<%BANNER%>

Zoning Variance Administration in Practice

Permanent Link: http://ufdc.ufl.edu/UFE0043540/00001

Material Information

Title: Zoning Variance Administration in Practice Influencing Factors and Trends
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Zhao, Jun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: variance -- zoning
Design, Construction and Planning -- Dissertations, Academic -- UF
Genre: Design, Construction, and Planning Doctorate thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Over time, local governments have long sought ways to integrate flexibility within the zoning process. The zoning variance represents one of these tools. However, this tool is perceived as less than precise because vague legal standards for issuing variances, undue hardship, or practical difficulties allow local decision-making bodies to exercise significant discretion. The author conducted a substantial review on zoning variance which covered definition of zoning variance, types of variance, conditions to grant variance, administrative body and criticism of zoning variance; through careful examination of these issues, we seek to answer the following questions: According to the literature, which factors affect the decision making of zoning administrative bodies? Moreover, how were these factors determined, and to what extent do they affect the final decision? Are these decisions related to the economic and social characteristics of communities? What are the spatial distributions of variance applications and decisions and how do these distributions reflect the correlation between the decisions and economic and social characters? How are the decisions of variances represented in the time dimension? Are the decisions consistent through time? The purpose of this study was to understand those factors which influence the decisions of administrative boards in the approval or denial of variances. Four hypotheses in the literature about the decision-making of administrative bodies were proposed by the author. The first of these is: Hypothesis 1: granting of variance applications is significantly higher than denial. This is a general view about the Board of Zoning Adjustment (BZA)'s decision. The other three hypotheses deal with the factors which might affect the BZA's decision. Hypothesis 2: opponents from affected neighborhoods influence the zoning administrative board's decision making; Hypothesis 3: type of variance application affects the BZA's decision making; Hypothesis 4: the zoning administrative board does not place significant weight on suggestions from other public agencies in its decision making. To assist in answering these research questions and testing the four hypotheses statistically, a case study was conducted. The author collected and compiled 2,140 variance decisions made by the Board of Zoning Adjustment in Washington D.C. from 1980 to 2009. The first hypothesis was tested by Binomial Test. Further, a simplified binary response model was developed to test the other three hypotheses and examine to what degree those factors affect the BZA's decision. Then the model considered additional variables that might affect the BZA decisions. According to the results of statistical tests and the binary response model, the author came to the conclusion that opponents from affected neighborhoods, type of variance, and suggestions from Office of Planning and Advisory Neighborhood Commissions did affect the BZA's decisions significantly. Except for the Office of Planning (OP) and the Advisory Neighborhood Commissions (ANCs) suggestions, other public agency inputs were not given significant weight by the BZA in Washington D.C. It was also found that land value influenced in the BZA decisions. The higher the land value, the lower the probability that the BZA would deny the application. In order to test a cluster of variance applications and decisions, the author applied Hot Spot analysis in ArcGIS to demonstrate the regions where variance applications and decisions were highly concentrated. The maps revealed that clusters of variance applications and decisions did exist in Washington D.C.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jun Zhao.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Jourdan, Dawn.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043540:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043540/00001

Material Information

Title: Zoning Variance Administration in Practice Influencing Factors and Trends
Physical Description: 1 online resource (183 p.)
Language: english
Creator: Zhao, Jun
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: variance -- zoning
Design, Construction and Planning -- Dissertations, Academic -- UF
Genre: Design, Construction, and Planning Doctorate thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Over time, local governments have long sought ways to integrate flexibility within the zoning process. The zoning variance represents one of these tools. However, this tool is perceived as less than precise because vague legal standards for issuing variances, undue hardship, or practical difficulties allow local decision-making bodies to exercise significant discretion. The author conducted a substantial review on zoning variance which covered definition of zoning variance, types of variance, conditions to grant variance, administrative body and criticism of zoning variance; through careful examination of these issues, we seek to answer the following questions: According to the literature, which factors affect the decision making of zoning administrative bodies? Moreover, how were these factors determined, and to what extent do they affect the final decision? Are these decisions related to the economic and social characteristics of communities? What are the spatial distributions of variance applications and decisions and how do these distributions reflect the correlation between the decisions and economic and social characters? How are the decisions of variances represented in the time dimension? Are the decisions consistent through time? The purpose of this study was to understand those factors which influence the decisions of administrative boards in the approval or denial of variances. Four hypotheses in the literature about the decision-making of administrative bodies were proposed by the author. The first of these is: Hypothesis 1: granting of variance applications is significantly higher than denial. This is a general view about the Board of Zoning Adjustment (BZA)'s decision. The other three hypotheses deal with the factors which might affect the BZA's decision. Hypothesis 2: opponents from affected neighborhoods influence the zoning administrative board's decision making; Hypothesis 3: type of variance application affects the BZA's decision making; Hypothesis 4: the zoning administrative board does not place significant weight on suggestions from other public agencies in its decision making. To assist in answering these research questions and testing the four hypotheses statistically, a case study was conducted. The author collected and compiled 2,140 variance decisions made by the Board of Zoning Adjustment in Washington D.C. from 1980 to 2009. The first hypothesis was tested by Binomial Test. Further, a simplified binary response model was developed to test the other three hypotheses and examine to what degree those factors affect the BZA's decision. Then the model considered additional variables that might affect the BZA decisions. According to the results of statistical tests and the binary response model, the author came to the conclusion that opponents from affected neighborhoods, type of variance, and suggestions from Office of Planning and Advisory Neighborhood Commissions did affect the BZA's decisions significantly. Except for the Office of Planning (OP) and the Advisory Neighborhood Commissions (ANCs) suggestions, other public agency inputs were not given significant weight by the BZA in Washington D.C. It was also found that land value influenced in the BZA decisions. The higher the land value, the lower the probability that the BZA would deny the application. In order to test a cluster of variance applications and decisions, the author applied Hot Spot analysis in ArcGIS to demonstrate the regions where variance applications and decisions were highly concentrated. The maps revealed that clusters of variance applications and decisions did exist in Washington D.C.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jun Zhao.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Jourdan, Dawn.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043540:00001


This item has the following downloads:


Full Text

PAGE 1

1 ZO N ING VARIANCE ADMINISTRATION IN PRACTICE: INFLUENCING FACTORS AND TRENDS By JUN ZHAO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Jun Zhao

PAGE 3

3 To my parents and my future family

PAGE 4

4 ACKNOWLEDGMENTS I am very fortunate to have the opportunity to study u rban and r egional p lanning at the University of Florida where I have been nurtured by my passionate, responsible and knowledgeable professors. Many thanks to my advisory committee chair Dr. Dawn Jourdan and my committee members Dr. Zhong Ren Peng, Dr. Michael Allan Wolf, and Dr. Jimmie Wayne Hinze. It has been such an honor having all of you in my committee. Dr. Dawn Jourdan, my energetic and in novative professor, dedicated much of her time to guide my research. In fact, t his dissertation could not have been accomplished without her guidance. Dr. Zhong Ren Pe ng always p rovided me with inspiration and techni cal support. He is my mentor who helped me to develop my academic career Dr. Michael Allan Wolf, an outstanding law school professor helped enhance my knowledge of land use law and gave me many suggestions from a leg al perspective Dr. Jimmie Wayne Hinze unreservedly imparted to me his knowledge on setting up a dissertation topic and doing research to write a dissertation. I sincerely appreciate everything my advisory committee has done for me. Special thanks to Dr. A nne R. Williamson for your consistent encouragement and support. You encouraged me to persevere and achieve my goal. I would like to express my heartfelt gratitude to you. I must also thank Dr. Shouyi Hao, who is my supervisor gree. You led me into the field of academy and let me know what makes an excellent professor. I would like to thank the Office of Planning, District of Columbia where my research inspiration came from. Special thanks to the Development Review and Zonin g Division where I did my unforgettable internship supervised by Mr. Travis Parker.

PAGE 5

5 I also owe a great debt to Dr. Vandana Baweja, Dr. Ilir Bejleri Dr. Christopher Silver, Marta Strambi Kramer, Juna Papajorgji, Dr. Russel Watkins, Dr. Joseli Macedo, and Dr. Jocelyn M. Widmer. I appreciate you all for giving me the opportunit y to work with you. Dr. Kristin Larsen, Dr. Ruth L. Steiner, and Dr. Paul D. Zwick deserve my gratitude for their willingness to offer me advice on research and job hunting. Thanks to my editor Lisa L Delacure who is an excellent editor in the Editorial Office at the University of Florida. Finally, my deepest thanks are extended to my beloved father and mother whose endless love and understanding enables me to finish seemingly impo ssible task s

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 A BSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 16 Statement of Problem and Research Questions ................................ ..................... 16 Research Outline ................................ ................................ ................................ .... 18 2 LITERATURE REVIEW ................................ ................................ .......................... 21 Background of Zoning ................................ ................................ ............................. 21 Zoning Variance ................................ ................................ ................................ ...... 29 Definition of Zoning Variance ................................ ................................ ........... 29 Type of Zoning Variance ................................ ................................ .................. 31 Disadvantages of permitting use varian ces ................................ ............ 33 Advantages of permitting use variances ................................ ................. 34 Conditions to Grant Zoning Variance ................................ ............................... 35 Administrative Body ................................ ................................ .......................... 41 Zoning Variance and Rezoning ................................ ................................ ........ 45 Criticism of Zoning Variance ................................ ................................ ............. 46 3 METHODOLOGY AND PROCEDURE ................................ ................................ ... 51 Research Design and Conceptual Framework ................................ ....................... 51 Hypot heses ................................ ................................ ................................ ............. 54 Approval vs. Denial ................................ ................................ .......................... 54 Opponents from Affected Neighborhoods ................................ ........................ 55 Area Variance vs. Use Variance ................................ ................................ ....... 57 Suggestions from Other Public Agencies ................................ ......................... 57 Binary Response Model ................................ ................................ .......................... 60 Introduction of Binary Response Model ................................ ............................ 60 Model Building ................................ ................................ ................................ .. 62 Case Study Method ................................ ................................ ................................ 63 Case Study ................................ ................................ ................................ ....... 63

PAGE 7

7 Study Area ................................ ................................ ................................ ........ 64 Data Collection and Data Processing ................................ ................................ ..... 64 4 CASE STUDY WASHINGTON, D.C. ................................ ................................ .... 70 Introduction of Washington, D.C. ................................ ................................ ............ 70 City Planning in Washington D.C. ................................ ................................ ........... 70 Zoning in Washington D.C. ................................ ................................ ..................... 73 History of Zoning ................................ ................................ .............................. 73 Zoning Regulations and Administration ................................ ............................ 74 Zoning Variance in Washington D.C. ................................ ................................ ...... 75 Regulation of Zoning Variance ................................ ................................ ......... 75 Zoning Variance Procedure ................................ ................................ .............. 80 5 ANALYSIS OF CASE STUDY FINDINGS ................................ .............................. 85 Zoning Variance Applications and Decisions A General View ............................. 85 Variance Applications ................................ ................................ ....................... 85 Decisions from the BZA ................................ ................................ .................... 86 Area Variance Approval ................................ ................................ .................... 87 Use Variance Approval ................................ ................................ ..................... 89 Office of P lanning Involvements ................................ ................................ ....... 91 ................................ ................................ .................. 92 Community Involvement ................................ ................................ ................... 94 Use of Properties Requested for Zoning Relief ................................ ................ 98 Wards in Washington D.C. ................................ ................................ ............. 100 Ward 1 ................................ ................................ ................................ .. 101 Ward 2 ................................ ................................ ................................ .. 102 Ward 3 ................................ ................................ ................................ .. 104 Ward 4 ................................ ................................ ................................ .. 105 Ward 5 ................................ ................................ ................................ .. 106 Ward 6 ................................ ................................ ................................ .. 107 Ward 7 ................................ ................................ ................................ .. 108 Ward 8 ................................ ................................ ................................ .. 109 Testing Hypothesis ................................ ................................ ............................... 110 Testing Hypothesis 1 ................................ ................................ ...................... 110 Testing Hypothesis 2 to 4 ................................ ................................ ............... 112 Format of model ................................ ................................ .................... 113 Model results and testing ................................ ................................ ...... 117 Assessment of t he model ................................ ................................ ...... 123 Further Investigation ................................ ................................ ............................. 125 Other Factors ................................ ................................ ................................ 125 Cluster ing of Zoning Variance ................................ ................................ ........ 126 6 CONCLUSIONS AND RECOMMENDATION FOR FURTHER RESEARCH ........ 153 Summary ................................ ................................ ................................ .............. 153

PAGE 8

8 Implications to Planning and Law ................................ ................................ ......... 159 Recommendation for Further Research ................................ ................................ 161 APPENDIX A CRUDE BINARY RESPONSE MODEL ................................ ................................ 165 B CATEGORIES OF VARIANCE REQUESTED ................................ ...................... 171 C HOT SPOT ANALYSIS ................................ ................................ ......................... 173 LIST OF REFERENCES ................................ ................................ ............................. 177 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 183

PAGE 9

9 LIST OF TABLES Table page 4 1 Overview of related z oning a gency r oles in Washington D.C. ............................ 82 5 1 Cross tabulation of a pproval and d enial r ate under different r ecommendations from the Office of Planning, the ANCs, an d the c ommunity ................................ ................................ ................................ ........ 129 5 2 Cross tabulation of a pproval r ate by p roperty u se and t ype of a pplicant .......... 129 5 3 Binomial t est ................................ ................................ ................................ ..... 129 5 4 Significance and p arameters of v ariable in b inary r esponse m odel .................. 129 5 5 2 Log l ikelihood s tatistic and the c oefficient of d etermination R ...................... 130 5 6 Hosmer and l emeshow t est ................................ ................................ .............. 130 5 7 Classification t able of b inary r esponse m odel ................................ .................. 130 5 8 Area under the c urve ................................ ................................ ........................ 130 A 1 Significance and p arameter of crude b inary r esponse m odel ........................... 168 A 2 Omnibus t ests of m odel c oefficients by s tep ................................ .................... 169 A 3 2 Log l ikelihood and R s quare by s tep ................................ ............................ 170 B 1 Categories of v ariance requested ................................ ................................ ..... 171

PAGE 10

10 LIST OF FIGURES Figure page 3 1 Conceptual f ramework of t his p aper ................................ ................................ ... 69 4 1 Map of Washington D.C. ................................ ................................ .................... 83 4 2 Zoning v ariance p rocedure in Washington D.C. ................................ ................. 84 5 1 Number of v ariance a pplications by t ype and y ear ................................ ........... 131 5 2 Approval r ate by t ype of v ariance and y ear ................................ ...................... 131 5 3 Area v ariance a pproval r ate by a pproval t ype in approved a pplications by every five year s from 1980 to 2009 ................................ ................................ .. 132 5 4 Use v ariance a pproval r ate by a pproval t ype in a pproved a ppli cations by e very f ive y ear s from 1980 to 2009 ................................ ................................ .. 132 5 5 T he final d ecisions from the BZA ................................ ................................ ...... 133 5 6 T he final d ecisions from the BZA by t ype of v ariance ................................ ....... 133 5 7 The Office of Planning i nvolvement by every five years ................................ ... 134 5 8 The ANCs i nvolvement by every five ye ars ................................ ...................... 134 5 9 The c ommunity i nvolvement by every five years ................................ .............. 135 5 10 Community o pposition r ates in c ases where the c ommunity was i nvolved by y ear ................................ ................................ ................................ .................. 135 5 11 Comparisons between d enial r ate, the Office of Planning o pposition r ate, the ANCs o pposition r ate and c ommunity o pposition r ate by y ear from 2000 to 2009 ................................ ................................ ................................ ................. 136 5 12 Use of p roperties requested for z oning r elief ................................ .................... 136 5 13 Approval r ate by p roperty u se ................................ ................................ ........... 137 5 14 Number of v ariance a pplications by p roperty u se by y ear ................................ 137 5 15 Distributions of v ariance a pplications from 1980 to 2009 by Ward ................... 138 5 16 Distribution of a pproved and d enied v ariance a pplications from 1980 to 2009 139 5 17 Distributions of v ariance a pplications and d ecisions in Ward 1 from 19 80 to 2009 ................................ ................................ ................................ ................. 140

PAGE 11

11 5 18 Distributions of v ariance a pplications and d ecisions in Ward 2 from 1980 to 2009 ................................ ................................ ................................ ................. 141 5 19 Distributions of v ariance a pplications and d ecisions in Ward 2 by d ecennial p eriods ................................ ................................ ................................ .............. 142 5 20 Distributions of v ariance a pplications and d ecisions in Ward 3 from 1980 to 2009 ................................ ................................ ................................ ................. 143 5 21 Distributions of v ariance a pplications and d ecisions in Ward 4 from 1980 to 2009 ................................ ................................ ................................ ................. 144 5 22 Distributions of v ariance a pplications and d ecisions in Ward 5 fr om 1980 to 2009 ................................ ................................ ................................ ................. 145 5 23 Distributions of v ariance a pplications and d ecisions in Ward 6 from 1980 to 2009 ................................ ................................ ................................ ................. 146 5 24 Distributions of v ariance a pplications and d ecisions in Ward 7 from 1980 to 2009 ................................ ................................ ................................ ................. 147 5 25 Distributions of v ariance a pplications and d ecisions in Ward 8 from 1980 to 2009 ................................ ................................ ................................ ................. 148 5 26 ROC c urve ................................ ................................ ................................ ........ 149 5 27 Map of v ariance a pplication hot s pots ................................ .............................. 150 5 28 Map of l ocations which has more than two v ariance a pplications ..................... 151 5 29 Map of approved v ariance c ases hot s pots ................................ ...................... 152 C 1 The statistic m odel of hot s po t a nalysis ................................ ............................ 175 C 2 Map of Z s core of aggregated v ariance a pplications ................................ ........ 176

PAGE 12

12 LIST OF ABBREVIATIONS ANCs Advisory Neighborhood Commissions BZA Board of Zoning Ad justment DCRA Department of Consumer & Regulatory Affairs FHA Fair Housing Act NCPC National Capital Park and Planning Commission OP Office of Planning OZ Office of Zoning PUD Planning Unit Development SSL Square, Suffix, Lot SSZEA Standard S tate Zoning Enabling Act Z C Zoning Commission ZEA Zoning Enabling Act

PAGE 13

13 A bstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ZO N ING VAR IANCE ADMINISTRATION IN PRACTICE: INFLUENCING FACTORS AND TRENDS By Jun Zhao December 2011 Chair: Dawn Jourdan Major: Design, Construction and Planning Over time, local governments have long sought ways to integrate flexibility within the zoning proce ss. The zoning variance represents one of th e se tools. However, this tool is perceived as less than precise because vague legal standards for issuing variances, undue hardship, or practical difficulties allow local decision making bodies to exercise sign ificant discretion. The author conducted a substantial review on zoning variance which covered definition of zoning variance, type s of variance, conditions to grant variance, administrative body and criticism of zoning variance ; through careful examinatio n of these issues, we seek to answer the following questions: According to the literature, which factors affect the decision making of zoning administrative bodies? Moreover, how were these factors determined, and to what extent do they affect the final de cision? Are the se decisions related to the economic and social characteristics of communities? What are the spatial distributions of variance applications and decisions and how do these distributions reflect the correlati on between the decisions and econom ic and social characters? How are the decisions of variances represented in the time

PAGE 14

14 dimension? Are the decisions consistent through time? The purpose of this study was to understand those factors which influence the decisions of administrative boards in t he approval or denial of variances. Four hypotheses in the literature about the decision making of administrative bodies were proposed by the author The first of these is : Hypothesis 1: granting of variance applications is significantly higher than denia l. This is a general view about the Board of Zoning Adjustment ( BZA ) deal with the factors which might affect the Hypothesis 2: opponents from affected neighborhoods influence the zoning administrative Hypothesis 3: type of variance application affects the Hypothesis 4: the zoning administrative board does not place significant weight on suggestions from other public agencies in its decision making. To as sist in answering the se research questions and test ing the four hypotheses statistically a case study was conducted. The author collected and compiled 2 140 variance decisions made by the Board of Zoning Adjustment in Washington D C from 1980 to 2009. Th e first hypothesis was tested by Binomial Test. Further, a simplified binary response model was developed to test the other three hypotheses and examine to what degree those factors affect the considered additional va riables that might affect the BZA decision s According to the results of statistical test s and the binary response model, the author came to the conclusion that o pponents from affected neighborhoods, type of variance, and suggestions from Office of Planning and A dvisory Neighborhood Commissions did affect the Except for the Office of

PAGE 15

15 Planning ( OP ) and the Advisory Neighborhood Commissions ( ANCs ) suggestions other public agency input s were not given significant weight by the BZA in W ashington D.C. It was also found that land value influenced in the BZA decision s The higher the land value, the lower the probability that the BZA would deny the application. In order to test a cluster of variance applications and decisions, the author ap plied Hot Spot analysis in ArcGIS to demonstrate the region s where variance applications and d ecisions were highly concentrated. The maps revealed that cluster s of variance applications and decisions did exist in Washington D.C.

