Data-Smart City Solutions

Search

By Nyasha Weinberg

Every year, thousands of city dwellers arrive home to overcrowded and illegally subdivided houses. Kitchens, dining areas, sheds and even under-stairs cupboards are converted to serve as bedrooms.  

Migrants, students, and families are forced into cramped living conditions by the immense pressures on housing in growing cities. Homes are illegally subdivided into increasingly tiny units to accommodate booming populations; often this involves haphazard electricity and plumbing arrangements, raising serious safety concerns.

The grave consequences of overcrowding need to be urgently addressed by governments. Cities must provide affordable, quality and safe housing. To fulfill this duty many cities, including Boston and NYC, are looking to data to solve the crisis.

Why Should We Be Concerned About Overcrowding?

Influxes of students, migrants, and job hunters exert considerable pressure on city housing. Maps generated from open data, like this example from New York, show that thousands of households across large cities are afflicted. Limited stock leads owners, landlords, and tenants to carve up buildings, packing in residents. In one house, a bedroom had been creased inside a walk-in freezer. The majority of landlords adhere to maximum occupancy and zoning laws, but some unscrupulous individuals keen to make a quick buck exceed legal limits. Certain neighborhoods within large cities are full of people forced by circumstance to live in substandard conditions, causing distress to buildings, occupants, and neighborhoods.

Overcrowded and deteriorating buildings endanger residents. This tragic truth was recently illustrated by a house fire in Boston involving students living in substandard conditions. Homes illegally converted to accommodate extra residents can contain “unexpected walls, barred-up windows, and locked interior doors” that trap residents and confuse firemen when blazes occur. In New York, 19 of the 49 fires over a two-year period that resulted in fatalities occurred in buildings altered in a manner prohibited by zoning laws.

Alongside the increase in fire risk, overcrowding also affects wellbeing: it is associated with respiratory disease, tuberculosis, mental health problems and higher mortality rates among women. A study of the effects of overcrowding on children showed negative outcomes on math and reading achievement, behavior, and physical health. Finally, illegal conversions create problems for city government: they strain essential services, inflate property values and cause difficulties for firefighters by rendering building plans inaccurate.

How can cities get an accurate picture of which households are overcrowded?

To get an accurate picture of which households are overcrowded, cities must work across agencies to collect data from multiple sources. Datasets must be shared; otherwise, actionable information about overcrowding cannot be produced. Officials must seek to work across state, federal, and third-party boundaries, overcoming legal and organizational resistance to information sharing. Difficulties, as outlined in Code for America’s Beyond Transparency: Open Data and the Future of Civic Innovation, still remain in connecting the diverse data provided by different agencies. As an example, the forty agencies that operate in New York have traditionally focused on their individual data collection responsibilities for internal use only. This localized approach makes information sharing extremely difficult. Each agency has a separate ontology of terms reflecting individual operations and these cannot be easily tied together.

The Boston fire occurred, in part, because privacy laws precluded information sharing between schools and city government. Having comprehensive information on where residents live is the first step in identifying problematic housing. This is especially true of at-risk groups like students who are more likely to live in accommodation that violates occupancy rules.

Cities have found some solutions and workarounds to the resistance towards data sharing from departments and agencies. In Boston, Mayor Marty Welsh emphasized that privacy laws protecting student information could still be followed, as only addresses rather than student names were needed to determine the location of potentially problematic buildings. All city schools but two have now been persuaded to give data to the city.

The additional student housing data enabled city officials to identify approximately 580 properties in need of inspection from a list of 25,000. The Inspectional Services chief stated this new information sharing provides a more accurate picture of the number of individuals sharing a unit, and has improved the possibility of inspectors identifying likely zoning violations.  Furthermore, as a result of improved systems, landlords have significantly improved at complying with regulations and addressing complaints on their own.

Fire and rescue services have also discovered the benefits of sharing information to leverage regularly updated dynamic data. Atlanta’s Firebird program collects data about properties and structures  from diverse departments such as the Office of Buildings and AFRD. The data enables Atlanta to monitor properties as they change hands or usage type.

Federal housing data also informs risk: the LA Times used 2008 and 2012 data from the Census Bureau’s American Community Survey to map crowded homes by zip code. Overcrowding was calculated by comparing the weighted number of households against the national average crowding rate. Elsewhere, the data is used by the U.S. Department of Housing and Urban Development to administer funding for community development programs to alleviate crowding, by philanthropies determining community sustainability and in NYC’s assessment of hidden households

DATA AND OVERCROWDING

Once information has been compiled, cities must allocate funding to support initiatives tackling overcrowding. There are two primary ways that data can assist: preemptively by preventing housing shortages, or effectively by identifying those overcrowded properties most likely to endanger residents.

PREVENTING OVERCROWDING

Overcrowding can be complex, but data can be used to calculate the need for additional housing. California faces the problem of having both the most expensive housing in the country and two of the most overcrowded counties in the United States. It has started to look at data to address this problem, developing a model that compiles data on house prices, education, incomes, and the weather to determine likelihood of future demand. This is compared with housing supply estimates based on land area, topographical constraints, and construction labor wages to determine housing sufficiency and likelihood of overcrowding.

If overcrowding looks likely, programs can be introduced to release pressure on housing. Examples include: relaxing development standards by increasing the number of multi-family units for development, facilitation of single room occupancy through zoning code revisions, and expanding affordability by working with nonprofits to assemble land and write down costs. By taking this data-smart preventative approach, cities design housing policies that address problems before they even happen.

PREVENTING THE CONSEQUENCES OF OVERCROWDING

Unfortunately, in many situations cities are already packed to the breaking point. Cities then need to effectively harness available information, using predictive risk analytics to monitor overcrowding. Making sense of available data can help target building inspections. By determining the most endangered properties and occupants, officials can triage inspections and conserve limited city resources, following the regular practice of the fire service.

Monitoring illegal conversions and levels of distress is best achieved by organizing datasets to help identify priorities. Cities should move beyond traditional tracking and management tools; they must not just collect historical data, but also data on other features correlated with overcrowding.

Street surveys and thermal imaging have not been as effective as initially hoped, but less traditional indicators are providing valuable insights. One London borough looks at sewage flow to calculate occupancy by comparing expected waste output with observed waste; trash left on the streets and regular calls to pest control teams indicate houses strained to breaking point. In New York, a project looked at whether landlords paid taxes, noise violations, and complaints to identify likely candidates for building code violations.

Such efforts have already proven their worth in reducing workload, relative to tools that simply track the period of time elapsed since last inspection. In New York, application of these strategies led the detection rate for major infractions to rise from 8% to 70% of inspections. Such an enormous jump in productivity shows that creating risk models can ultimately save cities money by pinpointing priority inspections.

Cities must work smarter, not harder.  For cities to grow sustainably, they need to carefully manage their housing stock without risking dangerous occupancy levels that threaten lives. Protecting residents involves collaborative work to gather data about housing, and clever data analysis to drive preemptive action and more efficient resource allocation once a problem has been identified. 

About the Author

Nyasha Weinberg

Nyasha Weinberg is a Kennedy Scholar from the UK currently studying for a MPP at the Harvard Kennedy School. She has previously worked for the UK government providing policy advice about the uses of data in emergencies management, for the World Food Program and on urban education policy in Rio de Janeiro. 

Comments

comments powered by Disqus