Decision science for housing and community development localized and evidence based responses to distressed housing

1 Introduction: Community-Based Organizations,Neighborhood-Level Development, and Decision Modeling 1 1.1 Challenges and Opportunities for Housing and CommunityDevelopment in the US, 1 1

DECISION SCIENCE FOR HOUSING AND COMMUNITY DEVELOPMENT

A complete list of the titles in this series appears at the end of this volume.

Operations Research and Management Science

DECISION SCIENCE FOR

HOUSING AND COMMUNITY DEVELOPMENT

Localized and Evidence-Based

Responses to Distressed Housing

and Blighted Communities

Published by John Wiley & Sons, Inc., Hoboken, New Jersey

Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or

by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com Requests to the Publisher for permission should

be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ

07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permissions Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of

merchantability or fitness for a particular purpose No warranty may be created or extended by sales representatives or written sales materials The advice and strategies contained herein may not be suitable for your situation You should consult with a professional where appropriate Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002.

Wiley also publishes its books in a variety of electronic formats Some content that appears in print may not be available in electronic formats For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Decision science for housing and community development : localized and evidence-based responses to distressed housing and blighted communities / Michael P Johnson, Jeffrey Keisler, Senay Solak, David Turcotte, Armagan Bayram, Rachel Bogardus Drew.

pages cm

Includes bibliographical references and index.

ISBN 978-1-118-97499-5 (cloth)

1 Community development–United States–Decision making 2 Urban renewal–United

States–Decision making 3 Housing rehabilitation–United States–Decision making I Title.

HN90.C6J64 2016

307.1 ′ 40973–dc23

2015014609 Cover image courtesy of Nancy Brammer/Getty and Dorann Weber/Getty

Typeset in 11/13pt TimesLTStd by SPi Global, Chennai, India

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

1 2016

1 Introduction: Community-Based Organizations,

Neighborhood-Level Development, and Decision Modeling 1

1.1 Challenges and Opportunities for Housing and CommunityDevelopment in the US, 1

1.2 Community Development in the United States, 6

1.3 Big Data, Analytics and Community Development, 9

1.4 The Foreclosure Crisis: Problem, Impacts, and Responses, 111.5 Community-Based Operations Research: A Novel Approach toSupport Local Development, 13

1.6 Why This Book Now?, 19

1.7 Book Description, 21

1.8 Conclusion, 24

SECTION 1 POLICY AND PRACTICE IN FORECLOSED

HOUSING AND COMMUNITY

2 Foreclosed Housing Crisis and Policy and

2.1 Roots of the Foreclosed Housing Crisis, 29

2.2 Impacts of the Crisis, 32

2.2.1 Foreclosure Rates, 33

2.2.2 Home Equity and Wealth, 34

2.2.3 Health, Education, and Household Mobility, 36

2.2.4 Disamenities and Spillover Effects, 37

2.2.5 Market-Level Consequences, 38

2.3 Responses to the Crisis, 39

2.4 Effectiveness of Foreclosure Responses, 41

2.5 Opportunities for Decision Modeling Responses to

the Foreclosed Housing Crisis, 43

3 Community Partners and Neighborhood Characteristics 45

3.4 Case 1: The Neighborhood Developers, 50

3.4.1 Organization Type and Mission, 50

3.4.2 Organization Service Area and Population, 55

3.4.3 Organization Engagement with Foreclosure Crisis, 553.4.4 Organization Technical Capacity and Familiarity with

Project’s Analytic Methods, 583.5 Case 2: Coalition for a Better Acre, 59

3.5.1 Organization Type and Mission, 59

3.5.2 Organization Service Area and Population

Demographics, 593.5.3 Organization Engagement with Foreclosure Crisis, 613.5.4 Organization Technical Capacity and Familiarity with

Project’s Analytic Methods, 623.6 Case 3: Codman Square Neighborhood Development

Corporation, 63

3.6.1 Organization Type and Mission, 63

CONTENTS vii

3.6.2 Organization Service Area and Population

Demographics, 633.6.3 Organization Engagement with Foreclosure Crisis, 643.6.4 Organization Technical Capacity and Familiarity with

Project’s Analytic Methods, 673.7 Case 4: Twin Cities Community Development Corporation, 673.7.1 Organization Type and Mission, 67

3.7.2 Organization Service Area and Population

Demographics, 683.7.3 Organization Engagement with Foreclosure Crisis, 683.7.4 Organization Technical Capacity and Familiarity with