PAGE 16

16 CHAPTER 1 INTRODUCTION Statement of Problem and Research Questions As a tool of land use regulations in the United States, zoning plays an important role in protecting the defining characteristics of local communities, as well as promoting local public welfare and economic development. Unde r zoning ordinances, the land is divided into different districts in which the dimensional characteristics of the lot, the use of the lot, and the buildings on the lot are regulated and specified by different criteria. Since zoning ordinances cannot descri be all situations encountered in complicated land use, strict observation of zoning ordinances might bring unfairness or hardship to individual properties. As a safety valve ( Burke & Snoe, 2004 ; Cohen, 1995 ; Reynolds, 1999 ; Shapiro, 1969 ) the zoning vari ance provides flexibility, allowing land use not expressly permitted by zoning ordinances under some circumstances. State or municipal legislative bodies authorize local administrative bodies (usually called the Board of Zoning Adjustment 1 ) to grant or d in the field of law due to the discrepancies between its theory and practice. In the three key tests commonly applied by many legislative bodies to decide whether to grant or deny an application, their ambiguous expression leaves administrative bodies more flexibility in final decision making. The boards make decisions at their discretion about whether the application is contrary to public welfare; whether it is substantially 1 The name and structure of administrative bodies ar e different in different areas. Usually it is called Board of Zoning Adjustment or Board of Zoning Appeals. In New York City, it is called Board of Standards and Appeals. Since Washington D.C. is a case study in this paper, Board of Zoning Adjustment (BZA ) would be used in this paper to represent all the administrative bodies. Since it is not the primary focus of this paper detailed information about this administrative body are not covered here. For more information, refer to Land Use Law by Salkin

PAGE 17

17 incomp atible with the comprehensive plan; and, whether the applicant would meet undue/unnecessary hardship or practical difficulties if he or she strictly abides by zoning ordinance. Some scholars have noted that zoning administrative boards are too lenient with zoning variance applicants, making the approval rate very high ( "Building Size, Shape, 1951 ; Reynolds, 1999 ) Protestors from neighborhoods tend to urge the board to deny variance applications ( Leary, 1958 ) Since use variances are considered more problematic for the pu rposes of zoning ordinance than area variances, boards are more reluctant to grant use variances ( Burke & Snoe, 2004 ) While the law encourages administrative boards to follow the advice of planning experts, Shapiro ( 1969 ) suggests that the advice of planners seems to have little effect on these boards (p.11) The re search objective of this study wa s to examine the facto rs affecting zoning o n var iance applications, as well as identify trends of variance application s and determination s from the time dimension perspective. Based on existing literature, the possible factors will b e identified and tested to achieve a thorough understanding of their respective influences. The research questions to be addressed in the study are: RQ 1. According to the literature, what factors affect decision making of zoning administrative bodies? Do these previously identified factors aff ect the board s decision significantly? Which factors diverge from the criteria for granting a zoning variance? Is the approval rate of zoning variance applications significantly higher than the denial rate?

PAGE 18

18 Do t he opponents of the variance application from the neighborhood affected Dose the type of variance really affect the ? Do the boards weight highly recommendations from other public agencies? RQ 2. Are the de cisions related to the economic and social characteristics of communities? Does clustering occur with respect to both zoning variance application s and approval s ? H ow do these clusters reflect the correlation between the decisions and the economic and socia l characters? RQ 3. Is there a method we can apply to verify the identified factors influence in different jurisdictions? What are the limitation s ? Research Outline My literature review on zoning and zoning variance i s conducted to better understand th e background, research questions, and research purpose. The origin, purpose, function, and practice of zoning are reviewed. Comments on zoning theory and practice are also listed. A detailed review of zoning variance is conducted, which includes definition s of variance type, conditions for granting variance, administrative bodies related to variance, and relevant criticism. Based on the literature, four hypotheses are proposed to examine the first research question. These four hypotheses are as follows: Hy pothesis 1 Granting of variance applications is significantly higher than denial. Hypothesis 2 Opponents from affected neighborhoods influence the zoning Hypothesis 3 Type of variance application affects the ecision making.

PAGE 19

19 Hypothesis 4 The zoning administrative board does not place significant weight on suggestions from other public agencies in its decision making. Two binary response models are developed based on the literature and the hypotheses. The initi al model is a basic one that includes only those variables that need to be tested in the hypotheses. The dependent variable is probability of granting the variances; the independent variables are opponents, suggestions from other public agencies, and type of variance. An improved model will include more variables that could potentially affect final decisions. In addition to the variables in the initial model, the improved model will include property area, land value, existing type of property, and type of v ariance applicant. Washington, D.C. is the case study area. It was one of the first cities to develop a comprehensive zoning ordinance, which occurred as early as the1920s. In 1958, it izes inter organization cooperation and community communication. The Office of Planning plays an important role in the application of zoning variances. It is also plays an important role in the public review of new zoning regulations. Additionally, the Off ice of Zoning provides online public access to variance cases back to 1960s, as well as the latest cases, which supplies reliable first hand data for my research. Based on more than two th ousand variance cases from 1980 through 2009, this study aims to sum marize the critical information of each case and compile this data for quantitative analysis. In addition, GIS is used to show the spatial distribution of variance applications and decisions, which provides a visual and direct exhibition on the

PAGE 20

20 geographic characters of variance. This is also a good way to display the cluster effect of variance.

PAGE 21

21 CHAPTER 2 LITERATURE REVIEW Background of Zoning People desire homes in pleasant neighborhoods with convenient schools and shopping centers, prosperous communities with spac e for economic activities that offer sufficient and diverse opportunities for employment, a variety of recreation facilities, patterns of urban development that will inspire community pride and participation in cultural and civic affairs, efficient and saf e transportation, and many other factors required of our physical environment for living a full life. In a rapidly growing urban area these physical characteristics can be achieved and maintained if there is sufficient forethought and planning, and if plan s are carried out. 1 County of Alameda Master Plan 3 (1957) ordinance. The Standard State Zoning Enabling Act (SSZEA) was adopted in New York City in 1916 ( Gardner, 2004 ) Zoning is a tool of land use regulation that is designed to protect and promote public health, safety and the welfare of the community, as well as to maintain economic stability and a esthetics ( Schmutz, 1931 ) By regulating the uses 2 of lots and the characteristics of buildings on the lots, zoning regulations, which differ in different districts ( Bassett, 1922 ) are intended to ensure "an orderly physical development of the city, borough, township or other community." 3 Put simply (although somewhat abstrac tly), they ensure "a place for everything and everything in its place" ( Perin, 1977 ) in o rder to make the city an orderly and better place ( Steele, 1986 ) From the perspective of economics, zoning tends to raise the general standard of living 1 Th e epigraph to this chapter is drawn from ( Donovan, 1962 p. 102) originally from County of Alameda Master Plan 3. 2 intended, or for whi ch it is occupied or maintained ( Gardner, 2004 p. 434 ) 3 ( Psota, 2005 p. 537) citing Hanna v. Bd. of Adjustment, 183 A.2d 539, 543 (Pa. 1962).

PAGE 22

22 through maximizing the total net product even though it may be disadvantageous to some individuals in each particular ins tance ( Bailey, 1959 p. 289 ) If use land value to represent the net project a social optimum is attained when each local authority imposes zoning regulations so as to maximize the land value in its jurisdiction" ( Helpman & Pines, 1977 p. 983 ) use development and strategy improve. It also responds to changes in social, economic, and political contexts. Con temporary zoning ordinances bear scant resemblance to those used in the first fifty years of zoning practice ( Owens, 2004 p. 302 ) Zoning originated from concern s about potential nuisance from the laundry services that were developing around residential neighborhoods ( Groves & Helland, 2002 ) which led to regulation of types of land use (residential, commercial, an d industrial). In the first years when zoning appeared, it was not popular as it put strict restrictions on the use of private properties. But modern crowding has brought about recognition of the wisdom of having certain districts devoted exclusively to residential, commercial, and industrial uses ( Dukeminier, 2002 ) In addition, it became popular among homeowners and developers who were concer zoning was originally been proposed by homebuilding developers, however, home owners became a major force in local politics. The purpose was to protect residential properties from decreasing values caused by surrounding industrial and high density residential uses as the improvement of infrastructure and public transportation around 1910 1920 ( Fischel, 2004 ) and to keep the housing market stable and housing prices

PAGE 23

23 predictable. Aside from developers and homeowners, social reformers and planners were also proponen ts of zoning regulations. Social reformers focused on the improved living environment resulting from zoning, while planners were more able to maximize efficiency by orderly assigning functions in each district ( Gardner, 2004 ) The primary purpose of zoning was to protect residential communities from being interrupted by the congestion, noise, traffic, pollution, and general ugliness associated with commerce and industry ( Shlay & Rossi, 1981 p. 705 ) especially for single family residential communities. Working in concert with this goal were other more detailed and comprehensive concerns regarding certain a spects of public services, such as light, space, and traffic, etc. These concerns led to the bulk function of zoning, which is to regulate buildings and the relationship between a lot and the building on it, such as height, number of stories, side yard, rear yard, front yard, lot area, lot occupancy, and floor area ratio, etc. The bulk function is used for three goals: control over density of population in living and working areas, adequate daylighting of buildings, and sufficient open space around buil dings for rest and recreation ( "Building Size, Shape, 1951 p. 507 ) Shlay & Rossi (1981) placed considerable weight on the protective function of zoning. It protects residential neighborhoods, protects property value, and protects the riders 4 in housing market, which helps guarantee that home values yield property tax revenues sufficient to cover the cost of supplying services ( "Zoning for the ," 1980 p. 752 ) 4 public resources without paying the fair share of his/her using. For example, in the housing market, a free rider could be the property owner who builds a high density dwelling in a low density community, which consumes more public services without paying the corresponding costs.

PAGE 24

24 The views about the protect ive function of zoning are relatively traditional and conservative. Steele (1986) summarized two different modes of zoning regulations based on different goals: one is to conserve urban communities, which refers to ove; the other is to use economic rationality to develop land use. Some scholars put high value on the proactive developing function of zoning, which helps to promote local economic development and enhance the economic stability of home ownership ( Dennis, 2000 p. 271 ) Since zoning can affect the price of housing by shifting either supply or demand or both ( Pogodzinski & Sass, 1990 p. 295 ) local government can use zoning as a tool to compensate for the deficiency caused by the housing market and to stabilize the local economy, as well as to correct failures in the housing and public service markets ( Zoning for the, 1980 p. 748 ) Particularly for commercial and indu strial districts, zoning helps to internalize external costs, especially the environmental cost. The economies of agglomeration 5 help further development of local business and economy. When an existing district encounters structural problems and the type o f use may no longer be suitable for redevelopment, rezoning 6 is a useful way to help address and curb further problems. It is shown that a residential to commercial rezoning often is viewed as a measure of increasing employment opportunities for local res idents ( Heffley & Hewitt, 1988 p. 373 ) The protective function and the proactive developing function of zoning represent inconsistent abstract approaches and often come into conflict. What is needed is some balance between them" ( Steele, 1986 p. 711 ) 5 The each other, especially for the companies in related industries. 6

PAGE 25

25 The Standard State Zoning Enabling Act (SSZEA) adopted by New York City in 1916 was already surprisingly complete, and was divided into three functions: control the height of buildings, control the dimension of properties, and protect functional districts divided by uses. However, the contents were conservative because this act should be granted by the highest courts of the state and nation ( Swan, 1949 ) Within nine years, 36 8 municipalities had adopted their own zoning ordinances ( Gardner, 2004 ) v. Ambler, in which th e U.S. Supreme Court upheld zoning plans in the village of Euclid, which were previously rejected by local courts ( Gardner, 2004 ) In the same year, the Zoning Enabling Act (ZEA) was enacted by U.S. Department of Commerce as a model to encourage states to enact their own legislation similar to ZEA ( "Seeking a Variance," 2000 ) Since the implementation of zoning, criticism from some lawyers, urban planners, and others has s eemed ceaseless. Zoning is criticized as being irrational, arbitrary, and venal in operation ( Steele, 1986 p. 712 ) Fichel (1978) commented that zoning was an erosion of private property rights. That is, the rights to determine the use of property and the dimensional charact eristics are transferred from the property owner to local zoning authorities. However, the existence of zoning becomes a potential bed of corruption that could create wealthy zoning officials and land speculators ( Benjaminson & Anderson, 1990 p. 68 ) "Zoning is the single biggest corrupter of the nation's local governments (p.68), said Dan Paul, a Miami attorney. Additionally, since all zoning restrictions have some exclusionar y effects 7 ( Mandelker, Payne, Salsich, & 7 Exclusionar ( Durkin, 2006 p. 445) Aside from economic status, which might exclude some groups from li ving in certain

PAGE 26

26 Stroud, 2005 ) they can lead to inequities and indirect racial and income discrimination. Some examples are zoning regulations on density, minimum building size, or the exclusi on of undesirable groups, which might include low or moderate income families and minorities, from living in certain communities. Though Fair Housing Act (FHA) prohibits housing discrimination based on race, color, religion, sex, disability, familial st atus, or national origin, discrimination based on economic status is not protected by law. A court will only overturn a zoning ordinance that excludes based on economic status if it is coupled with a discriminatory impact on one of the protected classes ( Durkin, 2006 ) Some advocates of smart growth and new urbanism believe suburban sprawl ( Wolf, 2008 ) Yet comments on zoning were not unanimously negative. Simply saying zoning is harmful or useful is arbitrary. In theory and practice, the views on the effects of zoning on the housing market, benefits to property owners, land use efficiency, the local economy, and public welfare are divergent. The complication of land use means that any regulation involving it cannot be wi thout criticism. Zoning is still a widespread and enduring fact ( Steele, 1986 p. 716 ) Moreover, Zoning is a nearly universal feature of land use regulation in the United States. Doubts as to zoning's legitimacy have long since been transformed into general acceptance ( Cohen, 1995 p. 307 ) Zoning as a regulatory tool of land use exerts its protective and proactive functions within the rule of free market. By the test of acceptance in the market place, zoning has been a smashing success ( Babcock, 1966 p. 737 ) Zoning, modified somewhat from its districts, other factors one might consider includes identities of exclusion, such as religion, gender, disabilities, etc. Low income families or minorities might face unreasonable costs if they live in certain areas because of zoning regul ations.

