Massachusetts Racial and Gender Profiling Final ReportExecutive Summary May 4, 2004
Northeastern University Institute on Race and Justice The goal of this report is to answer the mandate of Chapter 228 of the Acts of 2000 to identify and provide to the Secretary of Public Safety a listing of state police units or municipalities that appear to have engaged in racial or gender profiling. Using data on traffic citations and written warnings collected by the Massachusetts Registry of Motor Vehicles, this study examines the existence of racial and gender disparities in approximately 1.6 million traffic citations issued between April 1, 2001 and June 30, 2003. The citations analyzed in this report were received from the Massachusetts State Police, 340 municipal police departments and 25 other special police units. The intent of the Massachusetts legislation was to identify communities with racial or gender disparities which may be suggestive of racial profiling and to require those communities to collect additional information on traffic enforcement to determine if the disparities identified could be explained by factors other than race or gender. This model for dealing with allegations of racial and gender profiling is unique nationally and was intended to minimize the data collection burden on those departments where no indication of racial or gender profiling exists. It is important to note at the outset that research on racial profiling in traffic enforcement is a relatively new area of inquiry. Although numerous studies have begun to address questions of differential treatment in traffic stops, no absolute consensus exists about the best way to determine disparities. Racial disparities in citations can result from a number of factors that social scientists are just beginning to understand. Bias on the part of an individual officer is only one of several possible explanations for disparities in citations. Although the Massachusetts data does not allow for definitive findings of racial profiling on the part of individual officers or within a department, it does identify those jurisdictions with substantial racial or gender disparities in traffic citations. Methodology Although there are limits to the types of questions that traffic citation data can answer, this study addresses four different questions that commonly arise in public concern over racial profiling to determine if racial or gender disparities existed in any of the 366 jurisdictions who submitted data to the Registry of Motor Vehicles. 1. Are non-white drivers who are residents in a community cited more often than their representation in the residential population would predict? 2. Are non-white drivers overall cited more often their representation in the population of people driving on the roadways would predict? 3. Once stopped, are non-white drivers more likely to be subject to a search than white drivers? 4. Once stopped, are non-white drivers more likely to receive a citation than white drivers? In addition to examining the experiences of non-white drivers as a group, we have conducted additional analysis on each of the above four questions to determine if particular racial groups (Black or Hispanic drivers) or certain race and gender groups (non-white males) are disproportionately cited and/or searched. For example the measure of citations issued to community residents includes an examination of: 1) citations issued to non-white residents, 2) citations issued to Black residents, 3) citations issued to Hispanic residents, and 4) citations issued to non-white male residents. These subcategories are included to help departments focus their efforts toward the particular groups that are identified as having the greatest disparities. In all research on racial disparities in traffic enforcement, one of the most challenging issues defining an appropriate benchmark, or baseline, against which to compare the demographics of traffic citations. If for example, 15% of the citations in a particular community are given to Black residents, is that too many or too few? To answer that question we would want to know the proportion of traffic stops compared to an appropriate benchmark or base rate of those driving in that community. In Massachusetts the present study utilized two measures to determine the degree of disparity that exists in traffic citations across the state. First we compare the racial demographics of town residents who are cited against the residential population of that community. Second we compare the racial demographics of all traffic citations made by an agency (both citations of residents and citations of non-residents) to an estimate of the demographics of the driving population for that community. This approach, while more complex, utilizes the most reliable available comparison for each disparity measure. No external benchmark is needed to answer the final two questions about racially disparate treatment searches and written warnings. Instead, a measure of statistical significance is used to determine if non-white drivers are more likely than white drivers to be subject to a search or issued a written warning. Community and law enforcement feedback in analysis process During the analysis of the data for this report the Northeastern University team has been extremely fortunate to have worked with a dedicated group of police officials, community representatives and representatives of advocacy groups, legislators and representatives from the Executive Office of Public Safety and the Office of the Attorney General. This working group met monthly during the analysis process and provided extremely helpful advice and comment, all of which has made this report stronger. Also, with support from the Executive