Project’s Analytic Methods, 703.8 Case Contrast and Discussion, 71

3.8.1 Role of Community Partners, 71

3.8.2 Adaptation of Case Study Theory for Our Project, 733.9 Conclusion, 74

4 Analytic Approaches to Foreclosure Decision Modeling 75

4.5 Solution Design for Community Development using

Community-Based Operations Research, 102

4.6 Where Do We Go From Here?, 104

SECTION 2 VALUES, METRICS AND IMPACTS

5 Value-Focused Thinking: Defining, Structuring

and Using CDC Objectives in Decision Making 109

5.1 Introduction, 109

5.1.1 Overview, 109

5.1.2 Values and Objectives in Decisions, 109

5.1.3 Values and Objectives in Community-Based

Organization/CDC Decisions, 1105.1.4 Utility Functions and Decision Making, 111

5.1.5 Multiattribute Utility Functions, 112

5.1.6 Value-Focused Thinking, 114

5.1.7 VFT as Soft OR and Problem Structuring Method, 1155.1.8 The Resource Allocation Decision Frame, 115

5.1.9 Plan, 118

5.2 Methods, 118

5.2.1 Linear Additive Assumption, 118

5.2.2 Defining the Mathematical Model as

a Set of Linear Equations, 1195.2.3 Structuring, 120

5.5 Lessons for Applying VFT to CBOs, 151

6 Characteristics of Acquisition Opportunities:

6.3.1 Sets and Indexes, 159

6.3.2 Parameters and Functions, 160

6.3.3 Individual Resident Frame, 160

CONTENTS ix

7.4 The PVI Model, 180

7.4.1 The Foreclosure Process, 181

7.4.2 Modeling Foreclosure Phase Transitions with

a Markov Chain, 1827.4.3 Estimation of Proximate Property Value Impacts, 1847.5 Case Study: The Neighborhood Developers, 186

7.5.1 Data and Model Specifications, 186

7.5.2 Computational Results, 190

7.5.3 Clustering Effects, 191

7.6 Discussion, 196

7.7 Model Validity and Limitations, 199

7.7.1 Nonlinearities in Aggregate Impacts, 199

7.7.2 Representativeness of Data Sources, 200

7.7.3 Sensitivity to Transition Probabilities, 200

7.7.4 Impacts of Multiple Foreclosures, 200

7.7.5 Wider Range of Social Impacts, 201

7.7.6 Model Validity, 201

7.8 Conclusion, 202

8 Social Benefits of Decision Modeling for

8.1 Introduction, 207

8.2 CDC Practice in Foreclosed Housing Acquisition, 209

8.3 A Multiobjective Model of Foreclosed Housing

8.6 Conclusion and Next Steps, 244

9 Acquiring And Managing A Portfolio Of Properties 247

9.1 Introduction, 247

9.2 Dynamic Modeling of the Foreclosed Housing

Acquisition Process, 248

9.3 Model Formulation, 251

9.4 Policy Analysis Under Different Fund Accessibility Cases, 2539.4.1 Acquisition Policies Under No Fund Expiration, 2539.4.2 Acquisition Policies Under Fund Expiration, 2579.5 Case Study: Codman Square Neighborhood DevelopmentCorporation, 259

9.5.1 Data Description, 260

9.5.2 Implementation Under No Fund Expiration, 261

9.5.3 Implementation Under Fund Expiration, 265

10.3.2 FHAP with Simple Resource Allocation, 277

10.3.3 FHAP with Gradual Uncertainty Resolution, 28210.3.4 Model Variations and Extensions, 286

10.4 Case Study: Codman Square Neighborhood DevelopmentCorporation, 289

10.4.1 Data Description and Parameter Justification, 28910.4.2 Resource Allocations and Impacts of Model

Parameters, 29210.4.3 Policy Implications for CDCs, 303

11.2.1 Foreclosure Crisis and Responses, 308

11.2.2 Engagement with Community-Based

Organizations, 30811.2.3 Decision-Modeling Fundamentals:

Values and Attributes, 30911.2.4 Foreclosed Property Strategy Design Using

Decision Models, 31011.3 Research Insights, 312

11.4 Lessons Learned, 314

11.5 Community-Based Operations Research: A Reassessment, 31611.6 Research Extensions, 319

11.7 Conclusion, 320

CONTENTS xi

APPENDICES

B.1 Multiobjective Decision Making, 330

B.2 Multiattribute Decision Models, 333

This book represents the culmination of an effort to expand the horizons

of public sector operations research and management science to addresscritical problems in urban housing and community development It is based

on a belief that research that is empirical, problem driven, interdisciplinary,and mixed methods in nature can enable mission-driven, not-for-profitcommunity-based organizations (CBOs) improve upon what they dobest—solving problems to improve the quality of life in neighborhoodsfacing challenges of socioeconomic distress and limited technical andfinancial resources

Our work on this book originated with, synthesizes and expands upon

a multi-year, multi-phase research project to address neighborhood-leveleffects of the U.S foreclosed housing crisis At the time of the project’sorigin in 2008, when the worldwide Great Financial Crash and the housingmarket meltdown that was a proximate cause of the crash was peaking inintensity, it seemed that there was an opportunity to design decision modelsthat could speak directly to the needs, capacities, and challenges of CBOs,but through a conceptual framework—community-based operations research(CBOR)—that would allow for flexibility in methodological orientationand analytic methods In addition, this project offered the possibility of ascholarly response to the reality of community development that did notrely solely on the use of traditional methods in operations research andmanagement science that have had demonstrated success in other aspects ofpublic affairs such as transportation, public safety and emergency response,

logistics, and health services delivery Analytic methods used to addressthese important problems have tended to emphasize model complexity andanalytic sophistication beyond the resources of CBOs We believe thatthe problems in housing and economic development and infrastructuredesign that CBOs routinely address are particularly challenging: theyembody multiple competing objectives, multiple stakeholders, and multiplelimitations on process activities and resource availability These problemsmust be solved in a context of financial and political uncertainty and mustaccommodate planning horizons that vary from the very short (addressingimmediate responses to community concerns) to the very long (designingstrategy and initiatives to ensure the social and economic sustainability ofneighborhoods in uncertain environments).