PAGE 27

27 Euclidean origins, remains ascendant to this day, despite nagging concerns about its effect on potential newcomers, on the real estate market, and on the needs of neighboring communities ( Wolf, 2008 ) In addition, in order to cope with the side effects of zoning regulation s some other measures were developed. However, without a Supre me Court requirement, many states and courts have already begun making efforts through legislation in order to eliminate exclusionary effects ( Durkin, 2006 ) Inclusionary/ fair share housing policies are applied by many states to provide low or moderate income families with affordable housing under existing zoning regulations. For mixed use land development, planning unit development (PUD) is designed to give the developer more flexibility to develop a large lot for multiple land use. Planning professionals take on more responsibilities in the design of PUD lots ( Sampson, 2007 ) By using these tools, some negative effects of zoning cou ld be alleviated or eliminated. Since zoning ordinances cannot list all the possible situations individual properties might experience, their strict observation might have unfair or undesirable effects on individual properties. Zoning variance was designed as a tool to address the inflexibility of zoning and the problems that one size fits all zoning ordinances cause ( Sampson, 2007 p. 879 ) As zoning is a nearly universal feature of land use regulation ( Cohen, 1995 p. 307 ) as mentioned above, zoning variance is also a nearly universal feature in zoning regulation ( Sampson, 2007 p 888 ) Zoning variance is a tool to perfect a crude regulatory instrument ( Owens, 2004 p. 283 ) The legislative status of zoning variance is determined by eac

PAGE 28

28 variance is granted by certain administrative bodies 8 (quasi judicial bodies 9 ) according to rules and certain circumstances related to judgment ( "Replacing the Hardship," 1 987 ) Variance is derived from traditional Euclidean zoning 10 ( Cohen, 1995 ) It is commonly thought that these functions were originally meant to be administrativ e safety valves ( Burke & Snoe, 2004 ; Cohen, 1995 ; Reynol ds, 1999 ; Shapiro, 1969 ) that protect the constitutionality of zoning ordinance. Also, it prevents the city or county from being held liable under the Takings Clause 11 of the Constitution or the zonin g ordinance, from being declared unconstitutional under the substantive Due Process Clause of the Constitution ( Burke & Snoe, 2004 p. 531 ) Variance was seen as a pragmatic means of taking individual disputes out of the political and judicial realms that would likely be less hospitable to effective zoning practice ( Owens, 2004 p. 284 ) Property owners should first apply for variance before they claim hardship to the courts ( "Zoning Variance and ," 2005 ) Today the application of zoning variance has greater benefit for the commu nity and property owners, as it is "designed as an escape hatch from the literal terms of the ordinance which if strictly applied, would deny a property owner all beneficial use of his or her land and thus amount to confiscation. 12 The variance is a "permi tted violation of the zoning regulations ( Shapiro, 1969 ) which is a way to correct maladjustments and 8 Ibid. p.9. 9 judicial function resembles that exercised by a judge, and typically is denoted by public hearings for which notice is given and an opportunity to be heard is provided, as well as the application of the particular facts ( Sampson, 2007 p. 878) 10 "Euclidean zoning refers to that type of zoning characterized by the identification of use based zones, traditi onally residential, business, and industrial. Uses are typically allowed hierarchically, that is, uses ( Cohen, 1995 p. 330) Additionally, height and area regulations are also seen as Euclidean zoning ( Wolf, 2008 ) 11 12 ( Owens, 2004 ) citing Lincourt v Zoning Bd. of Review, 201 A.2d 482, 485 86 (R.I. 1964)

PAGE 29

29 inequities in the operation of general regulations 13 and the main purpose of allowing variances is to pr event land from being rendered useless ( "Zoning Variances," 1961 ) In addition, the variance helps property owners to avoid complicated procedures when they face hards hips under existing zoning regulations. The variance is a simple, cost effective means of providing such relief without the necessity of either ordinance amendment or litigation ( Owens, 2004 p. 371 ) Zoning Variance Definition of Zoning Variance The description of zoning variance is usually shown in state or municipal zoning ordinance act, land use law book, and on the website of local planning and development service agencies. The author summarized four main elements which are usually contained in the definition of zoning ordinance : the attribute, applicant, objects affected, and granter. attribute. Zoning v ariance is an author ization ( Salkin, 2008 ) It is a constitutional grant. It is also a relief from zoning ordinance, which relieves the applicant from strictly abiding by zoning code ( Salkin, 2008 ) For example, the BZA grants an application proposing to decrease the number of parking spaces below the required minimum number based on the existing build ing, land area and surrounding open space. If zoning ordinance was strictly applied, this building would be rendered useless. The second element is the applicant of variance. Some in the literature say it is individual, some specify it to be the property owner or land owner, and some just refer to applicant ( Burke & Snoe, 2004 ; Donovan, 1962 ; Madry, 2007 ; Mandelker, 2003 ; 13 ( Salkin, 2008 ) citing Visco v. City of Pla infield, 136 N.J.L. 659, 57 A.2d 490 (N.J. Sup. Ct. 1948)

PAGE 30

30 Reynolds, 1999 ; Shapiro, 1969 ; "Zoning Variances and Exceptions: The Philadelphia Experience," 1955 ) According to zoning orders 14 in Washington, D.C., an applicant could be an individual, a for profit organization, or non profit organiza tion. In New York City, the applicant could be be the head of any agency ( State of New York Legislative Bill Drafting Commission 2004 p. 177 ) The third element involves the objects affec ted. In accordance with zoning ordinance, variance affects both buildings and lands within zoning districts. When an applicant applies for variance, it must be based on the physical condition of the land ( "Zoning Variance and," 2005 p. 209 ) It could be variance on building, or variance on land, or on both. For example, variance on building could relate to the number of stories (or height); variance on land could relate to lot area (or width); var iance on both could be use (or lot occupancy). It also means the determination of variance application is based on the condition of the property, not the condition of the property owner. The decision should not be different on the same property due to diff erences between The fourth is the granter. The granter is an administrative body usually named Board of Zoning Adjustment, which is an independent quasi judicial agency in charge of granting variances, special exceptions, and appe als related to the enforcement or administration of the zoning ordinance ( Salkin, 2008 ; "Zoning Variances and Excep tions: The Philadelphia Experience," 1955 ) 14 Available on DC Office of Zoning website, http://dcoz.dc.gov/main.shtm

PAGE 31

31 In brief, a variance is an administrative authorization to relieve the characteristics of zoned property from strict accordance with zoning regulations under the application of property owner 15 Type of Zoning V ariance Variances occupy two categories. One is area/bulk/dimensional variance 16 ; the other is use variance ( Barry, 1993 ; Burke & Snoe, 2004 ; "Zoning Variances," 1961 ) The former refers to the modification of physical characteristics related to building or land or relationship between both (e.g., front yard, lot occupa ncy, height, story, open space, etc). The latter grants uses prohibited by zoning ordinance. Use variance can also be classified into two types. One is to establish or continue a use (e.g., establish a fast food restaurant in the district limited to resid ential use); the other is to change an existing use to another which is currently prohibited ( "Zoning Variances and ," 1955 ) (e.g., chang e use from a convenient store to a restaurant in a residential district) 15 Some definitions of variance: t permitted by ( Donovan, 1962 p. 103) A variance is "an authoriza tion to use property for a purpose prohibited by the zoning ordinance; and a use which is authorized by a 'variance' is not personal or limited to a particular owner of the property but rather 'runs with the land.' Balodis v. Fallwood Park Homes, Inc., 6 4 Misc. 2d 936, 283 N.Y.S.2d 497 (Sup 1967). legislative, quasi judicial determination that the use allowed is not offensive to the zoning ordinance with regard to the particular circumstances City of Garfield, 119 N.J. Super. 181, 290 A.2d 737 (App. Div. 1972). ( Reynolds, 1999 p. 127) a building or structure, or for the establishment or maintenance of a use of land, which is prohibited by a zoning ordinance. It is granted by an administrative body pursuant to power vested by statute or ordinance, and is a form of administrative relief from the literal import and strict application of the zoning regulations. A variance runs with the land and passes with the land to a subsequent purchaser 3) 16 ( Sampson, 2007 p. 881)

PAGE 32

32 Variances are not classified according to a universal standard in the United States. In fact, sometimes it is vague. As a result, it is hard to categorize on site parking, signs, an d density into those two types ( "Replacing the Hardship," 1987 ) For example, parking and floor area are treated as use variance in some states, while in others they are treated as area variance. The classification is dep endent on the language of the zoning ordinance the variance is requested from. If the zoning ordinance acts merely prohibits the use, then the requested variance is a use variance ( Salkin, 2008 volume 13 15 ) Currently eighteen states do not distinguish between use and area variances either legislatively or judicially ( Salkin, 200 8 volume 13 18 ) ; eight states have different standards for area variances and use variances; twelve states prohibit use variances; three states have unique provisions governing the issuance of variances ( Salkin, 2008 volume 13 34 ) ; other states statutorily give local municipalities and entities the power to ( Salkin, 2008 volume 13 36 ) It is commonly thought use variance is contrary to the purpose of establishing zoning regulations. Some courts views use variance is a greater threat to the integrity, and fairness, of a zoning scheme than area variances ( Cohen, 1995 ) Furthermore, use variance generally possesses a greater potential to enable a landowner unfairly to receive a substantially larger return on property than a similarly situated landowner who cannot engage in the use ( Cohen, 1995 p. 331 ) As a result, some states do not al low use variance.

PAGE 33

33 Disadvantages of permitting use variances One reason some jurisdictions do not allow use variance is that to allow use variances on the theory constitutes rezoning 17 and rezoning is a legislative function. justment to issue a use variance turns on whether the court finds that the legislature intended to confer such power and whether such a grant of power is an unconstitutional dele gation of legislative authority ( Juergensmeryer & Roberts, 2003 p. 188 ) Granting of use v ariances might constitutes an improper amendment of the zoning ordinance and use variances would usurp the amendment power ( Mandelker, 1997 ) Besides, it is commonly thought use variances are contrary to the purpose of threaten adjacent land with the es tablishment of an incompatible use, or with maintenance of a use that will change th e character of the neighborhood ( Salkin, 2008 volume 13 17 ) Especially for the la rge lots and where the proposed use is not otherwise allowed anywhere in the city. If such kinds of variances are granted, the changes on the local community are huge. Field studies also found substantial abuses in the granting of use variances ( Mandelker, 1997 ) Also there were arguments that though the proposed use would be advanta geous to the public, the board is not empowered to decid e what the public needs (Juergensmeryer & Roberts, 2003) 17

PAGE 34

34 Advantages of permitting use variances First, use variances embrace the purpose of setting variances in the zoning regulation. Some properties might encounter unnecessary hardship if strictly abide by the ordinance and the denial of use variances might constitute a taking. Where use variances are allowed, the courts find the guidelines spelled out in the enabling act sufficiently circumscribe t he decision making power of the board to overcome the unconstitutional delegation problem. The grant of the use variance, where permitted, is not viewed as an uncontrolled discretionary act but, rather, one that is limited by the necessity to find that unu sual conditions exist. Other courts have also rejected the unc onstitutional delegation theory ( Rohan, 1990 p. 43 ) Bair (1970) showed that a proposal to prohibit use variance in a city met strong that the board was being i llegally stripped of its powers (p. 479). Second, use variance may be faster and cheaper than a rezoning, and it may require less paperwork and few er hearings before fewer bodies (Juergensmeryer & Roberts, 2003) Especially in the highly d eveloped urban areas, more and more redevelopment happens and the needs for use variances are increasing. It is found that the application for use variances in New York City increased dramatically. In 2001 2002, around 64% of applications were use variance s, while in 1976 the percentage wa s 28% ("Zoning Variance and," 2005). If use variances are prohibited, it would increase tremendous burden on the legislative bodies to amend the map and rezone areas frequently. Bair (1970) showed the city attorney stated (not for publication) that issuance of use variances was

PAGE 35

35 convenient because it relieved the governing body of the necessity for considering a great many zoning amendments. The above indicates use variance is still very popular in some places. For the sta tes that allow use variance, the BZAs are usually stricter in granting use variances compared to area variances (Burke & Snoe, 2004). Some jurisdictions also have stricter standards compared to area variance s In some jurisdictions, use variances are only prohibited where the change to be allowed is significant ( Rohan, 1990 ) variance was a controlli ng factor in most of these case ( Mandelker, 1997 p. 2 52 ) Some scholars also point out it is arbitrary to say use variances are more harmful than area variances. In a low density residential area, the negative influence to granting 10 story addition is more than the granting of one seat hair style salon in a residential building. Conditions to Grant Zoning V ariance SSZEA is thought to have created some of the first definitive criteria serving to better define the circumstances under which granting a variance would be justified ( Sampson, 2007 p. 890 ) Under the guide of SSZEA, local legislative bodies make their own zoning ordinances. Different states have different jurisdictional structures that govern the granting of variances; some are under the governance of sta te statures, while others are under municipal law. In addition, under different jurisdictions, the conditions to grant variance also vary. Salkin (2008) summarized this well and presents a complete list of states. Though a variety of tests are used in gra nting variance, most states share some commonalties:

PAGE 36

36 (1) The variance must not be contrary to public interest, safety, and welfare ( Burke & Snoe, 2004 ; Donovan, 1962 ; Madry, 2007 ; "Zoning Variances and Exceptions: The Philadelphia Experience," 1955 ) This condition is in accordance with the principal purpose of zoning ordinance. (2) The variance would not be substantially incompatible with the comprehensive zoning plan ( Burke & Snoe, 20 04 p. 532 ) One of the conditions for any zoning ordinance is that it be in accordance with a comprehensive plan ( Madry, 2007 ) Though variance is an exemption from the zoning regulations, it still needs to observe the spirit of comprehensive zoning plan. (3) The strict application of zoning regulations would result in an undue/unnecessary hardship on the applicant ( Burke & Snoe, 2004 ; Donovan 1962 ; Reynolds, 1999 ) An applicant must satisfy each of these conditions to get the permission of variance. The important part in the above conditions is the term undue/unnecessary hardship, whi ch is the key test to grant a variance. It is the most common standard in zoning variance widely used in the United States ( Cohen, 1995 ) Barry (1993) indicates th e classic statement of hardship tests occurred in 1939 New York case of Otto vs. Steinhilber and it has been widely adopted in practice : Before the board may ... grant a variance upon the ground of unnecessary hardship, the record must show that (1) the la nd in question cannot yield a reasonable return if used only for a purpose allowed in that zone; (2) that the plight of the owner is due to unique circumstances and not to the general conditions of the neighborhood which may reflect the unreasonableness of the

PAGE 37

37 zoning ordinance itself; and (3) that the use to be authorized by the variance will not alter the essential nature of the locality. (p. 46) This Otto vs. Steinhilber test is not universally applied by every zoning ordinance, but it does have great i nfluence and is considered to be the classic statement ( Mandelker, 2003 ) Two terms are key to explaining undue/unnecessary hardship. The first is re asonable return. A reasonable return refers to fair return/profitable return, and "the most firmly entrenched and reiterated declaration is that mere financial hardship or an increased return from the property is not a sufficient reason for granting a v ariance ( "Zoning Variances and," 1955 p. 520 ) Yet some courts use reasonable use instead of reasonable return, or interpret both to have the same meaning. A property that would not be suitable for use if it strictly abided by the existing zoning regulation could be a justification of hardship ( Jacobs, 1958 ) If it is shown that the zoning regulation leads to the denial of all reasonable use, it actually equals a constitutional taking. In practi ce, most courts will generally approve the grant of a variance where the landowner would otherwise be denied the reasonable use of his or her property ( Cohen, 1995 p. 309 ) In Connecticut, applications to add a pool, porch, or addition to an existing structure ( "Replacing the Hardship ," 1987 p. 680 ) could not be seen as reasonable use. Usually applicants need to show enough ev idence that their properties cannot yield reasonable use. Sustained unsuccessful efforts to sell a property for permitted uses are often deemed a sufficient basis upon which to find that a reasonable return is not possible ( Cohen, 1995 p. 197 ) The Michigan Court of Appeals demands the landowner must show no reasonable return, hardship amounting to confiscation, or

PAGE 38

38 the deprivation of all reasonable use ( Cohen, 1995 p. 187 ) Cohen (1995) also shows New York State Court of Appeals listed some elements which the property owner needs to present in the application of varianc es as justification of no reasonable return, which is anecdotally called dollars and cents evidence ( "Zoning Variance and," 2005 p. 210 ) These factors are: (1) the amount the applicant paid for the entire parcel; (2) the present value of the parcel or part of it; (3) maintenance expenses; (4) taxes on the land; (5) mortgages and encumbrances; (6) income; and (7) other relevant factors, including the applicant's estimate of what constitutes a rea sonable return ( Cohen, 1995 p. 336 ) Courts have considered 3.6%, 6.9% and 9.9% as a sufficient rate of return, emphasizing that different circumstances may dic tate a different rate of return ( "Zoning Variance and," 2005 p. 210 ) The second is unique circumstance. Burke & Snoe (2004) explain that the relief for which the applicant applies does not sha re the general characteristics of the neighborhood. However, it does not mean that the property is the only one possessing a particular undue/unnecessary hardship. If a given hardship occurs in the entire neighborhood, it may reflect the unreasonableness of the zoning ordinance itself ( Cohen, 1995 p. 180 ) and may require the zoning legislative agency to amend the zoning regulation for a solution. The BZA does not have the authority to address a acknowledgement by the jurisdiction that zoning variance might inevitably cause hardship to certain property owners and requires the BZA to exert administra tive mechanisms to address this issue ( Cohen, 1995 ) It could not be considered a unique hardship if an applicant presents proof of inappropriateness of zoning reg ulations.

PAGE 39

39 Furthermore, this unique hardship should not be self created (or self imposed, self inflict, situation ( Jacobs, 1958 ; Reynolds, 1999 ; Rice, 2006 ) A hardship is considered self created if property is purchased s ubject to restrictions that are sought to be varied, and the applicant was aware or should have been aware of the zoning restrictions at the time of purchase ( Rice, 2006 p. 1127 ) The United States Supreme Court has addressed this and rejected a firm rule that bars a taki ng or hardships claim based on the purchase of property subject to the land use restrictions at issue ( "Zoning Variance and," 2005 p. 211 ) .Another situation is considered to be self created: the applicant's violation of the ordinance, knowing or unknowing, and his subsequent application for a variance based upon his expenditures [is] the hardship suffered. Unless the applicant is otherwise entitled to a variance, relief will be denied ( Jacobs, 1958 p. 822 ) In the Otto vs. Steinhilber test, the third requi rement (i.e, that the use to be authorized by the variance will not alter the essential nature of the locality ) reflects views, congestion, community character, noise, building size, arch itectural design, and environmental issues ( "Zoning Variance and," 2005 p. 210 ) as well as aesthetic concern. However, courts focus not so much on the impact, but on the purpose of the regulatio n and the interests and values sought to be protected ( Cohen, 1995 p. 337 ) In hardship tests, New York City requires one more condition for applicant to demonst rate hardship: the variance requested is the minimum variance required to alleviate the hardship ( "Zoning Variance and," 2005 p. 207 ) It reflects the principle of

PAGE 40

40 preserving zoning integrity and minimizing the negative influence the variance might cause. However, this test is not a strict requirement, even by court. Besides undue/unnecessary hardship, practical difficulty is another standard adopted by some states and administrative bodies usual ly apply it in practice. Practical difficulty is seen as a more lenient and relaxed standard compared to undue/unnecessary hardship since it emphasizes the feasible use of property ( Burke & Snoe, 2004 ; Madry, 2007 ) As mentioned above, area variance does not vary the mai n characteristics of a community (compared to use variance), so some states use practical difficulty as a standard to test area variance. In the case Village of Bronxville v. Francis, the court, for the first time, applied practical difficulty standard to support area variance ( Sampson, 2007 ) In many states, practical difficulty is employed for matters relating to area variance and undue/unnecessary hardship is employed for matters relating to use variance ( Barry, 1993 ; Burke & Snoe, 2004 ; "Seeking a Variance as a Prerequisite to C hallenging a Zoning Ordinance," 2000 ) This standard also reflects the general policy disfavoring use variances ( Cohen, 1995 p. 339 ) Salkin (2008) cited the case Anderson v. Board of Appeals, Town of Chesapeake Beach 18 which gave a relatively thorough explanation about practical difficulty: For practical difficulty, the applicant need show only that (1) compliance with the strict letter of the restrictions wou ld unreasonably prevent the owner from using the property for a permitted purpose or would render conformity with such restrictions unnecessarily burdensome; (2) a grant of the variance applied for would do substantial justice to the applicant as well as t o other property owners 18 Anderson v. Board of Appeals, Town of Chesapeake Beach, 22 Md App. 28, 322 A.2d 220 (1974).

