Case Study | VA Census
In Partnership with Illinois Department of Human Services
Virginia needed to increase the Census 2020 response rate of all households during the COVID-19 pandemic. In particular, they were interested in focusing on “hard-to-count” communities, those that historically have been undercounted in the Census. CCL developed a text based solution that reached more than 2 million registered voters.
Problem
Given the COVID-19 pandemic and shelter-in-place mandates, the Commonwealth of Virginia had to find alternative ways to continue Census outreach in the absence of grassroots or in-person community outreach efforts. Every uncounted person in the 2020 Census could result in a community losing $2,000 in funding per person; that’s $20,000 per uncounted person over a 10-year period.
CCL Solutions Used
✓ Text Reminders
☐ Interactive Surveys
☐ Eligibility Screening
☐ Intelligent Referrals
☐ Document Verification
✓ Call Scheduling
Solution
CCL partnered with the Commonwealth of Virginia to create a digital Census outreach solution. The text-based solution reached more than 2 million registered voters on their cellphones April 3 –4, 2020.
Results
Virginia posted some of the largest gains of any state in Census self-response rate over the two days. They outperformed other states by 21% during the days when people received our text reminders, more importantly, they increase was concentrated in “Hard-to-Count” communities.
Program Details
CCL generated a list of valid mobile phone numbers from the Virginia voter list. Virginia, with the help of CCL, crafted a single outbound message that was sent directly to 2 million registered voters over a two-day period, April 3–4, 2020. The text asked recipients if they have taken the Census, and included a link to the online Census form.
Key Takeaways
A key challenge in sending SMS text messages is respecting privacy and ensuring user consent. We were able to reduce opt-out rates dramatically by offering a “soft” opt-out option where someone could indicate that they had already completed the Census, rather than just opting out of all messages. The increase in self-response was primarily in hard-to-count communities.