthCare.gov may have been tougher than some state-based exchanges at enforcing other APTC user requirements, such as the requirement that users not be incarcerated, and the requirement that users be in the country legally.

The analysts found, for example, that HealthCare.gov terminated the coverage of 1,436 APTC users due to incarceration, and 4,556 due to users’ problems with proving that they were lawfully present. The exchanges in California, New York and Vermont terminated the coverage of just 170 people due to either incarceration or immigration status problems.

Related:

3 things NAHU told the IRS about ACA premium tax credits

Watchdog finds IRS still struggling with PPACA form processing

We’re on Facebook, are you?