PAGE 41

41 in the district; and (3) relief can be granted in such fashion that the spirit of the ordinance will be observed and public safety and welfare secured. (volume 13 28) New York courts use several factors to test area variance under t he practical difficulty standard: the economic impact on the landowner, the extent of the variance, the effect on the neighborhood, and whether the landowner could pursue alternative use economic harm, the burden shifts to the municipality to show that application is necessary to advance the purpose of the ordinance or prevent injury to the public ( Cohen, 1995 p. 340 ) Administrative Body Since zoning ordinances enacted by states cannot cover every conceivable situation, most states authorize local municipal agencies or communities to establish their own administrative bodies to to vary the ap plication of the provisions of the zoning ordinance ( Jacobs, 195 8 p. 821 ) This setting is good for addressing local problems on a case by case basis ( Durkin, 2006 ; Gardner, 2004 ) The state level agency has the right to r eview the decisions made at the local level to guarantee the integrity and observance of state zoning enabling statutes. The Board of Zoning Adjustment is delegated to authorize zoning variance by state or municipal legislative bodies. Aside from making de cisions on whether to grant variance or not, the BZA also assumes other responsibilities, such as granting special exception 19 and hearing appeals related to the 19 Special exception is a use allowed by the provisions of the zoning ordinance under some conditions. It must get approval from BZA to make sure this use does not bring negative effects to the neighborhood.

PAGE 42

42 enforcement or administration of the zoning ordinance. Usually five or seven members are electe d or appointed to compose the BZA. In Washington, D.C., the BZA is composed of five members, three of which are mayoral appointees, one a rotating member of the District of Columbia Zoning Commission, and one a designated representative of the National Ca pital Planning Commission 20 In New York City, the board also consists of five members featuring at least one professional and experienced planner, one architect and one engineer. The final decision can be made with at least at three members present ( State of New York Legislative Bill Drafting Commission 2004 ) The BZA does not have the authority to amend zoning ordinance in a manner that changes the essential character of a community, nor can it review the legislative their decision ( "Seeking a Variance," 2000 p. 2033 ) The state courts are the only the state and the public with a reasonableness ( Gardner, 2004 p. 423 ) Usually there are two steps in zoning variance administration. The first step is to de termine whether the application for a building conforms to the zoning code. If it conforms, the building permit is granted. This step is done by another zoning agency as opposed to the BZA. If the application does not conform to the zoning code, then the applicant must amend the original plan and resubmit the application for approval. Alternatively, the applicant may go on with the second step to apply for a zoning variance or special exception from the BZA ( "Administrative Discretion in Zoning," 1969 ) Usually, the cases the BZA reviews should first be reviewed by other zoning agencies to 20 District of C olumbia Office of Zoning, http://dcoz.dc.gov/services/bza/bza.shtm

PAGE 43

43 determine if they do or do not conform to the zoning code. Once the review is comp leted, a public hearing is required for public testimony. The purpose of public (Witten, 2007 p. 25 ). In order to grant a variance, it is required by ordinance a certain number of the bo ard members vote to grant. Ordinarily, for boards consisting of five members, if majority of the three vote to grant, the variance is approved. However, the regulations are different in different states and cities. In Denver, the board is composed of five members. Application of variance can only be granted when at least four members vote for approval ( Sampson, 2007 ) In most states, the BZA shall grant an application, grant an application subject to conditions, or de ny an application at their discretion. In New York City, the variance previously granted could be revoked or modified if the terms and conditions of such grants have been violated ( State of New York Legislative Bill Drafting Commission 2004 p. 175 ) An interesting finding is that in 2001 and 2002, all the grants in New York City were conditional ( "Zoning Variance and," 2005 ) This might reflect that the board aimed to keep changes to neighborhoods to a minimum, while giving the applicants opportunities to make some changes under the existing zoning ordinance. In the previously mentioned conditions where a variance is granted, there is no explicit line defining whether a variance is contrary to public welfare, whether it is substantially incompatible with a comprehensive plan, or whether it meets undue hardship or practical difficulties. Though many courts explained and showed tests of undue hardship and practica l difficulties, the descriptions of these two terms are still not very clear or strict.

PAGE 44

44 State courts have not dealt kindly with discretionary elements in zoning....many courts have condemned new discretionary plans on grounds either that the legislature m ay not constitutionally delegate its power to administrative bodies, or any more satisfactorily than appe al boards, nor have they developed more precise standards to guide the boards in granting exceptions and variances. ( "Administrative Discretion in Zoning," 1969 p. 682 ) Since it is the members in the BZA are final decision makers. Different members might have differing opinions based on their own discretion. Though they do not have the power to amend the z oning ordinance ( Burke & Snoe, 2004 ) they have the power to grant or deny variance based on their own comprehension of zoning ordinances and the situation of properties. Beca use it is impractical to adopt a zoning law that is both completely in the interpretation of the [zoning law is] delegated to an administrative body or 21 The general standards for granting zoning variance implies that the essence of the zoning variance is its flexibility ( "Administrative Discretion in Zoning," 1969 ) a way to handle different situations is in reality based on the context and special situation. If the purposes of zoning are to be accomplished, the master zoning restrictions or standards must be definite while the provisions pertaining to a co nditional use 21 ( Rice, 2006 p. 1137 ) citing Arceri v. Town of Islip Zoning Bd. of Appeals, 16 A.D.3d 411, 412, 791 N.Y.S.2d 149, 151 (2d Dep't 2005).

PAGE 45

45 or a variance, designed to relieve against uncertain eventualities, must of necessity be broad and permit an exercise of discretion. 22 Reynolds (1999) said, the power to grant or deny lies in the discretion of the members of the board of adj ustment, and their exercise of this discretion will not be overturned unless they act arbitrarily, capriciously, or outsi de the scope of their authority (p. 128). In Connecticut, T his deferential standard has been further relaxed by what might be called the uphold a ZBA 23 fairly exercised after a full hearing ( "Replacing the Hardshi p ," 1987 p. 688 ) It implies that while the decision made by the BZA might be controversial, the court might not reverse the ( Gardner, 2004 ) variances were frequently supported by courts in judicial cases over a relatively lengthy period (from 1962 to 2003). The exact percentage of cases upheld by cour ts was 85%. Courts are deferential to BSA decisions, as they generally are with governmental decision makers ( "Zoning Variance and," 2005 p. 230 ) Zoning Variance and Rezoning Zoning variance and which is not allowed by the provisions of the zoning ordinance. However, while a use 22 ( "Administrative Discretion in Zoning," 1969 p. 671 ) citing Tustin Heights Ass'n v. Board of Supervisors, I70 Cal. App. 2d 619, 634 35, 339 P.2d 9I4, 924 (Dist. Ct. App. 1959) 23 In Connecticut, the administrative body is called Zonin g Board of Appeal (ZBA).

PAGE 46

46 variance grants the owner an exemption (and leaves the ordinance intact), an ( Barry, 1993 p. 55 ) Usually rezoning i s used for a relatively large area which is not suitable for use under existing zoning regulations. The decisions on zoning variance and rezoning are usually managed by different public bodies. In Washington D.C., the BZA is the administrative body to deci de whether to grant or deny a zoning variance. The Zoning Commission (ZC) 24 is in charge of preparing and a djusting zoning map. The process between zoning variance and rezoning is also different. Zoning variance is first proposed by an applicant, who has de Rezoning is initiated by the public agency which is in charge of rezoning, based on economic and social changes and existing situation of an area. If a residential district sh ow s serious blight and a large number of population move s out of the area, it may reflect that the area is not suitable for existing zoning function. When there is clustering of zoning variances, especially use variances, it might show the sign that the wh ole area is not suitable for existing zoning setting. Criticism of Zoning V ariance In the practice of zoning variance, it seems that zoning boards focus more on flexibility as opposed to rigid discretion in every test. As early as the mid 1930s, variance practices were criticized as easy and erratic. 25 Madry (2007) described zoning variances more strongly, calling them notoriously badly administered (p. 486). The vague standards in determining whether applications satisfy the conditions encouraged the exercise of considerable discretionary power by zoning boards 24 In Chapter 4 the author will provide more information about the functions of each public agency. 25 ( Shapiro, 1969 p. 9 ) Citing Woodruf, A Zoning Primer 66 (Proceedings of the Annual Planning Conference 1935)

PAGE 47

47 ( "Administrative Discretion in Zoning," 1969 p. 668 ) The BZA has also been described as too lenient wi th applicants whose grants are wild and liberal ( "Building Size, Shape, 1951 ; Reynolds, 1999 ) and can generate a decision which bears little resemblance to zoning theory or legal norms ( "Administrative Discretion in Zoning," 1969 p. 668 ) Since some of the board members are a ppointed or elected from the neighborhood, they have "a natural disposition not to be too harsh ( "Zoning Variances," 1961 p. 1407 ) on their neighbors. In Michigan sta te, the members of Zoning Board of Appeals appointed by local government are not required to be familiar with land use policies and regulations ( "Seeking a Variance," 2000 ) In Connecticut, most of the board members do not have formal training ( "Replacing the Hardship ," 1987 ) The unprofessional backgrounds of some members, combined with political bias and pressure, can lead to inappropriate or controversial zoning decisions ( Gardner, 2004 ) It is also noted that the unique circumstance is no longer a significant aspect of the test for a variance ( Madry, 2007 p. 490 ) Regarding the test for undue/unnecessary hardship, it becomes a balance between the stakeholders. Many administrative bodies viewed the variance as a useful tool for balancing individual and community needs in a variety of circumstances rather than a device limited to amelioration of significant hardship ( Owens, 2004 p. 295 ) Applicants might easily gain approval if there is no opponent. Some sample data collected in various cities showed the rate of approval was very high as a proof of improper usurpations of pow er by the BZA. 26 Though the high approval rate does not alone indicate the improper decisions by the board, that suspicion seems particularly well placed, given the relatively strict standards of the 26 Reps, Discretionary power of the board of zoning appeals, 20 Law & Contemp. Prob. 280,281, 1955

PAGE 48

48 traditional variance approval criteria ( Sampson, 2007 p. 893 ) In a study of misrule, it was estimated that for applications of area variances granted in 1961 1962, on the basis of the facts alleged in the petition and the evidence in the minutes, in not more than twen ty, or approximately forty percent, of these cases were the legal requirements for a variance satisfied ( Dukeminier & Stapleton, 1962 p. 287 ) In Connecticut, the board displays lenience towa rd residential applicants, who usually apply for new swimming pools, garages, porches, or family rooms added to existing homes ( "Replacing the Hardship ," 1987 p. 691 ) In New Hampshire, it was found by the New Hampshire Supreme Court that 10% of variance grants were illegal during 1987 to 1992 ( Kent, 1993 ) Especially in cases when the applicants plan to demolish and rebuild their houses using area variances, they might get great support from the neighborhood, since new, go od looking buildings can help to increase surrounding property values. In this situation, the boards might weigh more heavily the benefits provided to a community versus the strict interpretation of zoning regulations ( Durkin, 2006 ) It was confirmed by a planner in New York City that the board works with city planners to weed out inappropriate subm issions and improve the process ( "Zoning Variance and ," 2005 p. 208 ) which might be another reason for the high approval rate. Also, the decision making process does not provide public participation. Though the public hearing is a channel between the community and decis ion makers, variance approval still depends final decision ( "Administrative Discretion in Zoning," 1969 ) Though the effect of individual zoning variances on a community might seem insignificant, (particularly for area variances), the overall effect of all granted zoning variances on a neighborhood could be remarkable. The quasi judicial nature of the

PAGE 49

49 variance process often forces the board of ad justment to focus on the needs of the petitioner and the impacts on immediate neighbors. What is often lost in such a process is the view of cumulative impacts of variances on broader community interests ( Owens, 2004 p. 319 ) Neighborhood character could be changed drastically by the cumulative effect ( Sampson, 2007 ) such as perception of density, esthetic layout, etc. How ever, the boards paid scant attention to this problem. The boards have an intuitive sense that the fabric of local zoning will probably not be damaged if a homeowner receives a minimal variance ( "Replacing the Hardship ," 1 987 p. 719 ) In New York City, clear evidence of clustering of variances showed up in some communities, which could lead to further granting of variances there and open up the possibility of rezoning. The Board is playing a role of shaping land use an d become a source of unexpected change in some communities, though it has never been authorized with this power by legislation ( "Zoning Variance and," 2005 p. 199 ) Another criticism is that whil e the granting of area variances might help to increase surrounding property values, an improper decision based on financial consideration may lead to exclusion ( Durkin, 200 6 ) In the literature, most discussed the legislative issues related to zoning variances and the discrepancy between theory and practice. Though it might be true that the BZA is somewhat lenient in its decision making and some factors might affect (or be highly correlated with) final decisions, most studies list a variety of factors affecting the decision making separately and do not show to what extent each one affects the final decision. Most studies use qualitative analysis with few data support. The author has not found any study applying econometric analysis to testify their conclusion.

PAGE 50

50 Furthermore, very few studies analyzed zoning variance dec ision from the time dimension. It is unclear how variance applications and decisions change over year Based on criticism of zoning variance, this paper focuses on zoning variance administration in practice in order to understand what factors influence the decisions of these boards in the approval or denial of variances. The author tries to determine the r elationship between the According to previous statements in literature, four hypotheses are proposed and wil l be tested by using quantitative analysis. Trends related to decision making on variances and the spatial distribution of zoning variances are also discussed in this paper.

PAGE 51

51 CHAPTER 3 METHODOLOGY AND PROC EDURE Research Design and Conceptual Framework Based on a c omprehensive review of the literature on zoning and zoning variance ranging from 1922 to 2008, the author will summarize facts and comments on zoning variance, as well as factors that might affect the methodologies used in each study will be examined and analyzed to determine better ways to test those facts and factors. This paper summarizes the factors that are thought to affect the decision making of zoning variance rather than the requirements of existing law according to published literature. Four hypotheses are proposed, each of which is discussed extensively in the literature. A basic binary response model is built based on these statements. In addition, in order to test other possible factors that might affect decision making, the author tests another binary response model including more variables which might affect the decision The binary response model is a model used widely in decision making analysis in which the dependent variable is the response probability (e.g., the probability of buying a Japanese car). The independent variables include the factors that affect the dependent variable. The theory and application of this model will be introduced in the section on The next step was to collect data for analysis. Two methods can be used to collect data regarding the determination zoning variance. The first method was to summarize zoning variance cases during a period based on the official documents. Using the dat a resear chers could investigate the relation between the BZA decision s and other factors (e.g., arrival of protestors and letters of opposition from the neighborhood, type of

PAGE 52

52 application. The advantage of this method is that it is objecti ve The shortcoming is we would miss some information which was not recorded in the documents. The second method is via survey. The surveyors send designed questions to the administrative officials and related agencies for their response. Using this method, some information that ca planning, engineering, and architecture, attitudes towards protestors, and other public agencies). Compared to the first method, this method is process oriented and the fir st method is result oriented. However, the deficiency of this method is that the responses aforementioned methods. However, due to limitations posed by time and practic ality, few researchers have used both. Owens and Brueggemann ( Owens & Brueggemann, 2004 ) conducted a comprehensive survey about zoning variance experience in North public agencies, and practitioners. This paper applies the second method as a and the interrelationship between the variables. Aside from data collection methodology, geographic scope of data and its time range also need to be considered. The cases examined could be limited within a certain administrative boundary, or in a larger region. The study could be one year cases, or examine 10 or more years. The larger the geographic scope, the more generous and applicable the conclusion. The longer the time range, the more stable the conclusion. Also, with larger geographic scope and a long study period, one can determine more

PAGE 53

53 characteristics o f zoning variance. For example, one can compare the variance determination between two administrative bodies to ascertain commonalities and differences. Another example is that in the long range study, one might find trends and changes in the variance dete rmination. This study applies single case study and the time period covers 30 years. The justification to use single case study method is shown Washington, D.C. is the study area, and the author will review all 2140 variance cases from the 19 80 through 2009. The documents of cases are available on the s from the Advisory Neighborhood Commissions (ANCs) and the Office of Planning (OP) contents of pub lic hearing, the organize this information for quantitative analysis. The reason to choose Washing, D.C. as the study area is explained in the section on Once the data collection is complete, the data will be exported to the software programs, ArcGIS and SPSS. GIS is a database management system that is used to store, display, and analyze spatial data for decision making. ArcGIS is software produced by ESRI to perform the function of GIS. SPSS is also widely used software for quantitative analysis. The author will conduct data processing through ArcGIS for spa tial analysis and SPSS for statistical analysis. Based on the results processed from SPSS, the author will confirm or disprove the hypotheses proposed previously. Besides, the author will further investigate other

PAGE 54

54 variables that might affect the sion. In addition, spatial analysis from ArcGIS will show the distribution of variance applications and decision. Hypotheses Approval vs D enial As mentioned above, many studies have found that the approval rates were significantly higher than the denial r ates in different areas. Studies showed that from 1925 to 1940, the approval rate in large metropolitan jurisdictions was a little bit more than 50%. During 1945 to 1960, the approval rate increased to around 70%, which might relate to the rapid urban deve lopment after World War II. From 1960 through 1990, the approval rate remained in the range of 70 80% in different levels of jurisdictions ( Owens, 2004 ) In New Yo rk City, the approval rate of variance was as high as 93% during 2001 2002, much higher than the 84% in recorded 1976 ( "Zoning Variance and," 2005 ) Sampson (2007) listed 14 studies on variance appr oval rates in different regions from 1938 to 2004, and showed in most areas that the approval rates were more than 50%. In seven studies, the rates were more than 70%. Of the 14 total studies, 12 were conducted before 1970, and two were conducted in the 19 80s, with the (2004) the study period was 1960 1990. Based on the many statements about approval rate, the first hypothesis is listed as follows: Hypothesis 1 Granting of variance a pplication is significantly higher than denial. It is nearly common consensus that approval rates are the highest that they have been since 1920s. However, study of variances approval rates from 1990 2010 is rare.

PAGE 55

55 few had long range study periods. The change and trends of approval rates in an area are rarely noticed. In addition to testing the hypothesis with regard to total number, this project also tests the hypothesis over time. Opponents from A ffected N eighborh oods important factor that affects the reaction of opponents. The appearance of protesters at public hearings has exercised c onsiderable sway over appeal board decisions in this country. Special interest groups and public hearing protesters might in large part balance each other in a system which indulged both ( "Administrative Discretion in Zoning," 1969 p. 680 ) In the absence of opponents in a public hearing, the ratio for granting variances was three times more than the ratio of opponents that appeared in the City of Philadelphia during t he study period from July 1954 through September 1954 ( "Zoning Variances and, 1955 ) The Zoning Board of Adjustment in Philadelphia gra nted 77% of applications without opponents present, compared to 24% with opponents present. In Boston, 81% were granted without opponents present, but 60% with opponents present ( "Administrative Discretion in Zoning," 1969 ) Leary (1958) said "it is frightening to think that the criterion used by the Board of Appeals for approval or disapproval of variance applications is the presence or absence of protests" (p. 14) Sha piro (1969) also commented that the absence of opponents would be one reason that the Board granted variances in Baltimore. One board official stated that the presence of protestants is the one factor which most frequently causes the Board to adhere to l egal requirements for variations

PAGE 56

56 ( Shapiro, 1969 p. 14 ) Madry (2007) cited "Wald's Statewide Rule of Four" 1 to support Wisconsi n cities [ ,] towns [ ,] and villages", and says that : I f four or more persons appear at the ZBA 2 variance hearing in person or in writing objecting to the application, then the variance is always denied. The ive across town. There were only one or two exceptions to this rule in 85 communities over four years. (p. 488) It is uncertain if the appearance of opponents denotes the application of zoning variance is incompatible with the local community. The opponent s might represent individual interests based on individual preferences or they might represent an interest the BZA to judge the influence the variance might cause ( "Administrative Discretion in Zoning," 1969 ) Since the previously mentioned studies did not consider other factors that might also have affected the results, they do not prov ide convincing evidence about Hypothesis 2 Opponents from affected neighborhoods influence zoning The opponents here could be neighbors, community org anizations, and other related stakeholders 3 They could choose to appear in the public hearing or send a letter expressing their objection. If enough required data about above information could 1 Wald Klimczyk is a city attorney in Janesville, Wisconsin. This rule was presented by him in the presentation Variances of Zoning Code Requirements: Current Practices by Wisconsin Cities, Towns & Villages (June 2000) (prepared for presentation to the League of Wisconsin Municipalities Institute). 2 Here ZBA refers to Zoning Board of Adjustment. 3 Oppositions by public agencies are n ot included here. Hypothesis 4 deal with the attitude of public agencies and their influences.

