,
16 tweets,
5 min read
Read on Twitter
Gather round, children, as I share a timeless tale of fundamentally broken #GovTech:
At the end of 2018, before I started my new job, my fiance explored the possibility of applying for Medicaid until we found a private insurance option that would let her see her doctors. (1/n)
At the end of 2018, before I started my new job, my fiance explored the possibility of applying for Medicaid until we found a private insurance option that would let her see her doctors. (1/n)
Like many states, Illinois's digital infrastructure for public benefits case management leaves a lot to be desired. So even after we found private insurance, which meant that all we had to do was withdraw her case in their system, I knew I had to set significant time aside. (2/n)
We received a letter containing instructions to access the system. But when we went to the site, the path outlined in those instructions didn't match what was there. Account creation is easy enough to interpret for us, but what about for somebody with low digital literacy? (3/n) 
After clicking the tiny account creation link below the login dialog (which wasn't styled as a link), things started getting worse. GovTech projects tend to be very resource-constrained, and the people making them do their best, but what happened next was beyond the pale. (4/n)
The case management system has clearly-stated password requirements, seen here. Unfortunately, the password acceptance filter seems to use different standards. Password after password was rejected, despite adhering stringently to the requirements. Eventually we had success. (5/n)
If a benefits applicant hadn't disengaged from the process when confronted with the unintuitive login screen, this experience could have made them give up in frustration. And if that wasn't enough, after successfully creating an account, we were completely unable to log in. (6/n)
Having spent some time working around GovTech systems, Megan and I know that sometimes there are inefficient processes that have to happen for seemingly simple things. So while another applicant may have totally given up here, we waited a day for the account to propagate. (7/n)
Tried again. No luck. We went back through the account creation process and made an entirely new user profile. Again, no luck. Wait one more day. Finally, two full days after getting the letter, we're let into the system for the first time with our second account. (8/n)
It goes without saying that once inside, the case management system didn't improve. This is true in both human terms (the site is filled with jargon, and none of the mailings help somebody understand what it all means) and in tech terms (holy free-text date fields, Batman). (9/n)
When we found the form that we needed to inform the state that Megan had private coverage, her health insurance provider (one of the largest insurers in the country, with two corporate offices in Chicago) wasn't listed in a dropdown that included many defunct providers. (10/n)
And as one final "fuck you", when we finally managed to submit the application withdrawal we were greeted with this. Lots of good a submission status number will do if you don't know who to give it to in order to find out whether your submission has been processed. (11/n)
It took hours of work over two days to get to this point, with two people who deeply understand GovTech. For folks who aren't as privileged with that professional background, this would be another whole level of nightmare. It's not a "laugh at bad tech" moment; it's scary. (12/n)
Over and over and over again, we see that systems built for our most marginalized communities are the least functional. This impacts more than just the individuals who discontinue their applications out of confusion and frustration. It has policy ramifications, too. (13/n)
Signup numbers are used to justify the budgets for programs like Medicaid. When somebody in need isn't able to access services because of bad tech, that isn't actively seen by policymakers looking at the "raw" numbers. Entire programs can be shrunk due to disengagement. (14/n)
That's why the work of organizations like @codeforamerica, @mRelief_form, and @humanutility are so important. They build user-friendly alternatives to broken GovTech processes. And more than that, they encourage #CivicTech activists to improve these systems from within. (15/n)
Because the fact is, bad GovTech won't get better if we just keep shouting at it. And marginalized voices won't be at the table to help build these systems right if we (as in you and I) don't directly include those voices in our activism and our work. So let's get to it. (16/16)
