A new Government Accountability Office (GAO) report reveals that 9 of 12 fictitious Obamacare applicants it created were able to gain insurance outside the normal enrollment period. Here’s the setup from the GAO report:

Consumers can enroll in health-insurance coverage, or change from one qualified health plan to another,6 through the federal and state-based marketplaces either (1) during the annual open enrollment period or (2) outside of the open enrollment period, if they qualify for a special enrollment period (SEP)…

Our undercover testing included fictitious applicants claiming to have experienced an event that would trigger eligibility to enroll in health insurance coverage during an SEP…

We designed the 12 fictitious applications to provide either no documentation or fictitious documentation related to the SEP triggering event, when instructed to do so.

Special enrollment can be triggered by losing a job, marriage, moving or several other major life changes. These events are supposed to be backed up by documentation to prevent people who chose not to buy insurance during the annual open enrollment period from buying it later after they become ill. Politico reported back in January of 2016 that this was an ongoing problem:

Obamacare customers are gaming the system, buying coverage only after they find out they’re ill and need expensive care — a trend insurers warn is destabilizing the fledgling health law marketplaces and spiking premiums for everyone.

Insurers blame the problem on lax rules that allow more than 900,000 people to sign up for coverage outside the standard enrollment season — for instance, when they change jobs or move — without sufficient proof they are eligible. No one knows precisely how many might be manipulating the system, but the plans say they run up much higher medical bills and then jump ship, contributing to double-digit rate increases and financial losses.

Given that this is one of the problems potentially driving insurers out of the marketplace, the government should be pretty eager to crack down. The government did announce an attempt to crack down in September. And yet, the GAO found 9 of 12 fake applicants were able to get and keep insurance which included thousands of dollars of government subsidies:

The federal or selected state-based marketplaces approved coverage and subsidies for 9 of our 12 fictitious applicants who initially applied online or by telephone seeking coverage during an SEP, as of October 2016. For these 9 applications, we were approved for APTC subsidies, which totaled about $1,580 on a monthly basis, or about $18,960 annually. These 9 applicants also each were approved for CSR subsidies, putting them in a position to further benefit if they used medical services.

The GAO created this figure outlining who was approved and who was rejected:

The current report by the GAO did not include any new recommendations to the Department of Health and Human Services (HHS) because the GAO has previously issued recommendations after similar investigations found a failure to detect fake applicants. A report in 2014 found 11 of 12 fictitious applicants were able to get insurance. In October 2015 17 of 18 were approved. So the exchanges do seem to have improved slightly at fraud detection, though a 25% success rate several years into the program is really nothing to brag about.