By applying principles from CBOR (Johnson, 2012; Johnson andSmilowitz, 2007) and its UK-based antecedent, community operationalresearch (Midgley and Ochoa-Arias, 2004), as well as modeling and analyticmethods from diverse sources such as urban operations research (Larsonand Odoni, 2007), problem structuring methods (Rosenhead and Mingers,2001), and public sector operations research broadly considered (Pollock,Rothkopf, and Barnett, 1994), we hope to contribute to the field of OR/MS

a suite of successful decision modeling applications for local impact.This effort could in turn inspire researchers and practitioners who seek toaddress other difficult problems in the urban context in which the needs ofsocioeconomically diverse communities might have a direct influence on thechosen analytic approach

Since we began to address local aspects of the foreclosure crisis andhousing and community development more generally, our team has expandedfrom three (Johnson, Turcotte, and then-University of Massachusetts Bostondoctoral student Felicia Sullivan) to a team of seven (the six authors ofthis book plus then-University of Massachusetts Lowell master’s studentEmily Chaves), augmented by University of Massachusetts Boston doctoralstudent research assistants Sandeep Jani, Merritt Hughes, Alvine Sangang,and Omobukola (Buki) Usidame, and University of Massachusetts Bostondoctoral candidate and editorial assistant Alma Biba All of the participants

in this research enterprise share a commitment to using decision analytics toimprove operations of urban CBOs and outcomes for the residents served bythese organizations In particular, we wish to learn how CBOs can addressthe critical problem of foreclosed housing acquisition and redevelopment forcommunity stabilization and revitalization

Our work in this area has evolved to address issues of housing policy, munity development, policy analysis, and multiple fields within OR/MS Wehave produced models, methods, applications, and findings that offer CBOs a

com-PREFACE xv

rich menu of resources to help them better achieve their objectives We havefound that even small, resource-constrained and mission-driven organizationsroutinely solve decision problems that are rich and complex Moreover wehave learned that these solutions offer marginalized and economically disad-vantaged communities to opportunity to define their own futures and to makeprogress toward meeting basic needs for good housing, education, employ-ment opportunities, social and physical environments, and human and familyservices We have also found, however, that decision analytics and relateddisciplines offer substantial but largely heretofore untapped opportunities toassist individuals and the local organizations that represent and serve them toachieve even better outcomes

Though the community development corporations (CDCs) with whom wehave collaborated may have different levels of capacity to incorporate deci-sion models into their daily practice, we have learned that the entire OR/MStoolkit has illuminated different aspects of the foreclosed housing acquisitionand redevelopment decision problem in different ways, generating a wholesuite of insights We believe that our book’s findings represent for our com-munity partners and for the readers of this book a sense that the whole of thearray of insights is greater than the sum of their parts Our decision modelingefforts provide decision makers with a rich set of lenses, each with differentframes Is acquiring a property like a card play in an uncertain game of black-jack, or finding the missing piece of a puzzle? Is it like choosing a dishwasherfor a kitchen, or prescribing a treatment for a patient or, simply laying thenext brick in a pathway? It is all of the above, and the skilled decision makercan think of using these different frames to connect the formal model-basedresults with the real-world problems of implementation, community building,and community development

The structure and form of this book bear some explanation, especiallysince we have written it with multiple audiences in mind: operationsresearch/management science (which draws researchers and practitionersmostly from business, management, and engineering-related fields), as well

as urban and regional planning, community development, public policy,and public administration (and social science disciplines such as economicsand sociology that form the basis for these professional domains) We havedivided the core of the book into three sections The first, “Policy andPractice in Foreclosed Housing and Community Development,” puts ourresearch into the context of housing, especially the recent foreclosure crisis,the organizational characteristics and foreclosure response practices of ourcommunity partners, and finally multiple traditions in data and decisionanalytics that are relevant to the models and methods we use in the book.The second, “Values, Metrics and Impacts for Decision Modeling,” uses

principles of decision modeling, primarily decision theory and data analytics,

to describe ways in which we have identified and quantified values andobjectives, the basis of decision models that are relevant to our project.The third, “Prescriptive Analysis and Findings,” contains three contrastingprescriptive decision modeling applications for foreclosure response Read-ers trained in OR/MS may wonder why a whole section is needed to set

up our problem; readers trained in planning and policy may wonder if themathematics-oriented material in the last section is really relevant to them

We believe that this rich detail is essential to engaging fully with a newapplication in public sector operations research and management science, par-ticularly within a domain we call community-based operations research Thebook presents our fullest understanding of practices and methods necessary tomeet community-based partner organizations where they are It also provides

us with the opportunity to explore certain problems with which CBOs arequite familiar—but which offer opportunities for improved responses—andwhich differ in important ways from most applications in the OR/MS litera-ture Therefore, the book represents an effort to dive deeply into problemsand practices within the world of CBOs in order to develop findings andinsights that may enable them to better fulfill their missions and, simulta-neously, enrich multiple academic disciplines and professional domains.This book represents one of the very first attempts to apply a fullymultimethod, mixed-methods, and multidisciplinary approach, rooted inoperations research and management science, to the problems of CBOs,especially CDCs Our work demonstrates that the entire OR/MS approachfits within our conception of CBOR Through this book, we hope thatpractitioners, researchers, and students will be persuaded that our findings,and others like it to follow, hold great promise for nonprofit and governmentactors to judiciously apply decision and data analytics to better achievefundamental goals of economic opportunity, resilient communities andsocial change

