Companies doing business in Denver have been able to file their sales and use taxes online through the City and County of Denver’s eBiz Tax Center since October 2012. To drive adoption of this time-saving tool, Denver’s Treasury Division sent out letters to 20,479 businesses each month to encourage usage.

These efforts increased the usage of this resource from 0% at the start to just over 50% in late 2015. However, Treasury management noticed that the uptake rate had plateaued over the last several months. They determined that perhaps their standard letter, stating “Go Green – File Online” had gotten stale and wasn’t having the impact they desired.

When the Behavioral Insights Team came to Denver in November 2015, they met with Treasury staff to discuss this issue and potential options. The team devised a couple of trials to increase site registration and encourage past users to come back to the site, using behavioral science concepts such as “loss aversion” and “social norms”.

The team drafted two additional letters to drive businesses to the eBiz Tax Center website, designed to get more clicks to the site and increase registration and online filing:

The three different letters were randomly allocated to different businesses and sent out at the beginning of January. The results of the trial showed that the social norm message outperformed all three letters in click-through rates, registration, and filing, and the two new letters (the loss aversion and social norm messages) outperformed the status quo “Go Green” message (noted as ‘Control’ below).

Letters - clickthroughseBiz Letter Result - Registrations

eBiz Letter Result - Online Filing

If the whole sample received the social norm letter, 267 more businesses would have paid their taxes online compared to the control, saving the city $1,692.78 in mailing and processing costs (compared to over-the-counter or mail-in transactions). Additionally, 546 more businesses would have registered to pay online. Due to these results, the Treasury Division will be replacing the “Go Green” letter with the “Social Norms” letter. However, additional analysis of user behavior on the eBiz Tax Center site is needed to determine why 50% of those who registered did not file online.

The second arm of this eBiz evaluation involved sending emails to businesses who had registered for eBiz and filed in the past, but hadn’t filed online for two tax periods. Over 12,000 businesses were identified to be in this category, and were randomly allocated to one of three groups: receive no email (control group), or receive a loss aversion or social norm email.

Interestingly, in terms of filing taxes online there was no statistical difference between those who received either email or no email at all:

eBiz Email Result - Online Filing

However, the loss aversion email was more effective than the social norms email at driving visits to the eBiz site (a statistically significant result at the 95% confidence level):

eBiz Email Result - visits

If the loss aversion email was used for the whole sample of 8,498 businesses, 1,185 more businesses would have visited the eBiz site from the email link. Denver can use this information to inform strategies that encourage individuals to use other online services, such as the Open Data Catalog, paying parking tickets, applying for services, etc.

It’s important to note, however, that this evaluation didn’t happen without a hiccup. A miscommunication caused more letters and emails to be sent out than was initially determined, resulting in some people mistakenly receiving a letter that shouldn’t have (they were one-time filers) and some people who received the email who recently used eBiz. As a result, Treasury staff had to field a lot of calls & emails from people wondering why they received the correspondence.

While this caused a headache for Treasury staff, the trial results aren’t significantly impacted or made null, as Joseph Boskovski from the Behavioral Insights Team explains:

“Hiccups can be expected when using administrative data in a new way or rolling out a new process. The good thing with running a Randomized Control Trial is that these problems are roughly evenly distributed across these groups, so it does not harm our ability to make causal statements.”

The Tax Compliance unit of the Treasury Division is excited by the results of this experiment, and will not only be using the more effective innovations from this evaluation but will be applying these concepts later to continuously improve their operations.

Bruce Moore, Director of Tax Compliance, commented that, “I’m constantly looking for ways to make it quick and easy for businesses to pay their taxes, and seeking methods to encourage online filing. This technique provides a sound, evidence-based strategy for testing ideas and focusing on what works.”

What’s more, these results can help other city agencies utilize similar nudge tactics to drive traffic online or increase participation in a program.

1 Comment »

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s