Demos and the Prison Policy Initiative respectfully submit this testimony to make recommendations for the Special Joint Committee on Redistricting as it seeks to assess lessons learned after the 2010 Census and to set goals for the next Census redistricting process. The focus of this testimony is on one specific aspect of the 2011 redistricting process in Massachusetts – the issue of prison-based gerrymandering. 

As background, Brenda Wright is the Vice President for Legal Strategies at Demos, a national, non-profit, non-partisan research and policy organization, established in 2000, with offices in Boston, New York, and Washington, D.C. The Demos Democracy Program works to ensure high levels of voting and civic engagement, and supports reforms to achieve a more inclusive and representative democracy. Brenda is an attorney with over 20 years of experience in redistricting, voting rights, and election reform, and was part of the legal team that represented community groups and organizations in the federal lawsuit challenging the Massachusetts House redistricting plan adopted after the 2000 Census.

Peter Wagner is Executive Director of the Prison Policy Initiative, a non-profit, non-partisan center in Easthampton which for the last decade has been the leading organization studying how the U.S. Census counts people in prison and working to quantify the policy and legal implications flowing from those technical decisions. 

Origin of the problem of prison-based gerrymandering 

The problem of prison-based gerrymandering stems from a long-standing flaw in the Census that counts incarcerated people as residents of the district where they are incarcerated rather than in their home district. Crediting incarcerated people to the census block that contains the prison, rather than to the home community that remains the legal residence of incarcerated persons for virtually all other purposes, results in a significant enhancement of the weight of a vote cast in districts with prisons at the expense of all other residents in all other districts in the state. It particularly distorts fair representation for communities of color which are disproportionately affected by high rates of incarceration. 

The rules that the Census Bureau uses for determining “residence” were adopted long before prison populations in the U.S. became large enough to have a significant effect on representation. The U.S. now has some 1.6 million persons in state and federal prisons, compared to only 200,000 as recently as 1970. In Massachusetts, the growth of the prison population in recent decades has been enormous.  As a percentage of population, Massachusetts now incarcerates three times as many people as it did as recently as 1980.  And prisons often are located in areas geographically and demographically removed from the home communities of incarcerated persons.  

Because of the rise in incarceration rates, the practice of allocating incarcerated persons to prison districts substantially skews redistricting. 

Analysis of new districts

According to our analysis of the new districts, four Massachusetts House districts, and three Senate districts meet minimum constitutional population requirements only by claiming prison populations as constituents. When Massachusetts relied on 2010 census data to calculate district sizes, the results appeared to conform to the 5% constitutionally-allowable deviations, but when the prison populations that were hidden in this data are taken into consideration, the actual population deviation falls outside of equal protection limits.

In the House, there are four districts that meet federal minimum population requirements only by claiming incarcerated people as residents. The 8th Plymouth, 37th Middlesex, 7th Middlesex, and 12th Worcester districts each have actual resident populations that are 5.6% to 7.4% smaller than the ideal district size:

    House District      



    8th Plymouth



    37th Middlesex



    7th Middlesex



    12th Worcester



We also note that the three smallest Senate districts appear to meet federal minimum population standards only because they include prison populations. The Second Suffolk, the Norfolk & Suffolk, and the Berkshire, Hampshire, Franklin & Hampden districts appear to be drawn right on the permissible line of having too little population to be districts, but they actually exceed that limit because those calculations relied on redistricting data that counted incarcerated people as residents of the prison. Each of those districts has an actual population that is 5.1-5.8% smaller than the ideal:

    Senate District



    Second Suffolk



    Norfolk & Suffolk



    Berkshire, Hampshire, Franklin & Hampden    




Solutions used in other states and the Massachusetts constitution

Massachusetts law makes it clear that incarcerated persons generally cannot claim a prison as their home residence. Until a ballot initiative amending the Massachusetts Constitution in 2000, incarcerated Massachusetts residents could vote, and the question of where they could vote was a frequent subject of litigation. In the late 1970s, people in prison were considered to be presumptive residents of their home addresses. Only in rare special situations could an incarcerated person argue that he intended to remain in the prison community permanently and thus could claim residency there; but by the early 1980s, even this narrow loophole was removed and all incarcerated voters in the state were required to vote as residents of their pre-incarceration homes. Cepelonis v. Commonwealth, 389 Mass. 930 (1983). Regardless of whether a specific prisoner was intending to never return home, state law barred him from adopting the prison address as his residence. In order to correct the distortions caused by counting incarcerated persons as residents of the prison, Maryland and New York amended their statutes to count incarcerated people at their home addresses for the 2010 round of redistricting. California and Delaware enacted similar laws that will take effect for the 2020 round of redistricting. Both the New York and Maryland laws have been upheld against legal challenges, with the Maryland law affirmed by the Supreme Court earlier this week. As a result, both states now have successfully enacted redistricting plans for the coming decade that are free of the distortions caused by prison-based gerrymandering. 