With all the recent fuss about big data and smart cities, it is not surprising tosee a new book about decision sciences applied to housing and communitydevelopment The book does indeed use new data and analytics to examineurban planning and revitalization strategies However, much to my delight,the book is long on problem framing and articulating suitable objectives andindicators, without resorting to unnecessarily complex mathematical formu-lations Yes, there are some equations and the book does take advantage ofnewly available and spatially disaggregated data about land use, propertyvalues, and financially troubled properties Likewise, the book includes con-strained optimization formulations of property acquisition and developmentstrategies for community development corporations (CDCs) across their ser-vice areas, and dynamic programming formulations of bidding strategies thatindicate when a bird in the hand is likely to be better than what is left in thebush But the focus of the book is less on complex models and “optimal”

strategies per se and more on problem formulations that facilitate clear

think-ing and meanthink-ingful comparisons of plannthink-ing and policy alternatives Thiswork takes seriously the multidimensional nature of community developmentimpacts; the diverse goals and skill sets of local nonprofits; and the inherentuncertainties about funding availability, political support, and developmentoutcomes

It may be worth reflecting for a moment on why the use of decisionsciences is so much more developed in private-sector business settingsthan in public-sector domains such as urban planning and community

development During the past few decades, airline scheduling, networkrouting, online shopping and delivery, taxi hailing services, and many othersupply chain and logistics operations have greatly increased the sophis-tication of the data and algorithms they use to optimize their operations.One obvious, and often cited, reason for the difference is the bottom-lineprofitability focus of private business Such use of decision sciences requiressignificant investment in analysts, data, and information infrastructure.Where the return on investment is clear, and accrues to the same entitiesthat commit the investments, then it is easier to raise the funds and hold theinnovators accountable for the performance of the new systems.

Certainly, in some areas of urban service delivery, financing and ability are fairly well identified and some “smart city” efforts have indeedtapped new data streams and technologies to improve urban logistics Trafficsignaling, snowplow routing, and various online fees and payment systemsare notable examples In community redevelopment and many aspects ofurban planning, however, the opportunity to capitalize on “big data” is muchless clear These domains tend to involve “wicked problems1” that are oftenopen ended, multifaceted, and politically controversial Such problems havecomplex social choice dimensions for which there is little agreement aboutvalues, beliefs, and desirable trade-offs How much public funding should

account-be invested in revitalizing a neighborhood with high poverty rates? Cansuch a program be successful for a particular geography and populationwithout addressing broader social policy issues such as unemployment, jobtraining, family responsibility? Suppose, moreover, that a community-basedprogram is “successful” in increasing economic activity and reducingblight and poverty rates If residents are displaced and the neighborhood isgentrified, can the program still be considered a success? As Schon and Rein(1994) argued in their book, “Frame Reflection: Towards the Resolution ofIntractable Policy Controversies,” policy and plan development in such set-tings is often shaped by “naming and framing” strategies that use diagnosticmetaphors to build consensus about problem framing in a way that suggests aparticular policy and programmatic choice Solving problems in housing andcommunity development requires serious assessment of the social impacts

of new programs in ways that private-sector program design that may benefitfrom decision sciences usually do not consider in their business plans Anexample of this is the so-called “sharing economy”

In Decision Science for Housing and Community Development, Johnson

and his co-authors do not “solve” community development problems as much

the applicability of management science methods to urban planning problems that typically involve complex social choices.

FOREWORD xix

as they help professional planners and community-based organizations toframe practical problems about development options and resource alloca-tion in ways that can benefit from new data and decision science tools It isappropriate, albeit somewhat ironic, that the book focuses on examples whereCDCs seek to mitigate the adverse effects of the recent housing foreclosurecrisis In many respects, the scope of the foreclosure crisis was exacerbated bythe use of complex private-sector financial instruments that greatly expandedhousing loans and optimized bank profits, but also opened the door to fraudu-lent loans and greatly underestimated the resulting systemic risk The publicwas not well served by these private-sector applications of decision sciences,

so it would be fitting if decision science can offer some help to the localgovernments and community organizations who are stuck with cleaning upthe mess Of course, the authors recognize that real, sustainable solutions toproblems such as stabilization and revitalization of local housing markets ulti-mately require action at a higher level in the political economy than the CDCs,which are their focus in this book

What I particularly like about the book is the extent to which the lem framing portions of the decision science modeling are developed throughdetailed descriptions of the case study settings and careful articulation of thesteps involved in defining multiple objectives and constructing practical mea-sures of effectiveness An entire chapter (Chapter 5) explains Ralph Keeney’s

prob-“value-focused thinking” approach to defining objectives and walks the readerthrough two “real-world” examples in which the authors work with two CDCs

to help them articulate their thinking about foreclosure problems and tion strategies Two subsequent chapters (Chapters 6 and 7) examine two par-ticular objectives of property acquisition strategies in detail Chapter 6 focuses

mitiga-on “strategic value” in order to understand both how a foreclosure acquisitimitiga-onfits into a CDC’s broader mission and also the extent to which some proper-ties might have disproportionate impact on a neighborhood depending upontheir location and relationship to other properties Chapter 7 focuses on the

“property value” effects of foreclosure and the extent to which any particularforeclosure acquisition might reduce or eliminate any negative effects of adistressed property on property values across the neighborhood Since theseeffects can depend on the length and specific stages of a foreclosure process,

a Markov chain model is developed both to address the uncertainty of theeffects over time and to relate the estimated property value impact of a poten-tial acquisition to the specific status of the property when it is acquired by aCDC In both chapters, as is customary throughout the book, specific casesare examined in detail so that the reader can see how the models value actualproperties and allow one to be explicit about various trade-offs and sensitivi-ties, as well as aspects of the valuation that might be ignored or undervalued

In Chapter 8, the authors formulate and solve a simple bi-objective decisionmodel that integrates the findings of the previous two chapters in order toprovide tangible representations of strategy alternatives that trade off impactsassociated with property value and strategic value.

