I call it a “data crime” when someone is abusing or misusing data. When we understand these stories and their implications, it can help us learn from the mistakes and prevent future data crimes. The stories can also be helpful if you must explain the importance of data management to someone.
The Story
The state of Rhode Island sent tax refunds to over 100 individuals and businesses signed by “Mickey Mouse” and “Walt Disney.” These were not the names of the Treasurer and State Controller, who should have signed the refunds. Having incorrect names meant that the refunds could not be cashed at the bank.
The state first blamed the error on a “technical glitch.” They later claimed that they “incorrectly sourced” the signatures.
Regardless of the cause, they then had to quickly process refunds that would be accepted by banks. There were people who badly needed this money and the government had to resolve it as soon as possible.
What We Learned
First of all, it’s never a “glitch.” A person caused this problem, although we don’t know who. While not meant to be malicious, it was an accident that wouldn’t have had to happen if good data management practices had been followed.
While the entire story never came out, I assume the issue happened when they wanted to test the printing of the refunds. Testing is a great thing to do, but it sounds like they decided to use a signature file that contained bogus names so that the refunds could never be cashed in case they fell into the wrong hands. If that was what happened, those test refunds got mailed out. That’s just one option for testing. There are other alternatives too.
They could have used check paper with “void” on all the checks. They could have used test names, like “test,” that weren’t real people or recognizable fictional characters. At the very least, they should have had a task on their project plan to switch the bogus signature file used in the test environment with the real signature file to be used in the production environment.
Maybe they didn’t have a test environment. Maybe they didn’t have a project plan. It’s hard to know what happened, but it seems to stem from carelessness. Good data management practices have been around for decades. Even before then, it was common sense: Good organization and project management skills will take you far.