Restrictive language in the Massachusetts Constitution made it difficult for the Committee to adopt the ideal approach of counting incarcerated persons at their home address for purposes of the 2010 redistricting process. But, looking forward, Massachusetts can take several steps to allow our state to end the harmful practice of prison-based gerrymandering before the next Census in 2020.


Recommendations for Massachusetts

The basic principle of our democracy is that representation is distributed on the basis of population. Crediting incarcerated people to the wrong location has the unfortunate and undemocratic result of creating a system of “Representation Without Population.” Massachusetts should make its best efforts to ensure that the 2010 redistricting becomes the last redistricting cycle in which thousands of incarcerated persons are counted in the wrong place. We therefore urge you to document in your report how the Census Bureau’s ancient decision to count incarcerated people as residents of the prison location challenged your efforts to fairly and equitably draw legislative districts.

Your report should also plot a course by which Massachusetts can eliminate this problem before the next round of redistricting.

As an initial matter, the Legislature should enact a resolution calling on the Census Bureau to change its policy and provide states with population counts that allocate incarcerated persons to their home communities. The Census Bureau is actively soliciting feedback from states about the uses of redistricting data, and Massachusetts’ opinion on this matter would carry great weight. A sample resolution is attached.

In addition, the Legislature should work to amend the Massachusetts Constitution to allow the Legislature to make its own adjustments to the Census data, so that Massachusetts can correct this problem even if the Census Bureau refuses to act.

Thank you for considering these views. 


Peter Wagner 
Prison Policy Initiative 

Brenda Wright



More information is available:

Importing Constituents: Prisoners and Political Clout in Massachusetts, is a district-by-district analysis of prison-based gerrymandering in Massachusetts state legislative districts after the 2000 Census:

Summary of Prison-based gerrymandering issues in Massachusetts (part of a report on all 50 states):

Preventing Prison-Based Gerrymandering in Redistricting: What to Watch For is a guide for advocates who want to minimize the effects of prison-based gerrymandering in their state or community:



Urging the Census Bureau to provide redistricting data that counts prisoners in a manner consistent with the principles of “one person, one vote.”

WHEREAS, obtaining an accurate count of the population is so vital to representative democracy that the framers of the United States Constitution addressed the issue of the census and apportionment in the opening paragraphs of the Constitution; and

WHEREAS, the Massachusetts Constitution requires that federal census data be the basis for state redistricting; and 

WHEREAS, the Census Bureau currently has a policy of counting incarcerated people at the address of the correctional institution, even though for other legal purposes their home address remains their legal residence; and

WHEREAS, this flawed Census data results in distortions of the one-person, one-vote principle in drawing electoral districts in Massachusetts, diluting the representation of the majority of districts that do not contain prisons;

WHEREAS, the simplest solution to the conflict between federal constitutional requirements of “one person, one vote” and Massachusetts’ technical constitutional requirements of using the federal census is for the Census Bureau to publish redistricting data based on the location of an incarcerated person’s residence, not prison location; and

WHEREAS, the Census Bureau has already recognized the demand from states and counties for data that better reflects their actual populations, and has agreed to release data on prison populations to states in time for redistricting, enabling some states to individually adjust the population data used for redistricting; and 

WHEREAS, Public Law 94-171 requires the Census Bureau to work with states to provide geographically relevant data and the Census Bureau has been responsive to state’s data needs for the past three decades; now therefore be it

RESOLVED, that the Massachusetts General Court hereby urges the Census Bureau, in the next Census and thereafter, to provide states with redistricting data that counts incarcerated persons at their residential address, rather than the address of the correctional institution where they are temporarily located; and be it further

RESOLVED, that a copy of these resolutions be transmitted forthwith by the Clerk of the Senate to the Director of Census Bureau.