By the time the more complex models of foreclosure acquisition strategiesare developed in Chapters 9 and 10, the reader has a rich understanding ofthe context in which CDCs might bid for foreclosed properties as part of theirefforts to revitalize neighborhoods by investing in distressed properties Atthis point, the mathematical model is less of a black box and more of a short-hand way to capture the relationships among key measures under the (many)assumptions made by the authors as part of the modeling process In thisway, the model solutions are more readily seen as “optimal” for a somewhatsimplified problem and best utilized as quantitative measures of key rela-tionships, guidelines, and trade-offs that are too complex to sort out withoutcareful articulation of objectives, values, and real-world interdependencies.Finally, Chapter 11 takes advantage of this careful, case-rich development

of concepts, measures, and models to outline useful findings and ties regarding the decision science approaches to foreclosure response andcommunity development The authors use the term “community-based oper-ations research” (CBOR) to represent the analytic approach used throughoutthis book for neighborhood revitalization, including the problem formulationprocess and value-focused thinking

opportuni-In this age of big data and smart cities, we are still a long way fromsolving “wicked problems” such as community development and neigh-borhood revitalization as if they were more straightforward logisticsproblems associated with urban service delivery Nevertheless, there aremany opportunities to crank up the level of sophistication with which citiesand community-based organizations articulate and explore their urbanplanning options and revitalization strategies The spatial encoding andstandardization of parcel-level databases of land use, ownership, real estatevalue, and the natural and built environment are greatly improved during thepast few decades Geographic information system technologies and methodshave greatly enhanced the value of urban analytics because visualization oftrends and urban performance measures at block and building scales helpfit modeling and model results into a broader, multiparty discussion aboutoptions, trade-offs, impacts, and the like

As we begin to view the emerging urban information infrastructure as akey to accumulating and maintaining “city knowledge”2as a public resource,

city knowledge.

con-to follow all the models, the detailed explanations of value-focused thinkingand model formulation, using the detailed case studies of CDC foreclosureacquisition processes, are a great introduction to how urban planners can usedecision science methods effectively.

Joseph Ferreira, Jr.3

June, 2015

REFERENCES

Carrera, F., and Ferreira, J 2007 The Future of Spatial Data Infrastructures:

Capacity-Building for the Emergence of Municipal SDIs International Journal

of Spatial Data Infrastructures Research, 2: 54–73.

Churchman, C.W 1967 “Wicked Problems,” Guest Editorial Management Science,

14(4): B-141–B-146 Web: 10.1287/mnsc.14.4.B141.

Rittel, H and Webber, M 1973 Dilemmas in a General Theory of Planning Policy

Sciences, 4: 155–169 doi: 10.1007/bf01405730.

Schon, D and Rein, M 1994 Frame Reflection: Towards the Resolution of

Intractable Policy Controversies New York: Basic Books.

Planning, Massachusetts Institute of Technology, jf@mit.edu.

This book is based upon work supported by the following sources: NationalScience Foundation, Grant No 1024968, “Collaborative Proposal: DecisionModels for Foreclosed Housing Acquisition and Redevelopment”; Joseph

P Healey Grant Program, University of Massachusetts Boston, Grant

No 51216, “Decision Modeling for Foreclosed Housing Acquisition in

a Large Urban Area”; and Joseph P Healey Grant Program, University

of Massachusetts Amherst, Grant No P1FRG0000000109, “CentralizedDecision Making in Societal Response to Foreclosures.” This book hasits roots in research previously completed under the National ScienceFoundation Faculty Early Career Development (CAREER) Program,

“CAREER: Public-Sector Decision Modeling for Facility Location andService Delivery.”

The authors would like to thank their respective institutions and ments for their support of the research, teaching, service, and mentoringactivities associated with the development of this book: Department of PublicPolicy and Public Affairs, University of Massachusetts Boston (Johnsonand Drew); College of Management, University of Massachusetts Boston(Keisler); Isenberg School of Management, University of MassachusettsAmherst (Solak and Bayram); and Department of Economics, Center forCommunity Research and Engagement, and Institute for Housing Sustain-ability, University of Massachusetts Lowell (Turcotte) We are grateful forthe expert support of research assistants Emily Chaves, Merritt Hughes,

depart-Sandeep Jani, Alvine Sangang, Felicia Sullivan, and Omobukola (Buki)Usidame and editorial assistant Alma H Biba.