PAGE 57

57 be acquired, this hypothesis could be tested using regression analysis. In addition, if this causal relationship exists to a significant extent, the degree of this relationship could also be estimated. Area V ariance vs. Use V ariance Some s tates do not allow use variance. For the states that allow use variances, it sh ould be noted that the BZA is reluctant to grant use variances compared to area variances ( Burke & Snoe, 2004 ) It is stated the use variance deviates from the purpose of zoni ng ordinance more than area variance. As mentioned previously, some states apply the standard of practical difficulty on area variance, which is already a less stringent standard in jurisdiction compared to use variance ( Salkin, 2008 ) Considering the above statements, a third hypothesis is will be tested: Hypothesis 3 Type of variance application affects the making. This hypothesis only applies to th e states that authorize use variance. It is not yet clear to what extent the BZA would grant area variance versus use variance. As mentioned above, some states apply the same standard to test both area variance and use variance application; some states alr eady apply stricter standard to test use variance than area variance. No matter what standar d they apply, whether use variance is less likely to be granted than area variance in terms of the is a way we can apply in both situations. It is no t required by regulations that type of variance should be one criterion when the BZA make its decision Suggestions from Other Public A gencies Some states require related public agencies and associations to submit analytical reports with recommendations fo r the

PAGE 58

58 agencies include planning agencies, transportation agencies, public work agencies, community associations, or historical preservation commission. In Washington, D.C., it is the responsibility of the Offic e of Planning (OP) to assess variance applications and present reports to the BZA. For some large development projects, the Office of Transportation also conducts analysis about congestion, parking and other traffic issues. As an outlet for citizen partici pation, the Advisory Neighborhood Commissions (ANCs) 4 also conduct meetings and propose their concerns to the BZA. It is often written in the BZA orders that it had given great weight to the OP and the ANCs reports for final decision making. However, it is not a regulatory prerequisite to have an ANC report on file before the BZA makes a decision. recommendations on the the advice of planners seem s to have little effect on (p. 11) the board in Baltimore. The city's planning department might solve the problem that the members of the BZA do not have enough knowledge in architecture and planning by cooperating closely with the appeal board, but su ch coo rdination d oes not always occur ( "Administrative Discretion in Zoning," 1969 p. 674 ) ( "Administrative Discretion in Zoning," 1969 ) In New York City, though City Planning is authorized to give recommendations to the board, it exercises that authority infr equently ( "Zoning Variance and," 2005 p. 213 ) In the study of misrule, when planning staff recommended for approval, the decisions by board were always ntrast, the board had sharp disagreement with planning staff when the planners did not support the applications. The planning staff 4 The District is divided into 37 ANCs.

PAGE 59

59 recommended denying 75 applications in a total of 102 applications during 1960 1961. However, the board only denied 26 ( Dukeminier & Stapleton, 1962 ) In a survey about variance decision making in North Carolina conducted by Owens, within the 441 y zoning boards ( Owens, 2004 ) However, it does not mean the final decisions were contrary to other recommendations. This percentage means 7% of zoning boards highly correlated to those recommendations. For other zoning boards, though their final decisions might be the same as those recommended, their decisions were often based on other considerations. Most of the existing literature indicates that the BZA are reluctant to accept recommendations from other public agencies, but some states require the BZA to as follows: Hypothesis 4 The zoning administrative board does not place significant weight on the suggestions of other public agencies in its decision making The ideal way to test this hypothesis is to conduct a survey about the weight that members of t he (2004) study. However, the results observed using this method might not be true, since members might not reveal what they really think. Another method involves probing the correlation the The deficiency of this method is also apparent: one can know the correlation between

PAGE 60

60 the say that the In addition to the above issues in the determination of zoning variances, there are other factors that might affect the final decision. As mentioned above, the political and social backgroun d and subjectivity of board members could be important factors in their decision making. However, these are hidden factors, since it is hard to examine such causal effects. The high value of proposed buildings might be approved for zoning variance more eas ily than those of lower value. In Boston, the board granted eighty eight percent of requested variances for buildings with an estimated cost of over $10,000, as compared with seventy four percent of those for buildings costing less than $10,000. Even more strikingly, the applications granted comprised only eighty one percent of the buildings by number, but ninety seven percent of the dollar value of the estimates. ( "Admin istrative Discretion in Zoning," 1969 p. 675 ) Also, Owens (2004) found the population of cities has positive correlation with the approval rates. Limited by the methodology and data available to the author, some factors cannot be tested. For a more com prehensive understanding of zoning variance decision making, other efforts will need to be conducted. Binary Response Model Introduction of Binary Response Model In social science, researchers are always interested in the factors that affect the final deci sion of an individual or a group. Some decisions only have two choices. For example, a family decides to buy a house or rent a house; a student decides to go on with higher education or not; or a dissertation supervisory committee makes the doctoral

PAGE 61

61 studen t to be a candidate or not. A binary response model (or binary choice model) is usually used in the above econometric analysis in which the dependent variable is a dummy variable that takes only two values, 1 or 0. The mathematic expression of a binary res ponse model is usually as follows: P denotes the probability when the dependent variable equals to 1; x denotes the vector of independent variables and i is the number of independent variables; is vector of parameter; F denote s the functional form of independent variables. The simplest functional form is linear regression: In practice, the linear regression model is no longer applied since it has serious shortcomings. In addition to being a linear regression model, a binary r esponse model is referred to as : a probit model if F is the cumulative normal distribution function. It is called a logit model if F is the cumulative logistic distribution function. The logistic and normal distributions are both symmetrical around zero a nd have very similar shapes, except that the logistic distribution has fatter tails. As a result, the conditional probability functions are very similar for both models, except in the extreme tails ( Horowitz & Savin, 2001 p. 44 )

PAGE 62

62 Model Building A binary response model is a good model to be applied in this study since decision making on zoning variance applications has two different results: grant 5 or deny. In regards to the functional f orm of model, we can use both probit and logit models. In Chapter 5, how to design and choose model will be explained in detail. According to the four hypotheses proposed in the last section of this chapter a simple binary response model is built as follo ws: O: opponent S: suggestion from other public agencies T: type of variance (use variance, area variance) This basic model is a first step for further investigation of the factors that affect the ine other possible factors, an improved model is built for testing: O: opponent S: suggestion from other public agencies T : type of variance (use variance, area variance) A: property area V: land value TP: existing type of property (residential, comme rcial, mixed use, other) TAP: type of variance applicant (individual, firm, other). 5 Here conditional approval is categorized as grant.

PAGE 63

63 Case Study Method Case Study The implementation of c ase stud ies is a widely used research strategy in social science research Briefly defined a case study is an empirica l inquiry that investigates a contemporary phenomenon within its real life context; when the boundaries between phenomenon and context are not clearly evident; and in which multip le sources of evidence are used ( Yin, 1984 ) A research er may apply a single case study or a cross case study method based on the research questions Cross case stud ies can be used for generalizing commonalities. However, we should keep in mind that by using the case study method we can not come to a universal conclusion which is a noteworthy shortcoming of this technique Single case studies are not suitable for generalizing; howeve r, they are suitable for testing general theories from multiple sources. In addition, based on findings about general causal effects from cross case studies, the single case method is used to study causal mechanism ( Gerring, 2007 ) Th is research applies single case study method to analyze the influencing factors in zoning variance decisi on making. Triangulation 6 is used as a justification of the propositions of hypotheses. C a usal effects between independent variables and the final decision are apparent in existing literature By using single case study, this research initially tests the c ausal implications of existing general views about the influencing factors; second this research digs into the causal mechanism. To what extent do these factors affect decision making is one of the research questions. Furthermore, as a 6 od s of gather similar information ( Byrne & Ragin, 2009 p. 343 )

PAGE 64

64 revelatory case stu dy, it provides a new perspective on how other factors affect the Study Area Washington D.C. is the case study area in this research. The author chose this city based on the considerations of the specialty and nonspecial ty of this a rea First, Washington D.C. is one of the first cities which initialized comprehensive zoning ordinances in the United States. Regulations on zoning and zoning variances are highly developed. The related agencies are created for specific responsibilities a nd have a specific framework and timeline in the procedure. In terms of the zoning variance process, there is no apparent difference from other cities. This normative attribute makes this city be representative of most other areas in the United States. Sec ond, Washington D.C. has a special attribute suitable for a study area in this research. It is a n old and highly developed city in regard to land use The n eed s for r edevelopment occur throughout this city in recent years as the old buildings be come increa singly unsound in structure and irrelevant in function It is appropriate to study the trends related to variance application and decision. In addition, n eighborhood characteristics are distinct ly different in terms of dwelling types, property values, hist oric preservation, and household income. It is also a n ideal place to study the social and economic factors which might affect the making spatially. Figure 3 1 shows the conceptual framework of this paper. Data Collection and Data Processing Successful statistical analysis and geospatial analysis in this paper rely highly on complete and reliable data sources. The author contacted the Office of Zoning and the Office of Planning in Washington D.C. for compiled zoning variance applications. Both

PAGE 65

65 have documented zoning orders, which are open to the public and c an be accessed through the official website of the Office of Zoning 7 The documented zoning orders include detailed information o n each variance ap plication, which contains applicant s name variances requested, property address, SSL (Square, Suffix, Lot) 8 hearing date, the OP the ANC s and other public agencies, neighbor the However, no effort has been made before to compile the aforementioned information together. Alt hough every case is unique in its conditions and variances requested, we can record the main common variables of each c ase. The author recorded 15 variables and a total number of 2140 cases from 1980 to 2009 according to the zoning orders provided online. All the data w ere saved in a table in Microsoft Excel format. Some variables might not be used for the final analysis; however, they were recorded in the original table for further reference. These 15 variables are as follows: zoning order number, hearing date, decision date, SSL, zone d istrict 9 type of property, type of applicant, variance requested 10 recommendation fro m Office of Planning, recommendation from t he ANCs recommendation from other public agencies, the opponents present at the public hearing, the vote s and 7 http://dcoz.dc.gov/search/searc h_orders.asp 8 SSL is a coding method to identify property in Washington D.C. In most cases, SSL is corresponded to have multiple addresses located on it. This often includes garden style apartment complexes as well as corner addresses with separate addresses facing each adjacent street. One address can al so sit upon multiple properties and o ne single family residence can sit upon multiple lots. The add ress records only ( District of Columbia O ffice of the C hief T echnology O fficer 2009 ) For the cases which had more than one SSL, the author recorded the first 9 Zone district mean regulations governing the use of land and the use, density, bulk, and height of buildings an d other structures are the same ( "Zoning Regulations ," 1958 p. 3 ) 10 A table in Appendix B shows detailed information o n variance requested, compiled by the author.

PAGE 66

66 standard ized before 2002 information o n variables could not be found in some cases. Spatial data were compiled and created for geospatial analysis. There are two ways to geocode the location of pro perties in this research. The first way is to record the addresses of properties and purchase the geocoding se rvice to geocode these data. P roviders of geocoding service s such as Geocoder, EZ locate, and Maponics c an be found on the internet The second w ay i s to relate the Excel table to existing shapefile. Since we do not have available address list s for all the applications, more time was expended to record all the addresses. Besides, purchasing online service s are more costly Since some shapefiles con tain ing the information of SSL are available on the D.C. government website we can connect the Excel table to these shapefiles and configure the shapefile to indicate the location of variances. The owner point shapefile, record lot 11 shapefile, and tax lot 12 shapefile c an be used for connection. The owner point shapefile is supposed to contain all the properties in Washington D.C. The author first connected the table with these three shapefiles separately, but later found a great amount of cases which could not be found in the above three shapfiles. The failure of connection stems from the subdivision of record lot/tax lot. 11 atory Affairs (DCRA) Office of the Surveyor (OS) DC Surveyor. They are official, platted, recorded subdivision lots created by the D.C public street building permit will be issued for that site in the District of Columbia, and all proposed Record Lots are carefully reviewed by Zoning Administration officials for c Record lots are defined only when requested by property owners, normally when they are seeking a ( District of Columbia O ffice of the C hief T echnology O ffi cer 2008 ) 12 nes tax lots under two circumstances: 1) when property owners ask for their real property tax bills to be consolidated, after they have bought several contiguous record lots; this is called a combine; 2) when part of a record lot is sold, but no new record lot is yet defined; this is called a split request. Tax Lots are not normally acceptable when applying for building permits and must be converted to Record Lots through the normal subdivision ll be issued. The only exception is if the lot ( District of Columbia O ffice of the C hief T echnology O ffi cer 2008 )

PAGE 67

67 Some lots were subdivided and new SSLs were allotted to the newly subdivided lots. SSLs for the old lots do not exist anymore. Another ex planation is that some data were not simply recorded in the shapfiles. In order to geocode all the cases, the author took five steps to finish geocoding process. Step 1 c onnect s SSL in the variance table (Excel format) with SSL in owner point table (ArcGIS format). 1408 cases were matched, in which 98 locations evinced at least two variance requests A new shapefile was created named ownerpt_match. This shapefile contains geospatial location, owner information and variance information. In s tep 2 s ome SSLs i n the owner point shapefile are also included in record lot shapefile. In the case of overlap happens, the author first deleted the matched 1408 cases in the variance table and then connected the left cases with record lot shapefile. 152 cases were matched in this step, in which 11 locations met variance requests twice or more. Since the record lot shapefile is a polygon file, we must convert it into a point file so as to be consistent with the owner point shapefile. A new point shapefile named recordlot_ma tch was created, which contains geospatial information, record lot information and variance information. Step 3 a ppl ies the same process i n Step 2 to tax lot shapefile. Two cases were matched and a new shapefile was created called taxlot_match. After this step was done, 1562 cases were matched, while 578 cases were left unmatched. In s tep 4 the U S addresses of the unmatched 578 cases were input into a n ITouchMap 13 coordinate converter and to obtain the latitude and longitude of all the 578 properties. ArcGIS functions to locate point s according to their latitude and longitude. A new shapefile named other_match was then created. 13 http://itouchmap.com/latlong.html

PAGE 68

68 Step 5 c ombine s ownerpt_match, recordlot_match, taxlot_match and other_match into one shapefile and name s All th e variance application cases are geocoded in ArcGIS.

PAGE 69

69 Figure 3 1 Conceptual Framework of this P aper

PAGE 70

70 CHAPTER 4 C ASE STUDY W ASHINGTON D.C. Introduction of Washington D.C. Washington, D.C. has been the capital of the United States since 1790. As one of typical example of a mix of both old and new elements, has good communities, historic architecture, modern service, and g reat parks and waterfront. More than 20 million visitors visit D.C. each year for cultural, commercial, and political exchange. The District is unique in that it operates simultaneously as a city, a state, and as the seat of federal government ( District of Columbia Office of Planning 2007 p. 2 ) The District is located in the mid Atlantic region of the United States, which is bordered by the states of Virginia and Maryland. The boundary lines are straight and Potomac River is a natural boundary on the one s ide ( Figure 4 1 ). The total area of the District is 68.5 square miles. The population now is around 0.6 million, of which 40.1% are white, 50.4% are black and 9.5% are others. In recent years, the population has gr own at an average rate of 0.39% since 2003. The median household income is $58,553 in 2008 1 City Planning in Washington D.C. first and most comprehensive plan ever designed for any city ( Caemmerer, 1939 ) The new national capital was parks, as well as the Capitol and the White House. Influenced by City Beautiful 1 Source: U.S. Census Bureau

PAGE 71

71 Movement, the McMillan plan endeavored to develop the park system in the District, which focused on the open space of the National Mall, and neighborhood parks ( District of Columbia Office of Planning 2007 ) These two plans paved the road to further design and improvement of the Dis trict in the following century. In 1950, the first Comprehensive Plan was produced by the National Capital Park and Planning Commission (NCPC), which was created in 1924. The Comprehensive Plan is a guideline for the future land use and development for a long time frame. This plan put significant efforts toward housing and transportation. In 1961 and 1967, another two Comprehensive Plans were established by the NCPC, including landscape change, expansion of central business district, and other urban renewa l projects ( District of Columbia Office of Planning 2007 ) In 1973, the District of Columbia Home elements and the Federal elements. The former are prepared by the of Planning under the administration of the Mayor; the latter are prepared by the NCPC. The Mayor shall be the central planning agency for the District. He shall be responsible for the coordination of planning activities of the municipal government and th e preparation and implementation of the District's elements of the comprehensive plan for the National Capital which may include land use elements, urban renewal and redevelopment elements, a multi year program of municipal public works for the District, a nd physical, social, economic, transportation, and population elements ( Office of the General Counsel 1973 sec. 423 )

PAGE 72

72 The first new Comprehensive Plan under this Act was adopted in 1984 and was amended periodically as needed. The 1989 and 1994 amendments added Ward Plans 2 The most recent 2006 Comprehensive Plan, which was amended in 2009, includ es 13 citywide elements and 10 area elements. The citywide elements are: Land use Transportation Housing Economic development Parks, recreation, and open space Educational Facilities Environmental protection Infrastructure Urban Design Historic Preservati on Community services and facilities Arts and culture Implementation The 10 area elements contain: Capitol Hill Central Washington Far Northeast and Southeast Far Southeast and Southwest Lower Anacostia Waterfront and Near Southwest Mid City Near Northwe st Rock Creek East Rock Creek West 2 as appropriate for each of the eight legally defin ed geographic areas of the city ( District of Columbia O ffice of P lanning 2007 ) congressional legislation authorizing election of members to the District of Columbia Board of government redraws ward boundaries after each decennial census, if necessary, to ensure that ward populations are as near to equal size as possible. When boundary changes were made in 1982, the Office of Planning prepared the preliminary ward boundaries, as closely to natural neighborhood n redrawn four times, after the 197 0, 1980, 1990 and 2 000 censuses ( District of Columbia O ffice of P lanning 2007 p. 27 )