Office of Public Safety, Northeastern University implemented the most comprehensive comment process that has ever been conducted nationally in conjunction with a racial profiling report. Over the three month period between January and March of 2004, Northeastern led six regional community meetings and actively solicited feedback from community members and law enforcement about their reactions to the findings presented in a preliminary report which was released in January 2004. Many of the comments from law enforcement officials and community members have been integrated into the final report, allowing us to address many of the problems or limitations of the study which may not have been identified without such an extensive comment period. Major Findings The process of drawing conclusions about racial and gender disparities in 366 different law enforcement agencies does not allow for the in-depth analysis that can and should occur in a particular community. This report is intended to highlight the main areas of concern, offer interpretations of different types of disparities and hopefully serve as a springboard for more detailed analysis that can best be done at the community level between local police and members of their community. A summary of the major findings is as follows: 249 law enforcement agencies in Massachusetts have substantial disparities in at least one of the four categories of analysis used in this report (Table A). Breaking down these disparities into the four measurements that were used in the report we find: -- 141 law enforcement agencies in Massachusetts have racial disparities above the statewide median in citations given to resident drivers who were non-white, Black, Hispanic or non-white males (Table B). Statewide the average disparity between non-white residents cited and non-white residents in the census population was .06% with non-white residents being cited slightly more frequently than their representation in the residential population. Statewide the largest racial difference between citations to residents and residential population was for Black drivers (average disparity of 1.3%) and non-white male drivers (average disparity 2.2%). -- 201 law enforcement agencies in Massachusetts have racial disparities above the statewide median in citations given to non-white, Black, or Hispanic drivers (Table C). Statewide the average disparity between non-white drivers cited and non-whites in the driving population estimate was 2.6%. Very consistent disparities were found when Black citations (average disparity of 2.3%) and Hispanic citations (average disparity 2.2%) were compared to the driving population estimate. -- Out of the 87 communities where a sufficient number of searches were conducted for analysis, 40 law enforcement agencies in Massachusetts are statistically significantly more likely to search non-white, Black, Hispanic or non-white male drivers compared to white drivers (Table D). Although searches following a traffic citation are a relatively rare event in Massachusetts (only 1.3% of all traffic citations resulted in a non-inventory search of the motorist or their vehicle statewide), non-white drivers were more likely to be searched following a citation than white drivers (1.3% of white drivers searched compared to 1.8% of non-white drivers). -- Out of a sample of 142 communities where an analysis of written warnings could be conducted, 83 agencies are statistically significantly more likely to give a citation to a non-white, Black, Hispanic or non-white male driver compared to a white driver (Table E). Statewide 72% of non-white drivers receive citations compared to 65.9% of white drivers. This suggests that in some communities in Massachusetts officers may be more likely to use their discretion to give written warnings to white drivers rather than to non-white drivers. Overall males were more likely to be cited than their representation in either the residential or the driving population estimate. Males were uniformly more likely to be subject to a search and to be cited than women. These findings were consistent across virtually all communities in Massachusetts. This report finds no indication that female drivers, in the aggregate, are more likely to be stopped, cited or subject to a search than their male counterparts. In fact, quite the opposite appears to be the case. Issues to Consider Based on the findings of the Massachusetts study, national research on the issue, and recommendations from the Working Group and other stakeholders we offer the following set of issues to consider which provide guidance to those who must begin to formulate policies based on the findings in this report. Nationwide over 6,000 law enforcement agencies are collecting data on all traffic stops. Traffic stop data collection has been recommended by numerous national professional police organizations as a necessary component of professional policing. As a part of good police practices, all law enforcement agencies in Massachusetts should establish a system to collect and monitor data on all traffic stop activity. Following national models for traffic stop data collection, a uniform set of data elements to be collected on all stops should be identified. It is important that any new data collection system include information on officer identification and the location of the stop in the required data collection elements. Additionally a specific timetable for data collection, auditing and reporting should be established. All local police agencies should begin or continue a conversation with members of their community about the existence of disparities in traffic stops, the goals of traffic enforcement and strategies to monitor and reduce such disparities.