Our research is inspired by the commitment and professionalism ofcommunity-based organizations engaged in housing and community devel-opment This book was made possible through the cooperation of ourcommunity partners: Coalition for a Better Acre (Lowell, MA), CodmanSquare Neighborhood Development Corporation (Boston, MA), The Neigh-borhood Developers (Chelsea and Revere, MA), and Twin Cities CommunityDevelopment Corporation (Fitchburg and Leominster, MA) We thank themfor their willingness to collaborate with us to uncover new ways to fulfilltheir missions

This book benefitted from the ongoing encouragement of James Cochran

We are grateful to Phillip L Clay and Joseph Ferreira for their comments andsuggestions The book has improved greatly from a review provided by ananonymous colleague

The authors are deeply grateful to their families and friends for their standing, encouragement, and patience

under-Michael thanks his co-authors for their outstanding contributions to thebook and the research and their professionalism and friendship that made thebook a reality

AUTHOR BIOGRAPHIES

Dr Armagan Bayramis an assistant professor in the Department of trial and Manufacturing Systems Engineering at University of Michigan –Dearborn She was previously a postdoctoral fellow in the Department

Indus-of Industrial Engineering and Management Sciences at NorthwesternUniversity She received her Ph.D in management science from the Uni-versity of Massachusetts Amherst and M.S and B.S degrees in industrialengineering from Istanbul Technical University

Dr Bayram’s research interests include the development of stochasticmodels and solution methods for capacity and resource allocation problems

Of particular interest are stochastic optimization and dynamic programmingmodels that involve nonprofit and healthcare applications

Dr Bayram’s honors and awards include several Best Paper Awards,including a Finalist Award in the 2013 INFORMS Doing Good with Good

OR Paper Competition and an Honorable Mention in 2013 INFORMSSection on Public Programs, Services and Needs Best Paper Award

Dr Rachel Bogardus Drewa freelance consultant working in the fields ofhousing markets and housing policy She received her Ph.D and M.S in pub-lic policy from the McCormack School of Policy and Global Studies at theUniversity of Massachusetts Boston and a B.A in economics from DartmouthCollege

Dr Drew’s expertise is in housing markets and policy, with emphasis

on the drivers of homeownership decisions for different populations Her

dissertation “Believing in the American Dream: How Beliefs InfluenceDecisions About Homeownership” is a multimethod and multidisciplinaryexamination of the sources and effect of commonly held assumptions aboutthe benefits of homeownership in the United States She has also publishedresearch on dynamics in the rental housing market, geographic patterns ofhousing relocations, and projections of housing demand.

Dr Michael P Johnson is associate professor in the Department of lic Policy and Public Affairs at the University of Massachusetts Boston Hereceived his Ph.D in operations research from Northwestern University in

Pub-1997 and B.S from Morehouse College in 1987

Dr Johnson’s research interests lie in data analytics and managementscience for housing, community development, and nonprofit service delivery.His methods enable nonprofit and public organizations, especially thoseserving disadvantaged and vulnerable populations, to develop programs andpolicies that jointly optimize economic efficiency, public welfare, and socialequity Current research projects include acquisition and redevelopment ofdistressed properties, resource allocation and urban planning for munic-ipal shrinkage and infrastructure redesign, and analytics and data needsassessment for community-based organizations

Dr Johnson’s work has appeared in a variety of journals, edited volumes,

and conference proceedings He is the editor of Community-Based tions Research: Decision Modeling for Local Impact and Diverse Populations

Opera-(Springer, 2012)

Dr Jeffrey M Keisler(B.S Computer Science and Mathematics, Wisconsin;M.B.A., University of Chicago; S.M Engineering Sciences, Harvard; Ph.D.Decision Sciences, Harvard) is professor in the Management Science andInformation Systems Department at the University of Massachusetts Boston

He has served as president of the INFORMS Decision Analysis Society and

as president of the Decision Analysis and Risk Specialty Group of the ety for Risk Analysis He is a fellow in the Society of Decision Professionals

Soci-He received the Decision Analysis Society’s Publication Award in 2013 Hiswork bridges theory and practice in decision modeling in private and publicdomains

Dr Keisler’s research interests include portfolio resource allocation, value

of information, multiattribute models, and the process of modeling itself ticularly within organizational contexts He was previously a decision analystwith Strategic Decisions Group, Argonne National Laboratory, and GeneralMotors He has published over 50 journal articles and book chapters

par-AUTHOR BIOGRAPHIES xxvii

Dr Senay Solakis an associate professor of operations management in theIsenberg School of Management at the University of Massachusetts Amherst

He holds Ph.D and M.S degrees in industrial engineering/operationsresearch from Georgia Institute of Technology and a B.S degree in electricalengineering from the US Naval Academy

Dr Solak’s main research interests involve portfolio management for nology and capital investment projects, with specific applications in the non-profit sector, and air transportation planning His methods focus on creatingvalue for organizations through better management of the uncertainty in suchproblems His research has been funded by NSF, NASA, FAA, and otherindustrial organizations, and his work has appeared in various top-tier journalsand conference proceedings

tech-Dr Solak’s honors include several Best Paper Awards, College OutstandingResearcher Award at the University of Massachusetts Amherst, the GeorgiaTech Supply Chain and Logistics Institute Global Logistics Scholar Award,and the US Naval Academy Distinctive Graduate Award

Dr David A Turcotteis research professor in the Department of Economics,senior program director at the Center for Community Research and Engage-ment, and editor of the Merrimack Valley Housing Report at the University

of Massachusetts Lowell He received his Sc.D from the University of sachusetts Lowell in work environment policy/pollution prevention/cleanerproduction and a M.S in community economic development from SouthernNew Hampshire University

Mas-Dr Turcotte’s research interests include regional housing needs, tive approaches to developing more affordable and sustainable housing, andassessment of housing intervention effectiveness in improving the health ofresidents Current research projects include in-home environmental interven-tion research with low-income children and elders and economic evaluation

innova-of bio-based alternative wind turbine blade manufacturing

Dr Turcotte is also an editorial review board member of Housing and Societyand a past president of the Coalition for a Better Acre, a communitydevelopment corporation in Lowell, MA