PAGE 73

73 Upper Northeast 3 The Wards Plans were replaced with area elements in 2006 Comprehensive Plan. Yet in practice, some planners and scholars still use Wards for geographic divisions in the District due to its long history o f application. Zoning in Washington D.C. History of Zoning As early as the establishment of Washington, D.C., President Washington proclaimed a height restriction of 40 feet on new buildings. In 1910, a comprehensive height regulation was established ( Caemmerer, 1939 ) It could be seen as the early forma tion of the conception of zoning. The District initialized the first comprehensive zoning ordinance, the Zoning Act of 1920, right after the first comprehensive zoning ordinance was enacted in 1916 in New York City. Types of use districts, height, and lot occupancy were regulated in this ordinance, which can be seen in three separate maps. The Zoning Commission was established and its basic structure was formed. The Zoning Act of 1938 was an extension and clarification of the Zoning Act of 1920. This Act es tablished the police power of the Zoning Commission and its responsibility. T he B oard of Zoning Adjustment (BZA) was also created in the Zoning Act of 1938. In the 1950 s a Comprehensive Plan suggested completed revisions on the zoning regulation map and t ext. In addition, suggestions regarding commercial zoning, off street parking and loading, and some other detailed issues were proposed. The Zoning Ordinance of 1958 adopted most suggestions from the Lewis Plan of 1958 4 in which Harold Lewis, who was one of and consultant s ( Peel, 1939 p. 3 ( District of Columbia O ffice of P lanning 2007 ) 4 Harold MacLean Lewis, A new zoning plan for the District of Columbia final report of the rezoning study, 1956, New York

PAGE 74

74 158 ) suggested a major zoning overhaul ( Kress, 2001 p. 2 ) including the responsibility, unification between zoning districts and comprehensive planning, a floor area ratio system, parking restrictions etc. The Zoning Ordinance of 1958 is still th e guidance document for the Zon ing Commission and the BZA In 1990, the Office of Zoning was created according to the Office of Zoning Independence Act of 1990 to assist the Zoning Commission and the BZA ( Kress, 2001 ) Zoning Regulations and Administration Zoning regulations in Wa shington, D.C. are indispensible as a component of the municipal regulations. Several public agencies are involved in the administration of zoning regulations. Zoning re gulations are set forth in Title 11 of the District of Columbia Municipal Regulations and the accompanying zoning maps. Zoning authority is exercised by the Zoning Commission of the District of Columbia and the District of Columbia Board of Zoning Adjustme nt. The Home Rule Act requires that zoning not be inconsistent with the Comprehensive Plan ( District of Columbia Office of Planning 2007 p. 25 ) Currently, the District has 34 zoning districts and 27 overlay districts. The zoning districts, which are basic di visions, refer to the different use of districts. Overlay districts could be combined with zoning districts to add special regulations onto the existing zoning districts. The zoning districts are divided into six categories: Residence District Special Pu rpose Districts Mixed Use Districts (commercial, residential ) Commercial Districts Industrial Districts Waterfront Districts

PAGE 75

75 In addition, there were subcategories in some of above categories. The overlay districts, which have the same structure with th e zoning districts, have 11 categories and some subcategories. These categories are: Langdon Overlay District Mixed Use Diplomatic Overlay District Hotel Residen tial Incentive Overlay District Ca pitol Interest Overlay District Neighborhood Commercial Overlay Distric ts Reed Cooke Overlay District Miscellaneous Overlay Districts Downto wn Development Overlay District Uptown Art Mixed Use Overlay District Capit ol Gateway Overlay District Southeast Federal Center Overlay District. There are four ma in public agencies involved in the zoning regulation and administration. They are: Office of Zoning (OZ) Office of Planning (OP) Advisory Neighborhood Commissions (ANCs) Zoning Administrator (ZA) 5 Inside the OZ, there are two important agencies set u p to exercise zoning authority: the Zoning Commission (ZC), and the Board of Zoning Adjustment (BZA). Table 4 1 s function s Zoning Variance in Washington D.C. Regulation of Zoning Variance Zon ing ordinance in Washington D.C. recognizes both area and use variance. It also recognizes two standards for peculiar and exceptional practical difficulties or exceptional and undue hardship 6 in its regulation and in administration. 5 Zoning Administrator is inside Department of Consumer & Regulatory Affairs (DCRA).

PAGE 76

76 Under § 8 of the Zo ning Act, the D.C. Official Code § 6 641.07(g ) ( 3 ) ( 2001), provides that the BZA may approve the zoning variance application : [ W ]here, by reason of exceptional narrowness, shallowness, or shape of a specific piece of property at the time of the original a doption of the regulations, or by reason of exceptional topographical conditions or other extraordinary or exceptional situation or condition of a specific piece of property, the strict application of any regulation adopted under D.C. Official Code §§ 6 64 1.01 to 6 651.02 would result in peculiar and exceptional practical difficulties to or exceptional and undue hardship upon the owner of the property, to authorize, upon an appeal relating to the property, a variance from the strict application so as to rel ieve the difficulties or hardship; provided, that the relief can be granted without substantial detriment to the public good and without substantially impairing the intent, purpose, and integrity of the zone plan as embodied in the Zoning Regulations and M ap. 7 Form 121 A pplicant s Burden of Proof for Variance and Variance Special Exception Applications establishes two categories of variances: area and use. These varian ce types are defined as follows : A n area variance is needed when the owner wishes to mak e some change to the physical structure or lot itself and the property does not or will not comply with the Zoning Regulations in some respect. A use variance is needed when the owner wishes to use the property in a way that is not permitted in that zone d istrict 6 Sour ce: 11 DCMR § 3103.2. 7 See also 11 DCMR § 3103.2

PAGE 77

77 under the Zoning Regulations ( District of Columbia o ffice of z oning 2008, para. 1) Three criteria 8 must be met for the grant of a use or area variance: 1. Are there peculiar and exceptional, practical difficulties, such as the property is exceptionally narrow, shallow, oddly shaped, and/or has unusual topography, soil conditions, or other special conditions: a. What makes it difficult for the owner to build on the property in compliance with the Zoning Regulations? (area variance) b. How will there be financial hardship for the owner in usin g the property consistent with the Zoning Regulations? (use variance) 2. Granting the application will not be of substantial detriment to the public good; and 3. Granting the application will not be inconsistent with the general intent and purpose of the Z oning Regulations. ( Kress, 2001 p. 33 ) The board applies the interpretation stated by the D.C. Court of Appeals case of Palmer v. Board of Zoning Adjustment 287 A.2d 535, 539 (D.C. 1972) to te stify the proof of practical difficulties and undue hardship. should be construed to require a showing of a situation where in the absence of a variance of t he property cannot be reasonabl y used in a manner consistent with the ( District of Columbia C ourt of A ppeals 1972 para. 43 ) As for 8 Columbia Zoning Commission: "[i]n order to obtain variance relief, an applicant must show that (1) there is an extraordinary or exceptional condition affecting the property; (2) practical difficulties will occur if the zoning regulations are strictly enforced; and (3) the requested relief can be granted without substantial detriment to the public good and wi thout substantially impairing the intent, purpose, and integrity of the zone plan (para. 12)

PAGE 78

78 be shown that compliance with the area restriction would be unnecessarily (para. 67). The court s have clarified the tests for area variance and use variance In Palmer, the court opined that area variances have been allowed on proof of practical difficulty only while use variances require proof of hard ship, a somewhat greater burden (para. 45). The court agreed that a more stringent showing is warranted with respect to the more drastic re lief inherent in a use variance (para. 50). On the basis of the difference between practical difficulties and u ndue hardship the D.C. circuit made it clear that practical difficulties is applied to the criterion of an area variance and undue hardship is applied as the criterion of use variance. In regard to a use variance, the ordinance allows consider ation o f financial hardship Traditionally, financial hardship is not considered to be undue hardship or unnecessary difficulties in most states. An applicant is not entitled to the best and highest use of the land or maximum benefit from the parcel ( "Replacing the Hardship," 1987 p. 682 ) It was almost unanimously confirmed by courts that mere financial hardship or an increased return from the property is not a sufficient reason for granting a variance ( "Administrative Discretion in Zoning," 1969 ; Cohen, 1995 ; "Zoning Variances," 1961 ; "Zoning Variances and," 1955 p. 520 ) Financial disappointment alone, including loss of profits or prohibition of the most profitable use of the property, will not justify a variance ( Jacobs, 1958 p. 822 ) In the case Palmer v. Board of Zoning Adjustment, it was recogniz ed by the court regulations will produce a reasonable income but, if put to another use, will yield a

PAGE 79

79 greater return e or loss of economic advantage is not sufficient to constitute hardship ( District of Columbia C ourt of A ppeals 1972, para. 67 ) I looked through zoning orders in Washington D.C. and listed the following cases which are representative of the attitude of the BZA towards financial hardship. In application No. 11000 when the applicant proposed a nursing home in the R 1 B district, the board demonstrated his Board has held in the past and continues to hold that economic and financial potential gain or loss will not alone substantiate relief from the strict in ( District of Columbia O ffice of Z oning 1974 p. 4 ) In the case of application No. 14019 the applicant pursued a use variance for all day co mmuter parking and the application was granted. T he board confirmed that easonable interim use of the subject property than the continuation of the subject parking facility ( District of Columbia O ffice of Z oning 1983 p. 3 ) In the case of application No. 11511, the applicant proposed to use the property as a flat in the R Zoning Regulations would cause an economic loss to her by reason of her l arge ( District of Columbia Office of Zoning 1974 p. 3 ) The BZA poi nted out the misconception about the status of the property by the applicant before she purchased the property could not be the proof of financial hardship. The board denied her application. In the case of application No. 11925, the applicant asked for a use variance to provide a social service center in R 3 district. The conomic hardship or financial inability to develop property is not a basis for hardship as claimed by applicant, unless that economic burden is caused by topography or oth er exceptional

PAGE 80

80 ( District of Columbia Office of Zoning, 1975 p. 3 ) and denied the applicat ion. The applicant s in th e case of No. 13405 alleged financial difficulties and asked for a lot area variance. T financial difficulties alleged by the application are personal and are not derived out of th ( District of Columbia Office of Zoning 1981 p. 3 ) According to the cases and decisions above, financial ha rdship is based on the property itself rather than circumstances pertaining to the owner. It should be caused by topographic or other exceptional circumstances. Strict observance of zoning ordinance would lead to financial infeasibility of the property. Th e conception about financial hardship in Washington D.C. is consistent with other states/cities which allow financial hardship as the reason of applying for zoning variance. Zoning Variance Procedure The variance application process usually takes around fo ur months since the applicant submits an application to the OZ until the Final Order is received by the applicant. In reality, the applicant first applies for a Building Permit to DCRA and then ZA determines whether this application is allowed by zoning or dinance. If a variance application is needed, ZA provides the applicant with advice. The applicant prepares and submits variance application to OZ. Once OZ confirms the required application is complete, OZ schedules a public hearing before the BZA. In addi tion, the application is forwarded to the OP, the A NC s property owners within 200 foot and other DC agencies for review. The h earing data are then published in the DC Register. Meanwhile, the applicant must post placards on the site as a notice to the ne ighborhood. T he OP is required to submit a report with recommendation s after the OP staff meet and discuss the proposal with the applicant. The BZA is required under D.C. Code § 6

PAGE 81

81 9 to the The BZ A is also required under D.C. Code § 1 309.10 (d) 10 to t he ANCs s The BZA conducts a public hearing. ZC may announce the decision following the hearing or at a subsequent meeting. ZC approves the Final Order which is then published in the DC Register and made available on the OZ website. Finally, the Final Order is sent to the applicant. Figure 4 2 shows the flow chart of zoning variance process in Washington, D.C. 9 D.C. Official Code § 6 -Recommendations, reports, review all be construed to prevent the Office of Planning from continuing to provide recommendations and reports to the Zoning Commission and the Board of Zoning Adjustment on any zoning case. The Office of Planning shall review and comment upon all zoning cases, and the Zoning Commission and the Board of Zoning Adjustment shall give great weight to the recommendation of the Office of Planning. Upon request of the Zoning Commission or the Board of Zoning Adjustment, the Office of Planning shall provide recommendat ions, information, or technical 10 D.C. Official Code § 1 recommendations of the Commission shall be given great weight during the deliberations by the gov ernment entity. Great weight requires acknowledgement of the Commission as the source of the Act of 2000, effective June 27, 2000 (D.C. Law 13 135, 47 DCR 5519 (2000)) (to be codified at D.C. Code § 1 261(d) (3)(a)), the Board must articulate with particularity and precision the reasons why the ANC does or does not offer persuasive advice under th e circumstances and make specific findings and conclusions with respect to and, as explained in this decision, finds their recommendations to grant the applic ation persuasive. The ( District of Columbia Office of Zoning 2001 )

PAGE 82

82 Table 4 1 Overview of r elated z oning a gency r oles in Washington D.C.

PAGE 83

83 Figure 4 1 Map of Washington D.C. 12 12 The street map of Washington D.C. was downloaded from ESRI website: http://www.arcgis.com/home/item.html?id=3b93337983e9436f8db950e38a8629af ; All other ArcGIS shapefiles of Washing ton D.C. were downloaded from http://dcatlas.dcgis.dc.gov

PAGE 84

84 Figure 4 2 Zoning variance procedure in Washington D.C. 13 13 This graph was designed by the author and Lucy Elder when they had their internship on The District Office of Planning supervised by Travis Parker.

PAGE 85

85 CHAPTER 5 ANALYSIS OF CASE STU DY FINDINGS Zoning Variance Applications and Decisions A General View This study consists of a total of 2140 variance application cases from 1980 to 2009 1 Since c ases withdr a w n 2 by the applicants or dismissed 3 by the board do not reflect the BZA decision s they were not recorded in this study. Of the variance applications, 1824 ( or 85.2%) requests were for area variance s and 316 ( or 14.8%) requests were for use varian ce s Variance Applications Figure 5 1 shows the number of variance applications by type of variance and year. The number of variance application s hit a peak in 1980, 159 in total. Since the beginning of the 1990s t he number of application s decreased to usually less than 60; while during the 1980s, the number of application s was usually more than 60. This decreasing trend in the early 1980s is in accordance with the economic recession lasting from 1980 to 1982. The e nergy crisis of 1979 led to inflation in the U.S. a situation halted by the government s decision to implement a tight monetary policy; the crisis culminated in domestic u nemployment rate s reaching 10.8% in 1982. Taking the hysteresis effect on economy in to consideration, it is reasonable to explain the decreasing trend in variance application from 1980 to 1985. From 1990 to 1991, high oil price s led to yet another economic recession in the U.S. The decreasing number of variance application s since 1990 may be partly due to the economic downturn. In 1995, 1 Note: Data for 1985 for four months from Ma y 15 to Sep. 25 are not available form Office of Zoning website. 2 An applicant may withdraw his/her request without penalty. The applicant is not required to explain the reason of withdrawal. 3 The failure to submit required materials or the failure to pr esent in public hearing will lead to the dismissal of the request.

PAGE 86

86 this number reached the lowest in 30 years with 28 applications in total. Area variance applications were the main contributors ea ch year ; 74.5% to 97.1% of the total applications each year. Use variance re quest s usually occupied less than 20% of the total cases each year. In 1997, the number of use variances a pplications reached the highest : 25.5% In 2004, there were 2.9% use variance applications, the lowest during the 30 year period This graph shows the number of variance applications in recent years were less than in the 1980s. It is uncertain to predict the future trend of variance applications according to this graph. However, for more than ten years from 1996 to 2009 the number of applications rema ined reasonably stable The n eed for variances appears to remain steady and evinces no sign of decline Decisions from the BZA The final decisions from the BZA could be categorized in two simple results: approval and denial. This two choice decision outc ome will be use d later in the binary response model analysis. Within those 30 years, a total of 1816 (or 85 %) requests were granted by the BZA and 324 (or 15%) requests were denied. Consistent with other studies in other cities Washington D.C. shows a hi gh approval rate of variance applications ( Figure 5 2 ) The average approval rate of variance s was 84.9% since 1980. T his a pproval rate was more than 70% each year except in 1982. There were two periods that the a pproval rates kept under 80%: from 1980 to 1983, and from 1992 to 1994. Since 1995, the approval rate remained above 80% each year ; s inc e 2005, the approval rate exceeded 90% each year. Before 1995, the approval rate fluctuated but this situation changed since 1995 namely approval rate increased gradually since 1995.