Table A: Any Disparity in the Four Summary Measures (N=249)
> Table B: Resident Citation Disparities (N=141 Above Statewide Positive Median)Abington Adams Amherst Andover Arlington Ashland Athol Avon Barnstable Belchertown Berkley Berlin Beverly Boston Bourne Boxborough Braintree Brockton Brookline Buckland Cambridge Canton Chatham Chelsea Cheshire Chicopee Chilmark Clarksburg Clinton Cohasset Dedham Dennis Dudley E. Brookfield Easton Edgartown Erving Everett Fall River Falmouth Fitchburg Foxborough Framingham Georgetown Gloucester Greenfield Hanson Harwich Hatfield Haverhill Holbrook Holyoke Hudson Ipswich Kingston Lancaster Lawrence Lee Leicester Leominster Lexington Littleton Longmeadow Lowell Lynn Malden Marion Marlborough Mashpee Mattapoisett Maynard Medford Medway Mendon Methuen Middleborough Milford Millbury Millis Milton Montague N. Adams Nantucket Needham New Bedford Northampton Northborough Northbridge Northfield Norton Norwood Oak Bluffs Oakham Orange Orleans Paxton Peabody Pelham Pittsfield Quincy Randolph Raynham Revere Rochester Rockland Royalston Salem Scituate Sharon Shrewsbury Shutesbury Somerville Southbridge Springfield State Police Stockbridge Stoneham Stoughton Sunderland Sutton Taunton Tisbury Townsend W. Bridgewater W. Springfield Waltham Ware Wareham Warwick Watertown Wayland Webster Wellfleet Wendell Westborough Weymouth Winchendon Winthrop Woburn Worcester Yarmouth Table C: Driving Population Estimate Disparity (N=201 Above Statewide Positive Median) Abington Acushnet Adams Amherst Andover Aquinnah Arlington Ashburnham Ashby Ashland Athol Auburn Avon Ayer Barnstable Belchertown Bellingham Belmont Berlin Bernardston Billerica Bolton Boston Bourne Boxborough Boylston Braintree Bridgewater Brockton Brookline Cambridge Canton Carver Charlton Chatham Chelmsford Cheshire Chester Chicopee Chilmark Clinton Dartmouth Dedham Dennis Douglas Dover Dracut Dudley Dunstable E. Bridgewater E. Brookfield E. Longmeadow Easton Edgartown Everett Fairhaven Falmouth Fitchburg Foxborough Framingham Freetown Georgetown Gill Goshen Granby Granville Groveland Hadley Hampden Hanson Harwich Hatfield Haverhill Heath Hingham Holbrook Holden Holliston Holyoke Hopedale Hopkinton Hudson Hull Huntington Kingston Lakeville Lancaster Lanesborough Lawrence Leicester Lenox Leominster Littleton Longmeadow Lowell Lunenburg Lynn Lynnfield Malden Marblehead Marion Marlborough Mashpee Mattapoisett Maynard Medford Medway Melrose Mendon Merrimac Methuen Milford Millbury Millville Milton Montague N. Adams N. Andover Nahant Nantucket Natick Needham New Bedford New Marlborough New Salem Northborough Northbridge Norton Oak Bluffs Oakham Orange Orleans Paxton Peabody Pepperell Pittsfield Quincy Randolph Raynham Rehoboth Revere Rochester Rockport Rowley S. Hadley Salem Salisbury Sandwich Saugus Seekonk Sharon Sherborn Shirley Shrewsbury Shutesbury Somerville Southborough Southbridge Springfield State Police Sterling Stoughton Stow Sturbridge Sudbury Sutton Swampscott Taunton Tewksbury Tisbury Topsfield Townsend Truro Tyngsborough Upton W. Bridgewater W. Brookfield W. Newbury W. Springfield W. Tisbury Walpole Waltham Ware Wareham Warwick Watertown Wayland Webster Wellesley Wenham Westborough Weston Westwood Whitman Wilbraham Wilmington Windsor Winthrop Worcester Wrentham Yarmouth Table D: Search Disparities (N=40 Positive Statistically Significant Differences) Amesbury