LIST OF FIGURES

Figure 1.1 The process of community-based operations research 17Figure 1.2 Characteristics of community-based organizations 19Figure 2.1 National homeownership rate, 1900–2013 31Figure 2.2 Percent of loans in foreclosure at end of quarter,

Figure 2.3 Percent change in median net wealth, 2007–2011 35Figure 3.1 Service area: The Neighborhood Developers 56Figure 3.2 Service area: Coalition for a Better Acre 60Figure 3.3 Service area: Codman Square Neighborhood Development

Figure 3.4 Service area: Twin Cities Community Development

Figure 4.1 Foreclosure recovery policy timeline 77Figure 4.2 Characteristics of nonprofit organizations 83Figure 4.3 Neighborhood typology for targeting funds 84Figure 4.4 Interactions between foreclosure risk and housing market

strength, community partner service areas 86

Figure 4.5 Market strength and foreclosure risk, Lowell, MA 88Figure 4.6 Market strength and foreclosure risk, Chelsea and

Figure 4.7 Market strength and foreclosure risk, Boston, MA 90Figure 4.8 Market strength and foreclosure risk,

Roxbury–Dorchester–Mattapan, Boston, MA 91Figure 4.9 Market strength and foreclosure risk, Fitchburg

Figure 4.10 Summary of foreclosure response potential by

Figure 4.12 Characteristics of nonprofit organizations relevant

Figure 5.1 Objectives network: Lowell simulated CDC 124Figure 5.2 Flip-chart notes, CSNDC value-focused thinking session

(a) Morning session (b) Afternoon session 130Figure 5.3 Transcript excerpt, CSNDC value-focused thinking

Figure 5.6 Strategy table, Twin Cities Community Development

Corporation, coded by organization purpose/role 141Figure 5.7 Strategy table, Twin Cities Community Development

Corporation, coded by type/status of project

Figure 6.1 Example neighborhood amenities, disamenities,

Figure 6.2 Map of Chelsea properties and amenities/disamenities

Created using ArcGIS 10 (ESRI, Inc, 2014) 164Figure 6.3 Strategic values with CDC frame, CDC-identified features,

LIST OF FIGURES xxxi

Figure 6.4 Strategic values with resident frame, all features,

Figure 7.1 Model of propagation of foreclosure impacts upon

Figure 7.3 Foreclosure state transition diagram 183Figure 7.4 Candidate and proximate properties 187Figure 7.5 Proximate property value discounts by stage and

Figure 8.1 Multiobjective solutions: objective space – constraint

Figure 8.2 Social value associated with solutions to the foreclosure

acquisition problem – constraint on the number of

Figure 8.3 Multiobjective solutions: decision space – constraint

on number of properties acquired, model 1 229Figure 8.4 Multiobjective solutions: decision space – constraint

on number of properties acquired, other models 231Figure 8.5 Multiobjective solutions: objective space – budget

Figure 8.6 Social value associated with solutions to the foreclosure

acquisition problem – budget constraint 238Figure 8.7 Multiobjective solutions: decision space – budget

Figure 9.1 (a) The change in the expected total PVI as a function

of accessible funds for different overbid rates under

no fund expiration (b) The change in the marginal

value of accessible funds under no fund expiration 262

Figure 9.2 (a) The change in the optimal PVI thresholds as a function

of available funds for different overbid rates under

no fund expiration (b) The change in expected total

PVI as a function of overbid rate for different funding

Figure 9.3 (a) The change in expected total PVI over time for different

funding levels under fund expiration (b) The change in themarginal value of accessible funds over time under

Figure 9.4 (a) Optimal PVI thresholds over time for an average

availability rate of 2.5 properties/week (b) Optimal PVI

thresholds over time for an average availability rate of

5 properties/week (c) The change in critical fund level

over time for different availability rates 268Figure 9.5 (a) The change in expected total PVI under fund expiration

(b) The change in optimal PVI thresholds under fund

Figure 10.1 The general decision process for the strategic foreclosed

Figure 10.2 Investment dependent social return function modeling the

synergistic effects of property acquisitions in a given

Figure 10.3 Categorization of CDC’s service area based on distinct

geographical regions Sample foreclosed property

availability information for each region and property

Figure 10.4 Change in optimal resource allocations and objective

function value over different budget levels 293Figure 10.5 Change in optimal resource allocations and objective

function value over different values of parameters

Figure 10.6 Pareto curves of financial and nonfinancial objectives

Figure 10.7 Pareto curves of equity and utility objectives

LIST OF FIGURES xxxiii

Figure 10.8 Trade-off graphs for equity objectives of base models of

Figure 10.9 Trade-off graphs for utility objectives of base models of

Figure A.1 Nondominated region and status quo point 324Figure A.2 Pareto frontier and potential Pareto frontier 325Figure A.3 Pareto frontier and indifference curves 327Figure B.1 Decision tree for development application 336

LIST OF TABLES

Table 3.1 Community Characteristics: The Neighborhood Developers

Table 3.2 Community Characteristics: Twin Cities Community

Development Corporation and Codman Square

Table 5.1 Example of Calculation of Scores for Objectives at Bottom

(Decision) Level of Hierarchy: Lowell Simulated CDC 128Table 5.2 Sensitivity Test Results: Lowell Simulated CDC 129Table 5.3 Sensitivity Test Results: Codman Square Neighborhood

