People often blame technology for widening the income gap and creating social ills.
Tipping Point’s use of data analytics shows that it can also improve lives.
The US-based non-profit, which fights poverty in the Bay Area for 1.3 million people, recently tied up with the San Francisco city government to find out the impact of parking fines on low-income drivers.
The weapon: data analytics. The result was the introduction of new reforms by the San Francisco Municipal Transportation Agency (SFMTA) that reduced the burden on low-income drivers.
Tipping Point’s efforts to tackle poverty are not new. The non-profit already had a track record of showing how seemingly small actions can disproportionately impact low-income families.
“But we also wanted to explore how tickets were given and managed. Were there any trends in the data that could relate to low-income communities? Was there any unconscious bias built into the system?” Jamie Austin, Senior Director, Impact + Learning, Tipping Point queried.
Tipping Point knew the answers hid within the government data. However, the size of the data was daunting and needed a solution that could handle such a vast volume.
"There were over 7 million citations to analyze and dozens of fields to look at. That's the sort of thing that will bring Excel to a grinding halt. Also, we were just building our data science practice, and while our Board was encouraging us to bring techniques of big data and data science to bear on social problems, we didn't have the tools. So, we wanted to find a way to use technology that was approachable, easy to use and manageable for the questions we were trying to ask," Austin said.
However, data management was not the only hurdle. Gaining meaningful insights was proving to be equally challenging.
“The data was difficult to manage, but after cleansing, the data seemed reliable and complete. The only thing missing was demographic information: Who was the driver? Where did they live? This initially seemed like a major hurdle. How can we assess the impact on low-income individuals if we don’t know anything about the driver except where they happen to park?” Austin asked.
So they turned to TIBCO Spotfire Data Science. The various features helped Austin and his team to answer these questions.
"First, [TIBCO Spotfire Data Science] can scale to pretty much any size of data. It uses these big data platforms to offload all the computations into something that's proven to scale. Second, it was really easy to use. We were able to cleanse the data and start testing hypotheses immediately, like for example which zip codes get the most tickets compared to the number of meters it has?" Austin said.
The TIBCO team also helped the data science manager to build a workflow that merged all the data and start the analysis quickly.
“Everybody could see what everyone else was doing, which made it seamless for collaboration,” Austin said.
In a few days, Tipping Point was able to answer crucial questions on the impact of parking fines.
“And the questions we were asking were pretty hard,” Austin said.
For example, Tipping Point studied whether parking fines affected low-income people or certain neighborhoods more by looking that vehicles' age as a proxy for income. It analyzed the license plate numbers, which are sequential in California, to determine the vehicle ages. Addresses were cross-matched to get neighborhood statistics.
"Both of those were pretty hard to do, but Spotfire Data Science made it not only easy but totally transparent so that members of the team could understand and check each other's work," Austin said.
The research and Tipping Point’s partnership with the Financial Justice Project raised awareness and drove results.
On May 15, 2018, SFMTA decided unanimously to lower towing fees for the 25% of San Franciscans who earned less than USD 50,200 for a family of four.
Since July 1, 2018, the same families did not have to pay either their first-time tow fee or the associated USD 269 administrative fee.
“This change alone is expected to save low-income communities -- those making less than 200% of the federal poverty line -- hundreds of thousands of dollars,” Austin said.
SFMTA also allowed eligible low-income families with outstanding debt to forgo late fees and pay the remaining citation fees through a reduced low-fee payment plan, or use the Community Service Program.
Understanding that beneficiaries often miss out on public programs due to low awareness, Tipping Point created a Facebook ad campaign.
According to Austin, the campaign reached 52,000 people and drove more than 1,500 to right website within ten days, "so that the percentage of individuals receiving low-fee payment plans per week increased by more than 160%."
Austin attributed the positive outcomes to the power of analytics and the capabilities of TIBCO products.
"The project would have never happened were it not for the Spotfire Data Science products and the patience, expertise, and commitment of the TIBCO team," he said.