Play Live Radio
Next Up:
0:00 0:00
Available On Air Stations
Donate today to give back in celebration of all that #PublicMediaGives. Your contribution will be matched $1 for $1.

Methodology: How We Analyzed the Data on Race in N.H.'s Criminal Justice System

Little research has been done in New Hampshire on race and the state's 10 county jails, which are run by county government.  No comprehensive data is available regarding these jails’ populations. But in our recent story, Racial Disparities Increase At Each Step Of N.H.'s Justice System, data provided to NHPR by the Valley Street Jail in Hillsborough County allows a glimpse into the details of who is incarcerated here, and why.

Here are the numbers behind our analysis.


This analysis primarily uses the 2014 Census Population Estimates for Hillsborough County, New Hampshire, and the United States.

Two or more races

The 19,074 people who identify as “two or more races” in New Hampshire present a wrinkle, however, as the jail and arrest data does not have a corresponding category. To accommodate this discrepancy, a distribution is created from the more detailed 2010 decennial Census, and then applied to the 2014 estimated total for individuals identifying with “two or more” races.  The counts are distributed evenly among mixed races. For example, if 10 individuals identify as both African American and white in 2010, the “white” category receives 5, and the “African American” category receives 5. If, hypothetically, there were 100 people counted in the “two or more” category in 2010, and 110 in 2014, these would be counted as 5.5 additional whites and 5.5 additional African Americans for the final totals.

This procedure is informed by methodology used by the Carsey School of Public Policy at UNH to calculate net migration numbers.


The U.S. Census, state and federal arrest data classify individuals as both a race (African American, white, Asian, etc) and an ethnicity (Hispanic or Non-Hispanic). However, county-wide and national jail data classifies Hispanics as an exclusive race alongside whites and blacks. To accommodate for this, as the Bureau of Justice Statistics does in its Jail Inmates at Midyear 2014 report, all Hispanics are designated as such, exclusively. Non-Hispanics are designated as white, black, or other.

Other races

Whites, blacks, and Hispanics are the focus of this report not because people of other races do not live in New Hampshire, but because members of other races are not a significant portion of the population of people arrested or jailed in the state.

For example, while people of other races make up 4.5 percent of Hillsborough County’s population, they are only about 1 percent of those arrested or jailed pretrial in the county.


Hillsborough County

The Hillsborough County Jail data consists of four jail population censuses provided by the Hillsborough County House of Corrections, taken at six-month intervals on Sept. 1, 2014, March 21, 2015, Sept. 1, 2015 and March 21, 2016. Aggregated, the dataset contains 2,079 records.  Duplicate records are not included in analysis except where individuals have more than one distinct booking date, or appear on multiple censuses. Of the 2,079 records, three have no race listed, and are also not included in analysis.

Average daily populations and comparative incarceration rates are calculated by averaging the four census dates. The analysis excludes individuals being held for other jurisdictions. Most of the analysis focuses on pretrial populations, to avoid the confounding factors of individuals sentenced to time served pretrial; and individuals sentenced to state prison. Both pretrial and sentenced populations were used in comparisons with state and national jail data, where pretrial data is not available.

Measuring disparity

The measure of racial disparity in arrests is adapted from the ACLU’s methodology for its investigation into low-level arrests in Minneapolis.

The measure of racial disparity in jail is calculated as the ratio of the black or Hispanic incarceration rate to the white incarceration rate.  To calculate the incarceration rate for blacks per 1,000, the total number of jailed black people is divided by the total black population in Hillsborough County, and then multiplied by 1,000. For example, on an average day, there are 5.8 black people in jail in Hillsborough County for every 1,000 black residents, and 1 white person for every 1,000 white residents.  5.8 is divided by 1 to determine that blacks are 5.8 times more likely to be in jail than whites in Hillsborough County.

This procedure allows for comparisons that take account of the fact that the population of blacks and whites in a given place differ; and that those population distributions differ between geographies.

A similar procedure is followed and described below, to determine disparities among arrest rates.

Nonviolent crimes

To understand who is behind bars for exclusively nonviolent crimes, a list of nonviolent offenses was created based on the Bureau of Justice Statistic’s definition: “Nonviolent crimes are defined as property, drug, and public order offenses which do not involve a threat of harm or an actual attack upon a victim. Typically, the most frequently identified nonviolent crimes involve drug trafficking, drug possession, burglary, and larceny.” The jail population database is then queried for individuals held only for those nonviolent offenses. 

Specific offenses

To understand for which offenses people of color are most disparately detained pretrial, racial distributions are calculated for the 15 most common charges. Those with the smallest proportion of white defendants are: criminal threatening, assault, resisting arrest, and disobeying an officer.  These offenses are often charged at the same time as other offenses. Drug crimes are the most common offenses for which people are detained pretrial. For that reason, disparities among drug offenses are also considered. While the sample size is considerable for drug and assault offenses, it is smaller for other disparately charged offenses. 

Of the four census days, the percentage of the population that is white ranged from 57 percent to 73 percent for resisting arrest charges, and from 45 percent to 64 percent for criminal threatening. The number of people held at a time for each of those charges ranged from 11 to 23 for resisting arrest, and from 13 to 22 for criminal threatening. The number of individuals charged with disobeying an officer, however, is too small to report: there are three or fewer people of color, and five or fewer whites jailed with that offense during any given census.

State and National Jail Data

National and state jail data is accessible from the Bureau of Justice Statistics’ Jail Inmates at Midyear 2014, an annual single-day census of the nation’s jails. New Hampshire jails had a 100 percent response rate in 2014.


Both county and state-wide arrest data are retrieved from the NH Department of Safety’s crime statistics interface, and consist of the number of offenders for both A and B offenses by ethnicity and race. 

For statewide data, only 2014 arrests are considered. In 2014, there were a total of 47,972 arrests recorded in the state. Two hundred twenty-six are missing race or ethnicity data, and are not included in analysis. According to the DOS, municipal police agencies reporting cover 93 percent of the population.

To increase accuracy with a smaller sample, data for Hillsborough County includes four years of arrests.  There are a total of 51,230 arrests recorded in Hillsborough County from 2011-2014, 197 of which are missing race or ethnicity data, and are not included in analysis.

Racial distributions and rates of arrest are nearly identical between the four-year average and 2014 data, however.  When comparing county-wide arrest data to state and national numbers, 2014 data is used.

As is explained above, Hispanics are designated as such exclusively.  Non-Hispanics are designated as white, black, or other.

Related Content

You make NHPR possible.

NHPR is nonprofit and independent. We rely on readers like you to support the local, national, and international coverage on this website. Your support makes this news available to everyone.

Give today. A monthly donation of $5 makes a real difference.