Table 5.4 Common and Contingent Objectives, All Cases 144Table 5.5 Drivers of Commonalities in Objectives, All Cases 147Table 6.1 Example Distances between Candidate Properties and

Table 6.3 Description of Features (Amenities and Disamenities)

Table 6.4 Amenity and Disamenity Weight Specifications 167

Table 6.5 Strategic Value Results 167Table 6.6 Correlations of Strategic Value Outputs 169Table 6.7 Average Strategic Value of Purchased Versus

Table 7.1 Summary Statistics on Candidate Foreclosed Properties 186Table 7.2 Transition Probabilities between Foreclosure Stages 188Table 7.3 Summary Statistics on Proximate Property Value Impacts 192Table 7.4 Proximate Property Value Impacts by Property Type and

Table 7.5 Characteristics of Proximate Foreclosed Units to Given

Acquisition Candidates, by Distance Band 193Table 7.6 Discounting Factors Associated with Foreclosed

Table 7.7 Estimated Clustering Effects, Foreclosed Acquisition

Table 8.1 Strategic Values, Property Values, and Assessed Values

for Foreclosed Housing Acquisition Candidates 216Table 8.2 Correlations between Input Parameters 219Table 8.3 Trade-Off Values: Constraint on Number of Properties

Table 8.4 Range of Objective Function Values, Both Models 233Table 8.5 Trade-Off Values: Budget Constraint 237Table 10.1 Sample Data Representing Possible Stochastic

Parameter Realizations for FHAP-S Case 2 × 2 292Table B.1 Probabilities of Events Associated with Development

Table B.2 Costs and Benefits of Various Development Application

1.1 CHALLENGES AND OPPORTUNITIES FOR HOUSING

AND COMMUNITY DEVELOPMENT IN THE US

Community development in the United States is a complex process thathas historically centered on meeting the diverse needs of low-income,low-wealth, and otherwise disadvantaged people and places for improvedshelter, education, employment, and health By doing so, communitydevelopment professionals support social and economic integration andthe alignment of capital with justice (Pinsky, 2012) This book represents

an attempt to apply current knowledge in decision science, particularly anemerging area called community-based operations research (CBOR); todevelop new analytic models, mostly quantitative and prescriptive; and tosupport the work of community-based organizations (CBOs) whose activitiesare intended to enable economic prosperity and social justice

There are many successful examples of community development TheDudley Street Neighborhood Initiative in the Roxbury neighborhood ofBoston, founded in 1984, generated a network of local developers and

Decision Science for Housing and Community Development: Localized and Evidence-Based Responses

to Distressed Housing and Blighted Communities, First Edition Michael P Johnson, Jeffrey M Keisler, Senay Solak, David A Turcotte, Armagan Bayram and Rachel Bogardus Drew.

© 2016 John Wiley & Sons, Inc Published 2016 by John Wiley & Sons, Inc.

community organizations to perform large-scale housing redevelopment.

It has since branched out to address issues such as public safety, communityplanning, and environmental justice through the lens of community eco-nomic development, leadership development and collaboration, and youthopportunities and development (Dudley Street Neighborhood Initiative,2014) DSNI’s success has served as a model for comprehensive communitydevelopment initiatives across the United States (von Hoffman, 2012) ThePurpose Built Communities program in Atlanta’s East Lake neighborhoodprovides affordable housing development, community engagement, andeducation and early learning programs (East Lake Foundation, 2014) PBC’sefforts in East Lake from 1995 to the present have been associated withdramatic declines in violent crime, improvements in housing quality, andimprovements in educational outcomes and have been replicated in eightcommunities across the country (Belsky and Fauth, 2012)

Since 1997, the Harlem Children’s Zone (HCZ) in New York City hasput the needs of children at the center of its efforts to provide comprehen-sive services to families These services include educational resources (char-ter schools, parenting workshops, college readiness programs), family andcommunity programs (family support services and one-stop-shop connec-tions to government resources, legal services, and tax preparation) and healthimprovement programs (nutrition education and facility-based recreation, fit-ness, and nutrition resources) (Harlem Children’s Zone, 2014) HCZ’s socialoutcomes, though limited in various ways and expensive to produce, serve

as a model for high-impact social investments (Belsky and Fauth, 2012).Community development initiatives such as the three presented here embodyprinciples of success including local initiative, support from diverse financialand governmental sources, and a focus on tangible results that can be scaled

up and replicated (Grogan, 2012)

However, the environment within which community development works isone of high social inequality and substantial barriers to social advancement.Two prominent areas of challenges are income and economic opportunityand affordable housing Recent figures from the U.S Census show that while9.8% of non-Hispanic whites live in poverty, 25.6% of Hispanics and 27.2%

of blacks live in poverty; similar disparities are seen for persons whose income

is 50% or less than the poverty rate Moreover, while children are 23.7% of thetotal U.S population, they make up 34.6% of persons in poverty and 35% ofAmericans living in deep poverty (NCLEJ, 2013) Accounting for householdtaxes and cash transfers, the relative poverty rate in the United States of 17% isexceeded only by OECD countries Mexico, Israel, and Chile (Krueger, 2012).According to a measure of equality called the Gini coefficient, the UnitedStates has the fourth most unequal distribution of disposable income among