PAGE 87

87 When isolating area variance s and use variance s the approval rate of the former was 87.9%; whereas the latter was 67.7% or 2 0% lower than the rate of area variance s The approval rates of use variances were usually lower than area variances, ranging from 23.5% to 100%. It is apparent that the approval rate s of use variance s fluctuated more than th ose of area variance s A possible explanation for this difference in approval rates is that the number of use variances each year was relatively small There are no clear trends of approval rate for either area variance or u se variance. However, the chart ( Figure 5 2 ) shows that the approval rates in recent years were more ste ady and higher than the rates of the 1980s. It is unknown why approval rates are so elevated It could be that the BZA members tend to be lenien t ; or that other public agencies perform as filters which receive variance applicants and provide them with c onsultation before they pursue the next step of the zoning variance process; or perhaps applicants or their agents are simply prudent The estimated variance application process itself takes around four months, not including the preliminary work done when the applicant submits an application. T he burden of proof is largely on t h e applicant so it is imperative that the applicant pursues a variance for which there is reasonable certainty of approval Time, money and energy are considerations that often lead to prudent decisions. Area Variance Approval By examining the approved cases in depth, the requests can be grouped into three categories: full approval, conditional approval and partial approval. Full approval refers to the case in which all aspects of the pursued variance request s are approved without condition. The term full here has two meanin gs : t he first meaning is that the BZA approves the proposal as submitted and does not require additional conditions as the

PAGE 88

88 premise of granting the variance ; t he second means that the BZA approves all the proposed variances. It is common that an applicant might propose several variances in one application. Conditional approval means that the BZA approves an application subject to some specific conditions. For examp le, in the case of Zoning Order 15605, applicants propose d t he use of the first floor as an office space in a 2 story apartment building in a low density residential area. The BZA granted this application subject to the following five conditions: (1) the u se shall be limited to the specifications in the application: apartment management office use and administrative offices for a tour company; (2) No exterior sign is allowed for the advertisement of these offices; (3) Number of employees shall not exceed fo ur; (4) Commercial vehicles are not allowed to park on the site. (5) No loading or unloading shall be allowed on the site ( Distr ict of Columbia O ffice of Z oning 1992 ) Partial approval refers to case s, in which an application is granted partially. This can occur when the applicant has submitted at least two or more variance requests and where the BZA denies at least one of the proposed variance requests Figure 5 3 shows the ratio of different approval types in the approved area variance cases by every five year s In most years the BZA gave full approval to the majority of the applications which usually occupied more than 60% of the total approved cases except in the period of 1985 1989 Conditional approval was the second largest category, which usually occupied more than 20% of the approved cases. Partial approval was rarely observed in the decisions. There were 9 cases identified as partia l approval within this 30 year period, which represented 0.6% of all applications from 1980 through 2009. In the approved area variance application, the average full

PAGE 89

89 approval rate was 70.1%, and the average conditio nal approval rate was 29.3% from 1980 through 2009. This diagram ( Figure 5 3 ) shows that the proposals for area variance applications were more likely to be accepted fully. The high approval rate for applications most receiving f ull approval reflect s that the BZA decisions on area variance applications were not as strict as prescribed by regulations. Use Variance Approval F igure 5 4 show s the use variance approval ratio for the approved a pplications by ever y five year s. It demonstrates a different pattern compared to area variance approved applications. The percentage of use variance requests receiving conditional approval wa s usually more than 40%, considerably higher than that of area va riances. The average conditiona l approval rate for the 30 year period was 49.5% of the use variances requests and the average full approval rate was 50%. For the year s 1980, 1981 and 2004, all the approved use variances were granted conditionally. This gr aph indicates that in granted cases, the BZA showed a more cautious attitude towards use variance cases than in area variance cases. The combination of a lower approval rate for use variance applications and a lower full approval rate for approved applicat ions partially reflect s similar finding s by other scholars: the BZA is more reluctan t to approve use variance s than area variance s Additional analysis must be applied to ver if y whether this statement is accurate in terms of statistical significance. In Ch apter 5 a statistical method was applied and the result revealed a significant difference. Fro m 1980 to 2009, the BZA gave 58% variance applications full approval, and 27% received conditional approval ( Figure 5 5 ) There were 10 cases of partial approval. Alt hough full approval constituted more than half of all application s it should be not e d that this does not entail that the original submissions were granted by the BZA without

PAGE 90

90 any revision s Before final submis sion to the BZA, applicants can engage the OP staff for policy explanation s and other forms of guidance. Some original requests might be deleted or revised if there is suspicion that they would probably be deni ed Some withdraw n a nd dismissed cases also ha ve similar reason s, namely that applicants become aware that they lack proof to jus tif y their burden after discussi ons with the OP staff or other public agencies. When separat ing area variance requests with use variance requests the final decisions are qu ite different Area variance requests had a denial rate of 12% compared to 32% of use variance requests ( Figure 5 6 ) These two pie charts show that use variance requests were more likely to be denied by the BZA, which is consiste nt with previous studies conducted by other scholars in other metropolitan regions For area variance requests, 62% received full approval and 26% received conditional approval. For use variance requests 34% received full approval and 34% obtained conditi onal approval. It shows that for use variance even the approved cases were less l ikely to be granted without condition s One possibl e explanation is that the court requires a somewhat greater burden on use variance than area variance, as discussed in Chapt er 4 In Washington D.C., an applicant for an area variance must prove the existence of practical difficulties By contrast an applicant for a use variance must prove the existence of an undue hardship Of the 316 use variance applications, 203 ( o r 64.2% ) were for commercial use Sin ce 7% of the property in Washington D.C. is commercial, there was a disproportionately higher application rate for use variances by commercial property owners ( the percentage is show n in Figure 5 12 ). Of these 203 applications, 149 ( or

PAGE 91

91 73.4% ) properties were located in residential zones. Of these 149 applications, 67 ( or 45.0% ) received conditional approval, and 77 ( or 51.7% ) were denied by the BZA. The relative denial rate may reflect the recognition of the negative effects of use variance s of commercial properties in residential zones On the other hand, the negative effects brought about by the use variance s of residential properties in residential zones or commercial zones are likely to be viewed less adversely Since commercial activities in residential zones are not in accord with the primary theme of zoning ordinance s, they m ight cause more negative effect s, resulting in the BZA denial of applications or the placement of addition al conditions on use variance requests to restrict their adverse impact on residential neighborhoods Office of Planning Involvements As mentioned in the section o n Zoning Variance in Washington D.C. in this chapter, the OP plays a very important role in providing recommendation s techni cal support and offering other rel e vant information to the BZA 4 It is required by law that the BZA should give serious consideration to the OP recommendations. There is no standard requirement for documenting the activitie s of the OP in the minutes of public hearin gs and published the BZA orders. Th is author compiled all the cases from 1980 through 2009, and the results for the activities of the OP are shown in Figure 5 7 The activ ities of the OP are divided into five categories: support, opposition, other, no action, and NA. Support comprises full support, conditional support, and partial support. Opposition means the OP recommend s that the BZA deny the application. Other signifies the submission of an OP report with concerns or recommendations with no 4 See also note 10 in Chapter 4.

PAGE 92

92 clear stance. No action means it is recorded in the document that the OP did not prepare a report for submission to the BZA for reference. NA. indicates that the documents do not have information about whether the OP submit ted a report or give suggestions to the BZA During the period f rom 1980 to 2000, the majority of the documents did not record whether the OP submitted a report, or whether the applications were supported/opposed by the OP. T his situation changed in 2001 when the recommendations of the OP were documented. In addition, the OP reports were clearer in terms of its stance ; whether to support or oppose a variance request S ince 2005, the involvement rate of the OP was 100% each year. This graph shows that during the past decade, the documentation of involvement and recommendations have become standardized. It was found in the year 2000 that the Office of Planning drafted a handbook fo r zoning planners: Zoning Application Processing. In this book, detailed explanation s are provided about the zoning process and the timeline for zoning adjustment, variance, and special exception s which specif y the handbook is essentially a guide on the zoning process for the internal use. Its issuance of this handbook could be the reason that in recent years the OP involvement has both increased and standardized. The Involvements The setting of the ANC s in Washington D.C. is a peculiar feature com pared with other cities. The y are formalized neighborhood associations, composed of elected residents from each neighborhood. The basic role of th ese immensely important bod ies is to consider a wide range of p olicies and programs affecting their neighborhoods, including traffic, parking, recreation, street improvements, liquor licenses, zoning,

PAGE 93

93 economic development, police protection, sanitation and trash collection, and the District's annual budget. 5 As show n in the section o n the Zoning Variance in Washington D.C. in Chapter 4 local regulations require the BZA to give serious consideration to the ANC s recommendation during the decision making process 6 The ANCs involvement in zoning variance application d ecisions has two major advantages. First, it provides local residents with the opportunity to particip ate in governmental decisions, which helps to prevent abuse of power in the variance process. Second ly compared to individual participation in public hea rings, the ANCs recommendation is relatively neutral which represents the whole welfare of the local community. As shown in Figure 5 8 t he ANCs involvement has also increased graduall y since 1980, from around 50% at the beginn ing of the 1980s to around 75% during the 2000s. Between 1980 and 2000, the involvement rates were higher than those of the OP. However, it is not reasonable to conclude the ANCs took more activities in the whole process during that period. Lack of standar dized documentation might lead to inaccurate conclusions o n OP and ANCs activities. Though the rate of no action has decrease d considerably in recent years, the involvement of the ANCs is not been comparable to that of the OP since 2000 There were certa in instances where no ANCs report was submitted to the BZA. Since 2001, the average rate of no action is still around 20%, whic h is more than that of the OP. The ANCs are established to speak for the non political representative s of the who le community The ANCs are com posed of local residents in area neighborhoods and are to ensure that the citizen s voice s are heard in the political decision making process. No action from an ANC indicates in some degree that there 5 Source: Website of the Office of Advisory Neighborhood Commissions. http://www.dccouncil.washington.dc.us/officeoftheadvisoryneighborhoodcommissions 6 See also note 11 in Chapter 4.

PAGE 94

94 has been acquiescence to the applicati ons. An OP report is more technical oriented and focuses on whether the request satisfies the criteria of variances. These report s concentrate on neighborhood attitudes toward the application. Figure 5 8 shows that in recent years ANC s involvement in variance requests has become actively supportive of variance requests. Since the BZA places considerable weight on both OP and ANC s recommendation s the ir high support rates might be one reason behind approval rate on zoning variance requests. Community Involvement Community involvement 7 refer s to any party /person who attends public hearing s or submits a letter or a petition t o express his/her point of view or represents an agent in a public hearing. They c an support, oppo se, or express their concerns about applications. During the period 1980 to 1984 the involvement rate was 60% or higher but then decreased ( Figure 5 9 ) Since 1990 involvement rate s have remained around 20%. In the beginning of 2000, the term party status 8 was quoted in the documents. Before that there was no clear distinction for the level of community involvement in terms of the number of persons involved The new definition differentiates party from 7 The ANCs are excluded from community involvement here. The author separated ANC involvement from community involvement in this paper based on their different legal status and the stakeholders they represent. The ANC is a gover nment body and its members are elected from local neighborhood s It represents the whole neighborhood and makes decisions based on the benefits of their constituency. Community involvement refers to the activities of individuals or associations. Most of th em are directly affected by persons whose properties are adjacent to the propert ies which have applied for a variance. 8 actions by other parties. T hose without party status can still testify at the public hearing(s) and submit comments. The granting of party status by the Zoning Commission is by no means certain; at a ZC PUD hearing several years ago, the Zoning Commission denied party status to the Capitol Hill Restoration Society, even though it exhibited the most expertise on the matter. http://www.ancnorm.org/wp content/uploads/2011/01/Party Status.pdf

PAGE 95

95 perso F orm 140 P arty Status Request an explanation is provided about person and party. Any person or representative of an organization may provide written and/or oral testimony at a public hearing. A person who desires to participate as a party i n a proceeding, however, must make a request and must comply with the provisions on this form. A party has the right to cross examine witnesses, submit proposed findings of fact and conclusions of law, receive a copy of the written decision of the Zoning C ommission or Board of Zoning Adjustment, and exercise any other rights of parties as specified in the Zoning Regulations. Approval of party status is contingent upon the requester clearly demonstrating that his or her interest will be more significantly, d istinctively, or uniquely affected by the proposed zoning ac tion than that of other persons ( District of Columbia o ffice of z oning 2011, para. 1) As me ntioned in the literature review in Chapter 2, opponents or supporters should not affect the While Washington D.C. zoning ordinance s do not include a imony for the BZA decision making, the setting of party status indirectly reflects the f In contrast to ANC s involvement, community involv ement represents a interest over interest. Usually the involved parties are owners and therefore commonly stakeholders with property They are naturally concerned with the possib le negative changes that might affect their own properties. Neighborhood association s are also major participant s in community involvement. When many neighbors oppose a variance application it might urge the

PAGE 96

96 existing neighborhood association might be urge d to take measure s in opposing the application. In cases where the community is involved with variance application s oppositions from individuals and asso ciations, especially since 1990 constitute the majority ( Figure 5 10 ) In the 1980s, 56.1% of applications were opposed by the community In 1990s, the proportion of community involvement in all cases decreased, but the ratio of opposition increased to 77.7%. In the 2000s, this ratio averaged 72.4%. This phenomen on reflects that opponents were very motivated to attend public hearing s or to write letters to express opposition to a variance request On the other hand no involvement from a communit y reflects its acquiescence to the proposals, though this cannot be m isunderstood as tacit support When a co mmunity does not get involved with a variance it is assumed/ surmise d that the variance would not pose a significantly negative impact to the existing community. Figure 5 11 shows the comparison among denial rate, the Office of Planning Opposition rate, the ANCs opposition rate and community opposition rate by year since 2000. As mentioned above, due to the problem of unstandardized do cumentation, the data about the OP involve ment might not accurately describe the situation. Since t he comparison would be biased u sing the data before 2000 this chart shows the comparison from 2000 to 2009. The chart clearly indicates that the denial rate is consistently lower than that communit y opposition rate. Community opposition rate kept more than 10% each year while the denial rate was under 10% except in 2002. In some years, the community opposition ra te was twice or even triple the denial rate. Since 2003, the OP also shows a higher op position rate than the In

PAGE 97

97 fact, during s ome year s the OP opposition rate was higher than the community opposition rate. The ANC s opposition rate was generally sma ll relative to the other groups and in s ome year s was higher than the denia l rate, in some year s lower. It is difficult to visualize from this chart whether the affected by the OP, the ANC, or the community. In Chapter 5 a binary response model was applied to answer this question. Table 5 1 provides the approval and denial rate s under different recommendations from the Office of Planning, the ANCs, and c ommunit ies For the variance applications opposed by the OP, 62% were denied by the BZA and 38% wer e approved by the BZA. When the OP supported the applicatio ns, the BZA granted 93% of application requests and denied 7%. It implies that the OP support for variance requests considerably increased the probability of having a request granted. For cases in which the OP recommended denial, the BZA also denied a large proportion of the applications but this relationship was not as strong as when a request received the OP support W hen the ANCs opposed applications, 43% were denied by the BZA with 57% receivi ng approval. When the ANCs supported the applications, 93% were approved and 7% were denied by the BZA, similar to the proportion supported by the OP. From these number s it is evident that the attitudes of the ANCs towards variance applications appeared to influence both the approval rate s and denial rates The numbers also reflect that the ANC s opposition has less influence on the BZA than that of the OP. In Chapter 5 the binary response model will show whether and to what extent the OP and the mmendation affect the

PAGE 98

98 Regarding community activities, 67% of the a pplications were granted in cases when there were opponents, and 33% applications were denied. When there were supporters for the applications, 80% were granted and 20% wer e denied by the BZA. W hen the OP, the ANCs and neighbors all opposed the applications, 77% were denied by the BZA. When the OP, the ANCs and neighbors in the community all supported the applications, 98% were granted by the BZA. The consensus among the OP, the ANCs, and the community is more likely to influence the BZA to make the decision consistent with their recommendations. Use of Properties Requested for Zoning Relief According to the use of properties requested for zoning relief, the cases were group ed into five categories: residential commercial, mixed use, industrial and other. Other refers to non profit use, such as churches, non profit office and community centers, etc. Figure 5 12 shows that residential use constituted 60% of the variance requests during the 30 year period Commercial use made up 24% and ranked as the second largest use. Other accounted for 14% and mixed use was 2%. There were no industrial variance application s Alt hough eleven cases in total were in industrial zones, th e se uses were related to commercial, residential and mixed uses. In terms of approval rate by property use, the approval cases were not evenly distributed. The highest approval rate was 17.7% higher than the lowest approva l rate ( Figure 5 13 ) V ariance applications with mixed use were m ore likely to be granted, with an approval rate as high as 97.8%. Other use also had a high approval rate 91.3%. The approval probabilit y for resid ential use was less than mixed use and other use. At 80.1% commercial use had the lowest approval rate. The supportive stance of the BZA to mixed us e and other use is apparent. In the era of global warming and shortage of

PAGE 99

99 energy in highly concentrated popu lation centers smart growth policy is widely accepted in Washington D.C. Planners typically promote stategies to make the community safer and healthier, as represented by community walkability access to public transportation, and general convenience of l iving. Mixed use development is one of those planning tools. Since this type of development is consistent with the theme of smart growth, the BZA is more likely to grant the application Other refers to some non profit uses which are set up for the public welfare. The effectiveness of these non profit uses is demonstrated by the high percent of variance requests that are granted. Though difference s existed in terms of approval rate by requested property use, the approval rates were still relatively high re gardless of the usage category of the requested property. Figure 5 14 shows the number of variance applications by property use each year. Residential variance requests occupied the highest pro portion each year. T h e gap between residential and other uses has declined in the later years. The number of residential v ariance applications by year shows the largest variation ranging from 16 to 101. The number of commercial variance applications decreased gradu ally from 19 80 to 1989, and has remained under 20 since 1990, which ranged from 4 to 18. The number of other uses variance applications was more stable than other types of property use since the 1980s, with no more than 20 cases each year. Variance application s for mi xed use properties occupied a very small part each year. There were 14 years in which no variance application was propo sed for mixed use property. Five mixed use applications were proposed during the 1980s, and 8 during the 1990s. Since 2000,

PAGE 100

100 applications for mixed use properties increased slightly usually around 3 each year. In 2002, a maximum number of 8 was reached Approval rates might also show a discrepancy with regards to the type of applicant. This assumption comes from the concern that firms and non profit organizations are more adept than individuals in preparing variance applications Well prepared variance applications are obviously more likely to be granted Correspondingly the variance requests of individuals are more likely to be rejected t hat those of firms and non profit organizations. Table 5 2 shows how approval rates are different due to type of use and type of applicant. In general, ind ividual s received the lowest approval rate 79.8 Firm s ran k in the middle, with an approval rate of 88.9%. Other applicants received the highest rate 93.9%. In terms of different types of property use, individual applicants met the strictest decisions when they applied for variances in which the type of property use was commercial or other. The approval rates for both are below 70%. Alt hough in terms of residential variance applications, individual s received request approval as high as 82.9%, the approval rate for firms was 8.2% higher than that of individual s w hich was 91.9%. Other performed the best regardless of property use application Wards in Washington D.C. As mentioned in Chapter 4 since 1968 wards 9 were applied to geographic divisions in Washington D.C. They were drawn based on natural neighborhood boundaries and redrawn according to the US Census population distribution. Each ward has an equalized population which provides us with a standard to compare how zon ing variance applications var y in the District in terms of number and geographic distribu tion. 9 See also note 31.