Attleboro Auburn Barnstable Bedford Boston Bridgewater Brockton Brookline Cambridge Dunstable Fairhaven Fall River Framingham Haverhill Leicester Lowell Lynn Malden Mansfield N. Adams New Bedford Newton Northampton Peabody Quincy Salisbury Southborough Spencer Springfield State Police Stoughton Taunton Tewksbury Townsend W. Bridgewater Waltham Wareham Wilmington Worcester Table E: Warnings vs. Citations (N=83 Positive Statistically Significant Difference) Abington Amesbury Andover Arlington Auburn Barnstable Bellingham Belmont Berlin Billerica Blackstone Boston Braintree Bridgewater Brookline Cambridge Canton Chelsea Chicopee Dedham Dracut E. Bridgewater Everett Falmouth Fitchburg Framingham Franklin Holland Hudson Lawrence Leicester Leominster Lexington Lincoln Longmeadow Lynn Marlborough Mattapoisett Medway Methuen Milford N. Andover N. Attleboro Natick Needham New Bedford Newburyport Newton Northampton Norwood Palmer Peabody Pittsfield Plymouth Quincy Randolph Revere Rockland Shirley Shrewsbury Somerset Somerville Southborough Southwick Springfield Stoughton Sturbridge Sudbury Truro Upton W. Bridgewater W. Tisbury Wakefield Walpole Waltham Watertown Wellesley Westborough Westfield Weston Weymouth Worcester Yarmouth Table F: Measures of Disparity for All Jurisdictions Jurisdictions with Any Disparity (N=249) 1. Jurisdictions Above Threshold on All Four Measures (15) Barnstable Boston Brookline Cambridge Framingham Leicester Lynn New Bedford Peabody Quincy Springfield Stoughton West Bridgewater Waltham Worcester 2. Jurisdictions Above Threshold on Three Measures (42) Abington Andover Arlington Auburn Berlin Braintree Bridgewater Brockton Canton Chicopee Dedham Everett Falmouth Fitchburg Haverhill Hudson Lawrence Leominster Longmeadow Lowell Malden Marlborough Mattapoisett Medway Methuen Milford N. Adams Needham Northampton Pittsfield Randolph Revere Shrewsbury Somerville Southborough State Police Taunton Townsend Wareham Watertown Westborough Yarmouth 3. Jurisdictions Above Threshold on Two Measures (87) Adams Amesbury Amherst Ashland Athol Avon Belchertown Bellingham Belmont Billerica Bourne Boxborough Chatham Chelsea Cheshire Chilmark Clinton Dennis Dracut Dudley Dunstable East Bridgewater East Brookfield Easton Edgartown Fairhaven Fall River Foxborough Georgetown Hanson Harwich Hatfield Holbrook Holyoke Kingston Lancaster Lexington Littleton Marion Mashpee Maynard Medford Mendon Millbury Milton Montague N. Andover Nantucket Natick Newton Northbridge Northborough Norton Norwood Oak Bluffs Oakham Orange Orleans Paxton Raynham Rochester Rockland Salem Salisbury Sharon Shirley Shutesbury Southbridge Sturbridge Sudbury Sutton Tewksbury Tisbury Truro Upton West Springfield West Tisbury Walpole Ware Warwick Wayland Webster Wellesley Weston Weymouth Wilmington Winthrop 4. Jurisdictions Above Threshold on One Measure (105) Acushnet Aquinnah Ashburnham Ashby Attleboro Ayer Bedford Berkley Bernardston Beverly Blackstone Bolton Boylston Buckland Carver Charlton Chelmsford Chester Clarksburg Cohasset Dartmouth Douglas Dover E. Longmeadow Erving Franklin Freetown Gill Gloucester Goshen Granby Granville Greenfield Groveland Hadley Hampden Heath Hingham Holden Holland Holliston Hopedale Hopkinton Hull Huntington Ipswich Lakeville Lanesborough Lee Lenox Lincoln Lunenburg Lynnfield Mansfield Marblehead Melrose Merrimac Middleborough Millis