Research Profile

Brandeis University's National Equity Research Database (NERD) for Boston

Dr. Clemens Noelke

Dr. Clemens Noelke









The National Equity Research Database (NERD) for Boston is a project of the Institute for Child, Youth and Family Policy at Brandeis University’s Heller School for Social Policy and Management. NERD Boston is led by Research Director Clemens Noelke and the team behind the Institute’s flagship research project, We sat down with Dr. Noelke to talk about the project.


Tell us a little about yourself.

We are a research team based at the Institute for Child, Youth, and Family Policy, located at the Heller School for Social Policy and Management at Brandeis University. Our flagship research project is, a website and data project that involves research, policy analysis, data products, and dissemination on racial/ethnic inequities in child opportunity.

One of our data products is the Child Opportunity Index (COI), featured in a Boston Globe cover story. The COI has users in Boston and across the U.S. and is currently undergoing a major update. Like all BARI members, everyone on our team has over the years done research on and contributed to equity-related causes in the Boston metro area. For example, we developed the NERD prototype for a workshop convened by the Bertelsmann Foundation to support the Mayor’s Office of Resilience and Racial Equity. Our work is funded by the WK Kellogg Foundation.


What is NERD and NERD Boston?

The Boston edition of the National Equity Research Database, or “NERD Boston,” is a database of 286 socio-economic indicators on a wide range of subjects, including demography, education, employment, health, housing, earnings and ancestry covering the Boston metropolitan area.

NERD Boston is unique in providing:

  • A wide range of indicators,
  • For Boston neighborhoods,
  • Broken out by race/ethnicity (160 of the 286 indicators),
  • Over time, from 2009 to 2016.

Beyond Boston neighborhoods, every indicator is also available for the Boston metro area, the City of Boston and all other municipalities in the Boston metro, as well as the Commonwealth, New England, and the nation so users can compare data across the metro area and benchmark findings against larger areas.

NERD Boston is drawn from our much larger National Equity Research Database (NERD). It contains all indicators included in NERD Boston, but for ten geographic summary levels covering the entire country. NERD and NERD Boston indicators are calculated from the American Community Survey (ACS) Summary Files.

We are currently in beta mode, i.e., data and documentation may still be incomplete or modified in future releases. We expect to release a major update in early 2019, which will add a new wave of data, additional indicators from the American Community Survey, as well as other sources.


How is this different from other big database projects?

We’ll answer in order from least to most important:

NERD Boston puts an unprecedented amount of socio-economic information directly into the hands of users. Users who are familiar with data in spreadsheet format can look up statistics or perform simple analyses right away. Other users will find it is trove of information with 6.6 million data points and lots of potential for research, discovery and visualizations, without the hassle of downloading data from census websites followed by hours upon hours of data management.

Beyond ease of access, many researchers currently rely on census tract data to analyze and visualize inequalities across Boston neighborhoods. However, census tracts are not especially meaningful geographic units. Moreover, census tract data from the ACS – now the primary source of public, local data on socioeconomic characteristics – is notoriously imprecise, especially if the population and underlying sample are broken down by race/ethnicity.

NERD Boston addresses both problems: By pooling observations across census tracts and aggregating up to the neighborhood level, we improve statistical precision considerably. In fact, we improve precision enough that we can report statistics for smaller population groups, like people of different races/ethnicities, with acceptable margins of error. This gives us estimates for geographic units that have a deeply-rooted social and historical meaning for residents and are commonly used in many research and planning contexts.

Most importantly, perhaps, is that NERD allows users to obtain a fuller view of the large racial and ethnic inequities that exist within and across Boston neighborhoods. Before NERD, users seeking socio-economic information on different racial/ethnic groups in Boston neighborhoods needed to do the extremely time- and resource-consuming work we just described. Existing data profiles and reports reflect these barriers, and have focused on aggregate data that is more easily obtained: poverty rates for the entire city, perhaps broken down by race/ethnicity but not by neighborhood, or neighborhood poverty rates for the total population but not for each racial/ethnic group. In a spatially segregated city that is becoming more diverse, these aggregate statistics often hide more than they reveal.


It sounds like this will help reveal some more ‘hidden’ inequities in our city. Can you share some data from NERD Boston?

Let’s look at poverty rates, for example. The most recent city-wide data (2012-2016) show the poverty rate for the total population to be 21 percent, 13 percent among white, and 25 percent among African-American households. However, as the graphic below shows, this city-wide average conceals huge inequalities across neighborhoods. The poverty rate among African-American household ranges from 10 percent in Hyde Park to 56 percent in Charlestown. Reporting neighborhood-wide statistics can be equally misleading. For example, neighborhood-wide poverty rates for Charlestown and South Boston are 19 percent and 17 percent, respectively, but these population-wide numbers conceal huge differences between households by race. In Charlestown, poverty rates are 56 percent for African-American households versus six percent for white households. In South Boston, African-American poverty is 55 percent versus eight percent for white households. In relative terms, poverty rates for African-American households are three times as high as poverty rates for white households in the South End and Jamaica Plain, seven times as high in South Boston, and nine times as high in Charlestown.

We have observed similar patterns of inequality across other indicators. Additional sample data analyses and figures are included in our User Guide. However, these examples above do not consider how these inequities have changed in recent years, or how neighborhoods fare relative to the other municipalities in the Boston metro area.

Poverty Rates among African-Americans and White Households, 2012-2016

Click to enlarge image

Poverty Rates among African-American and White Households, 2012-2016
Notes: NERD-Boston, based on 2012-2016 American Community Survey data. Error bars are 90% confidence intervals.


This is really interesting. How can I get access to NERD Boston?

Thank you for asking (and reading this far). The full database is available via our website: We have also made available a small slice of the data in tableau visualizations, and here are links to the table of contents listing all available indicators and our User Guide that includes our motivation for building NERD, more sample data, and technical information. Lastly, we have published the crosswalk we used to link tracts to neighborhoods as defined in the User Guide and a neighborhood shapefile that we used to make the maps on Tableau.