Exclusive: How New York City used analytics to solve urban challenges

How do you protect housing renters from discrimination; engage parents in the education system; and preventing a large-scale epidemic of a deadly disease? For Dr Amen Ra Mashariki, the answer is always the same: data.

Mashariki was the city’s second Chief Analytics Officer, leading the Mayor’s Office of Data Analytics from 2014 to May this year. His job was to translate the city’s challenges into questions that could be answered using analytics, and get agencies to use their data in new ways.

“Whenever the Mayor ran into a complex problem, he called on my team,” Mashariki says. Mashariki now heads up urban analytics at Esri, the GIS company. He talked with GovInsider about how tech has transformed his city, and could transform yours too.

Preventing housing discrimination

New York’s Human Rights Law prohibits landlords from discriminating against tenants. Yet income discrimination is one of the top housing-related complaints received by the city’s Commission on Human Rights.

It had “years upon years” of data from calls made by victims, Mashariki says, including the landlords’ names and addresses.

Last year, the city’s analytics team worked with the commission to “predictively identify where and which landlords are likely to discriminate”.

The team combined complaints data with crime, education, city planning and housing data, turning it into a “map that geographically identified where these occurrences were likely to happen”.

The commission then sent in actors and actresses posing as tenants to identify landlords who were discriminating. Some had housing vouchers and others didn’t, to test how they were received by landlords and building management companies.

In total, it conducted over 300 such tests. Within two months, the commission charged five landlords and housing brokers - which together control about 20,000 units across the city - for repeatedly discriminating against tenants based on their income.

The commission has also stepped up efforts to file investigations on behalf of the city, with the 120 cases field last year from just 22 in 2014.

Preventing a legionnaires epidemic

In 2015, 128 people were infected and 12 people died as a deadly disease spread across the city. Legionnaires bacteria was being spread through untreated water in cooling towers on buildings.

With over 1 million buildings in the city, tens and thousands of lives at risk, and limited resources for inspection, Mashariki and his team were asked to use data to identify locations with the biggest risks.

The city, however, had no existing list of all cooling tower locations. The team worked round-the-clock for weeks, pulling in fragments of data from multiple agencies.

They built a machine learning algorithm to identify buildings likely to have contaminated cooling towers. The team raised the hit rate for identifying cooling tower locations from 10% to 80% with data, Mashariki says.

8 in every 10 attempts to identify buildings with cooling towers was successful, allowing inspectors to identify contaminated cooling towers faster.

Targeting communications

This year, Mayor Bill de Blasio announced a new scheme offering free pre-kindergarten for all three-year-olds in the city. To be a success, however, the scheme had to be inclusive of all people and have a high enrollment.

The government launched a citywide campaign, sending officials door-to-door to tell people about the new scheme. But with close to 9 million people living in the city, officials had to be micro-targeted to ensure all eligible children had a chance to enroll.

“We needed to give every New Yorker the opportunity to know about and enroll in this programme,” Mashariki says. “We used data to identify parents who were likely to need and want free pre-kindergarten.”

His team increased enrollment by 117% using analytics. With targeted and accurate information, frontline officials were able to better engage and educate parents about the programme.

This allowed people to pre-enroll before the start of the scheme. The data accuracy also ensured residents had a “strong customer experience” with the government.

People weren’t being repeatedly bothered by different city officials and canvassers did not go knocking on the wrong doors.

“We changed how government was perceived by a lot of New Yorkers through that effort”, Mashariki believes.

These case studies all show how vital analytics can be for a modern city. And, when it comes to killer apps, location is the missing link.

It brings the data together, helping officials “identify process, methodologies, capabilities and strategic ways for urban centres to solve complex problems”.

“One of the things I’ve learnt is that every in a city happens somewhere, so location is of utmost importance,” he says. Esri are the experts – download the following white paper to find out more.