Millville N. Attleboro Nahant New Marlborough New Salem Newburyport Northfield Palmer Pelham Pepperell Plymouth Rehoboth Rockport Rowley Royalston S. Hadley Sandwich Saugus Scituate Seekonk Sherborn Somerset Southwick Spencer Sterling Stockbridge Stoneham Stow Sunderland Swampscott Topsfield Tyngsborough W. Brookfield W. Newbury Wakefield Wellfleet Wendell Wenham Westfield Westwood Whitman Wilbraham Winchendon Windsor Woburn Wrentham Jurisdictions with No Disparity (92) Acton Agawam Ashfield Barre Becket Blandford Boxford Brewster Brimfield Brookfield Burlington Carlisle Charlemont Chesterfield Colrain Concord Conway Cummington Dalton Danvers Deerfield Dighton Duxbury Eastham Easthampton Egremont Essex Gardner Grafton Groton Gt. Barrington Halifax Hamilton Hanover Hardwick Harvard Hinsdale Hubbardston Leverett Leyden Ludlow Manchester Marshfield Medfield Middleton Monroe Monson Monterey N. Brookfield N. Reading New Braintree Newbury Norwell Norfolk Otis Oxford Pembroke Peru Petersham Phillipston Plainfield Plainville Plympton Princeton Provincetown Reading Rowe Rutland Sandisfield Savoy Sheffield Shelburne Southampton Swansea Templeton Tolland Tyringham Uxbridge W. Boylston W. Stockbridge Wales Warren Washington Westford Westhampton Westminster Westport Whately Williamsburg Williamstown Winchester Worthington Jurisdictions for Which Disparity Could Not Be Measured (N=25) AMTRAK B&M Railroad Bridgewater SC Bunker Hill CC CSX Environmental PD Fernald State School MA Maritime Police Massasoit CC MBTA Metro Police Lwr. Basin Metro Police Marine Mt Wachusett CC Registry of MV Somerville Housing Auth. State Fire Marshal Templeton Dev. Center Tufts University Univ Of Mass Amherst Univ Of Mass Boston Univ of Mass Datmouth Univ of Mass Lowell Univ of Mass Worcester Westfield SC Worcester Co. Sheriff Final Report Prepared by: Dr. Amy Farrell, Associate Director, Institute on Race and Justice Dean Jack McDevitt, Director, Institute on Race and Justice Lisa Bailey, Assistant Director of Community Outreach, Institute on Race and Justice Carsten Andresen, Research Associate, Center for Criminal Justice Policy Research Erica Pierce, Research Associate, Center for Criminal Justice Policy Research The Institute on Race and Justice at Northeastern University brings together academics from the College of Criminal Justice, School of Law, Department of Sociology and Anthropology, and African American Studies Program to examine questions on race and justice facing urban communities. Our primary goal is to conduct meaningful research in race and justice and provide leadership on issues of racial injustice. The mission of the Institute is founded on the premise that academic institutions can provide rigorous and objective analysis that can be used by members of the community and policy makers to make policy changes that advance the cause of social justice. This research model attempts to enhance scientific inquiries with the input and experiences of community stakeholders who struggle with issues of racial injustice. Combining traditional empirical questions with a community-based problem-solving model of research provides a national model for action-based research in the field of race and justice. For more information about the Institute or the findings in the Massachusetts Racial and Gender Profiling study please visit our website at www.irj.neu.edu. 
|