PAGE 101

101 Figure 5 17 shows the distribution of variance applications by Ward. As this map shows, variance applications were highly concentrated in the middle of the city. Ward 2 had the largest number of variance applications, 561 or 26.2% of the total applications. Ward 6 had 459 applications, most of which were located in the north middle of the ward. Ward 3 ranked the third in terms of number of variance applications. However, the distribution of requested properties in this area w as more diverse. Ward 4, Ward 7 and Ward 8 had the lowest nu mber of applications, 152, 133 and 98 respectively The number of applications in Ward 2 is five times more than that of Ward 8. Figure 5 16 shows the distribution of appr oved and denied variance applications from 1980 to 2009. 1816 applic ations were approved and 324 were denied. Ward 1 Ward 1 is the smallest ward within the eight wards, cover ing 1778 acres. As mentioned before, each ward has a similar population. It implie s the density of population in Ward 1 is the highest among the eight wards, which has an average number of 41 persons per acre. Ward 1 is predominantly residential, with 82% of its l and area being used for housing ( District of Columbia Office of Planning 2007 p. 33 ) The median household income was 36 902 dollar s as of 2000, and the total number of housing units was 34 632. Mount Morgan and Columbia Heights are the two major neighborhoods in Ward 1. Lands for commercial use spread out along the major roads. T o the south of Ward 1, there are a few lands for industrial use. Figure 5 17 shows the distribution of variance applications and the From 1980 to 2009, 236 (or 11.0%) variance requests happened in Ward 1, in which 147 were residential, 52 were commercial, 8 were mixed use and 29 were other. In Ward 1, 32 (or 13.5%) applications out of 236 were denied by the BZA. As the graph

PAGE 102

102 shows, variance applications spread out all over the Ward 1. Especially in the south ern part board ing Ward 2, a relatively higher density is to be found; t his area is part of the old city neighborhood. Ward 2 Ward 2 is located in th e center of the District and covers a variety of land use types. It is the busiest area in which t he political, educational, historical, cultural and business activities concentrate. There were 39 352 housing units as of 2000, and the median household income was 44 742 dollar ( District of Columbia O ffice of P lanning 2007 ) Ward 2 had 564 (or 27.6%) variance applications the largest amount in the area 72 (or 12.8%) cases out of 564 cases were denied by the BZA, which is 3.2% lower than the average denial rate of 15% for the District ( Figure 5 5 ). Within the eight wards, Ward 2 was one of the most highly co ncentrated wards in terms of variance applications. The White House, the Capitol, Washington Monument and a great number of museums are located in the middle of Ward 2, in which only a few variance requests were located. However, in the Central Business District (downtown) area located in the north of the White House, variances requests were highly concentrated. The d owntown area is the largest commercial area in the District and in terms of p roperty use a total number of 207 variance requests were for commercial use, which occupied 39.2% of the total number. It shows a majority of commercial variance requests happened in this area. 174 (or 83.6%) variance requests related to commercial use were approved, which was a little higher than the average approv al rate 80.1% for commercial use ( Figure 5 13 ). Residential areas are located around the north, east and west of the downtown area. Desirable locations adjacent to the downtown area contribute to high property

PAGE 103

103 valu es. The recent trend toward urban living occurring in cities across the United States has made this area increasingly desirable and has c ontributed to higher home costs ( District of Columbia Office of Planning, 2007 p. 37 ) The Georgetown neighborhood is th e traditional residential area, in which can be found Georgetown University, Dumbarton Oaks Park, Montrose Park, and Rock Creek Park are nearby. The average home value in Georgetown is much higher than other areas. Besides, Kalorama and Foggy Bottom are al so traditional neighborhoods with high home values. New develop ments and redevelopment s expand to east part of the old city neighborhood. From 1980 to 2009, 274 variance applications were related to residential use. It is apparent that the amount of varian ces applications and the distribution of variances are related to the activities of development. Mixed use and other use had 22 and 61 variance requests separately. With the highest number of variance applications, Ward 2 c an serve as a good study area f or a deep er look into the eight wards. Figure 5 19 shows the distribution of variance applications and the decisions by d ecennial periods. From 1980 to 1989 there were 293 (or 52.0%) out of 564 applications more than half of the total applications during the 30 years. The BZA granted 249 requests, of which 133 were full approv ed and 116 approved on condition 44 (or 15.0%) out of 293 requests were denied ; 126 (or 43.0%) out of these were for commercial use, and 128 (or 43.7%) were for residential use. Variance requests during 1990 to 1999 were much less than th ose from 1980 1989. 119 (or 21.0%) out of 564 requests were proposed during this period. 18 (or 15.1%) cases were denied by the BZA. More than half of these request s were for residential use, and requests for

PAGE 104

104 commercial use were 12 less than residential use. From 2000 to 2009 the amount of variance requests was 33 more than that of 1990 1999. V ariances requests were highly concentrated in the east of War d 2, which was in accordance with the increased development activities in the east part. 10 (or 6.6%) out of 152 cases were denied by the BZA. The denial rate here was much lower than the average denial rate. In addition, conditional approval rate was also much lower than before. Applications during 2000 to 2009 were more easily to be granted with full approval than before. Ward 3 Ward 3 covers the west part of the District ( Figure 5 20 ) The lands of this ward are mainly used for residential use. Low density residential properties are dominant in this area. More than 40 % of the areas in Ward 3 are zon ed as R 1 Districts (One family detached dwellings). According to Census 2000, the total housing units numbered 38 73 4. It features the highest median household income, 71 875 dollar s in 2000. There were fewer new residential developments in Ward 3 compared to other wards. New developments are mainly for mixed use and concentrate around Metro stations Friendship Heights and Tenleytown Metro. Commercial properties are mainly clustered around Wisconsin Avenue NW and Connecticut Avenue NW. From 1980 to 2009, there were 324 variance requests in Ward 3, in which 260 (or 80.2%) cases were of residential, 32 cases were commerci al, 31 cases were other, and 1 case mixed use. 33 cases out of 324 requests were use variances, in which 11 were denied. 17 out of 33 cases were related to commercial use in residential zones, 7 of which were denied by the BZA, 9 were granted with conditio n, and only 1 was granted with full approval. 6 use variance requests were related to non profit use in residential zones, in which 1 case was denied. Seven use variance requests were related to

PAGE 105

105 residential use in residential zones, which included the conv ers ion or construction of apartment house s, or flat s, in low density zones, the creation of an addition. Two requests proposing conversions into high density buildings were denied. Four requests were granted partially or with condition. One request was ful ly granted. The high denial rate of commercial use variance in residential zones shows that the BZA was very prudent in granting commercial use variance in these zones. As Figure 5 6 shows, the BZA seems less likely to grant use variance compared to area variance. From the above number, we can learn that the fewest grants were for those which proposed commercial use variance in residential zones. Ward 4 Ward 4 is located in t he north of the District and covers 5 763 acres ( Figure 5 21 ) The whole area is dominant by residenti al land use, which is 87 % of the total area. According to US Census 2000, Ward 4 had 31 044 housing units and 65 % of the total housing units were single family housing. R 1 District (one family detached dwellings) and R 2 District (one family semi detached dwellings) are the main zones in Ward 4, which cover around 50% of the total area. The median household income was 46 408 dollar s which ranked as the second highest h ousehold income in these 8 wards. Commercial use lands mainly spread along George Ave from the south to the north and Kennedy Street NW from the east to the west. Most properties for variances applications were situated around the major roads of Ward 3. T he total number of residential, commercial, other, and mixed use applications was 100, 19, 32 and 4 respectively 34 (or 22.37%) out of 152 variance applications were denied by the BZA, which is much higher than the average denial rate of the District. Thi s is e specially true for commercial variance applications 8 (or 42.1%) out of

PAGE 106

106 19 applications were denied by the BZA. Besides, 13 (or 61.9%) out 21 use varia nce applications were denied. This statistic suggests that in the low residential density dominate d areas, use variance applications are not easily granted. Ward 5 Ward 5 is situated in the northeast of the District ( Figure 5 22 ) It is the most diverse region in terms of land use, which includes residential, c ommercial and industrial areas Lots for commercial and industrial use s are concentrated and relatively large. Low density residen tial zones are mainly located to the northeast of the ward. In the south and south east are mainly high density residential z ones. In 200 0, the total housing units numbered 32 258 and the median household income was 34 433 dollar s 175 variance applications were for properties situated in Ward 5, in which 103 were for residential use, 39 for commercial use and 33 for other use. 44 (or 25.1%) out of 175 applications were denied. In the southeast and the middle of the ward variances applications tended to be concentrated. As show n in the southeast of the ward, 4 applications with adjacent properties were denied by the BZA. They we re all for use variance and the properties are located in residential zones. All 4 variances applied for commercial use in residential areas. Economic benefits drive property owners to change the use of property, especially when existing commercial propert ies are nearby. However, as pointed out in Chapter 4 mere financial hardship or an increased return from the property is not a sufficient reason for granting a variance ( "Zoning Variances and," 1955 p. 520 ) Unless a large scale rezoning plan is proposed for local economic development by a great number of residents, city planners, and local policy makers, there is no sufficient reason t o transform the use of property from commercial to residential just for financial considerations.

PAGE 107

107 Ward 6 Ward 6 is in the heart of the District ( Figure 5 23 ) The natural boundary of the Anacostia River separate s Ward 5 from Ward 8. The northwest of the ward is featured as available for commercial use and mixed use ; the area is connected to the downtown area in Ward 2. In the s outh around the Navy Yard, the land use is also for commercial and mixed use s The northe ast part is covered with residential lands, which are mainly R 4 Districts (row dwellings, conversions, and apartments). Much of R 4 District would lie within urban renewal areas as assigned by the Redevelopment Land Agency, the demolition of substandard structures and replacement with low density apartm ent houses should be encouraged ( "Zoning Regulations," 1958 p. 12 ) The total housing units were 35 510 and median household income was 41554 according to US Census 2000. Affected by the new urbanism, urban renewal is described as smart growth, mixed use and transit oriented development. Blighted communities are gradually transformed through mixed use o f retail, office, and residential. Redevelopment also brought a large number of variance applications. From 1980 to 2009, there were 459 variance requests in Ward 6, in which 233 were proposed during 1980s, 112 were proposed during 1990s, and 114 were pro posed during 2000s. 252 (or 54.9%) out of 459 requests were for residential use, 142 (or 30.9%) were intended for commercial use, 55 (or 12.0%) were for other use, and 10 (or 2.2%) were for mixed use. All these 10 requests for mixed use were approved by th e BZA. 62 (or 13.5%) out of 459 requests were denied. Th is denial rate was less than the average denial rate of the District. It might reflect that the BZA was less strict when large scale redevelopment occurred for the revitalization of local communities

PAGE 108

108 Ward 7 Ward 7 is situated in the east of the District ( Figure 5 24 ) 1/4 belongs to federal government, most of which is park land. Using Fort Dupont Park as a divider, R 1 Districts (one family detached dwellings) are mainly concentrated in the south of the ward. R 2 Districts (one family semi detached dwellings) and R 5 Districts (general residence) are mixed together in the north of the ward. According to US Census 2000, the total housing unit s numbered 33 651; and the median household income w as 30 533 dollar s which was the second lowest income within the 8 wards. Around 97% of the population in Ward 7 were black, 2% were white, and 1% were other. The commercial land i n the middle west of the ward constitutes the Minnesota/Benning Business District, which include s large shopping centers, retail stores and other services. According to the hand book prepared by Office of Planning, this business district was not fu lly utilized ; parts of the area consist of empty parking lots, open storage, vacant buildings, and undeveloped land ( District of Columbia Office of Planning 2007 p. 53 ) From 1980 to 2009, a total number of 133 variance applications were proposed in Ward 7, in which 92 (or 69.2%) were for r esidential use, 16 (or 12.0%) were for commercial use, and 25 (or 18.8%) were for other use. Unlike Ward 2, more than a half of the applications were proposed during 2000 to 2009. Comparable to the rank of median household income, the total amount of varia nce applications ranked the second lowest within the 8 wards. It might reflect that the number of variance applications is related to the local economic activities and situation. 20 (or 15.0%) out of 133 applications were denied by the BZA, in which 13 we re intended for residentia l use 4 for commercial, and 3 for other. 19 out of 133 applications were use variances, in which

PAGE 109

109 8 (or 42.1% ) were denied. Compared to 32% the average denial rate of use variance, the rate was much higher in Ward 7. Ward 8 Divid ed by Potomac River, Ward 8 is located in the south of the District ( Figure 5 25 ) It is the largest ward within the 8 wards, which covers 7 556 acres ; h owever, the taxable area is only 962 acres. Public and Instit utional land, park land and water bodies occupy the most part of this ward According to US Census 2000, the total housing units were 29654 and the median household income was 25 017 dollar s both of which ranked the lowest within the 8 wards. The resident ial land in Ward 8 is mainly occupied by multi family units, which was around 70% of the total housing units. The homeownership rate was around 20% in 2000, which was much less than the average homeownership rate : 40% in Washington D.C. Encouraged by hous ing projects and financial assistance programs, this area is under large scale development and redevelopment since 2000. From 2000 to 2005, around 8 000 housing units were constructed or rehabilitated ( District of Columbia Office of Planning 2007 ) Residential developments also brought parallel developments to commercial and public services. The total amount of variance applications from 1980 to 2009 w as 97, much less than other wards. 63 (or 64.9%) requests were proposed from 2000 to 2009, which was in accorda nce with the large scale development and redevelopment ac tivities since 2000 in Ward 8. Alt hough the total number of requests was not as many as other wards, developers might apply for Planned Unit Developments (PUD) for more flexibility, which is not stri ctly constrained by zoning regulations. Within 97 requests, 50 were for residential uses 21 for commercial and 27 for other. 12 (or 12.4%) out of 97 were denied by the BZA, which was a little bit lower than the 15% of average denial rate in

PAGE 110

110 the District. 18 (or 18.6%) out of 97 were use variance applications, in which 6 (or 33.3%) were denied. Testing Hypothesis In Chapter 3 f our general hypotheses were proposed They are: Hypothesis 1 Granting of variance applications is significantly higher than deni al. Hypothesis 2 Opponents from affected neighborhoods influence the zoning Hypothesis 3 Type of variance application affects the Hypothesis 4 The zoning administrative board does not pl ace significant weight on suggestions from other public agencies in its decision making. Zoning variance data in Washington D.C. from 1980 to 2009 are used in this paper to test the above four hypotheses. Hypothesis 1 was tested using the Binomial Test. H ypothesis 2 to 4 were tested in the basic Binary Response Model developed by the author S tatistica l and econometric analysis c an assist scholars and practitioners in the planning and law fields to better understand the situation of zoning variances. Testi ng Hypothesis 1 Figure 5 2 shows the approval rate of variance applications by year, which provides a general view about the indicated the approval rate was maintained above 60% from 1980 to 20 09. Alt hough it is apparent in this graph that the granting of variance application is higher than denial, statistical analysis is a better way to test hypothesis without subjective judgment. The decisions were divided into two groups: approved application s from 1980 to 2009 and denied applications from 1980 to 2009. To test whether two groups are statistically different, we can use the test analysis (Paired Sample T test) or the Binomial Test in a nonparametric test. The

PAGE 111

111 test analys is has relatively strict assumptions about the samples. It requires that the distribution of the samples is normal distribution, and all the samples are independent and random. Nonparametric test does not assume data should belong to any particular distrib ution. Compared to the test analysis, the nonparametric test is preferable in this study. The Binomial Test in a nonparametric test is used to verify whether the distribution of dichotomous data is the same as t he expected distribution. Assum in g that the proportion of variance applications granted by the BZA is not different from the proportion of variance applications denied, meaning the prop ortions for both are the same ( 50% each ) Set out th e null and alternative hypothesi s in this test: H0: the distribution of variance applications granted by the BZA is the same as the distribution of variance applications denied by the BZA. H1: the distribution of variance applications granted by the BZA differs from the distribution of variance application s denied by the BZA. We get the Binomial Test results ( Ta ble 5 3 ). In th e table Group 1 denotes the applications denied by the BZA and Group 2 denotes the applications granted by the BZA. The column N shows the number of cases i n each group. As show n 324 cases were denied by the BZA and 1816 cases were approve d The fourth column Observed Prop. shows the frequency of each group in the observed cases. The fifth column is the proportion of decisions which were denied by the BZA. The last column Asymp. Sig. (2 tailed) shows the p value. The p value is 0.000 which is less than 0.05. Th us the statistic al test performed resulted in the rejection of the null hypothesis in favor of the alternative. I t means that at a confidence leve l of 95% the proportion of variance

PAGE 112

112 applications denied by the BZA is different from the proportion of variance applications approved. In another word, the denial rate is significantly different from the approval rate. Considering that the average approval rate is much higher than the average denial rate, we come to the conclusion that the granting of variance application is significantly higher than denial. Testing Hypothesis 2 to 4 Chapter 3 explained the method of Binary Response model and its applicati on. The author also constructed the framework of a basic b inary r esponse model according to the research questions. In this section, detailed model construction, data compil ation and process ing results, and result analysis were demonstrated. First, set out th e null and alternative hypothesi s in this test: (1) Hypothesis 2 H0: o pponents f rom affected neighborhoods do not influence the zoning administ H1: o pponents f rom affected neighborhoods influence the zoning administ rati ve (2) Hypothesis 3 H0: t ype of variance application does not affect the H1 : t ype of variance application affects the (3) Hypothesis 4 H0: t he zoning administrative board does not place significant weight on suggestions from other public ag encies in its decision making. H1: t he zoning administrative board place s significant weight on suggestions from other public ag encies in its decision making.

PAGE 113

113 Format of model There are two kinds of b inary response model: probit model and logit model A binary response model is referred to as : a probit model if F is the cumulative normal distribution function. It is called a logit model if F is the cumulative logistic distribution function. The logisti c and normal distributions are both symmetrical around zero and have very similar shapes, except that the logistic distribution has fatter tails. As a result, the conditional probability functions are very similar for both models, except in the extreme tai ls. ( Horowitz & Savin, 2001 p. 44 ) As Wooldridge (2002) pointed out, the two models tell a consistent story. The signs of the coefficients are the same across models, and the same va riables are statisti cally significant in each model ( Wooldridge, 2002 p 468 ) So with regards to the functional form of model, we can use both probit and logit models. However, if all of the independent variables are categorical or mixed with continuous and categorical variables, the logit model is then usually employed ( Wuensch, 2011 ) Since the binary response model for the have categorical variables, the logit model is best applied in this study The equation of binary response logit mode l is as follows:

PAGE 114

114 where is the predicted probability of the event which is given the value of 1, are independent variables, are the parameters of independent variables, and e is mathematical constant. The above equ ation can be transformed to the following equation: which shows the ratio of the probability of event to the probability of no event. From the online documents provided by the D.C. Office of Zoning, the author collected the needed information on recommen dation s from the OP, the ANC s other public agencies 10 opponents present at the public hearing s and other forms of being present at the public hearings, such as letters, petitions and signatures. The variables of recomme ndation s from the OP, the ANC s ing several outputs: support, opposition, no action, no objection and no data. The author simplified the above outputs and classified them into thr ee groups: support group, opposition group, and no action/ no objection/ no data group. The first two groups have clear views on the variance applications and the last group has no clear opinion. Statistical analysis requires setting K 1 dummy variable to represent categorical groups. For each dummy variable a score of 0 will indicate that the subject does not belong to the group represented by that dummy variable and a score of 1 will indicate that the subject does belong to the group represented by that dummy 10 Since it is required by law t hat BZA should give great weight to recommendation from OP, the author separated OP from other public agencies in this model.

PAGE 115

115 variable. One of the groups will not be represented by a dummy variable. If it is reasonable to consider one of your groups as a reference group to which each other group should be compared, make that group the one which is not represented by a dummy variable ( Wuensch, 2011 p. 11 ) Consider ing th e high approval rate of variance application, the opposite forces which influence the BZA to deny applications perform very important rol es. Whether these forces affected the and to what degree is research questions. These forces might include opponent s from the neighborhood, and opposition from public agencies. Based on this consideration, the oppositio n group was set as reference group in this model. Other public agencies include but are not limit to the Department of Transportation, the Department of Public Works, the Historic Preservation Review Board, the Department of Human Services, the National Capital Planning Commission, the Capitol Hill Restoration Society, etc. These agencies t a k e part in the decision proces s based on the type of lands buildings, and locations. These agencies are recommended by the Office of Zoning for review. Since not eve ry agency got involved into each case, the author combined their opinions together. There were situations where one agency supported the application while another agency recommended denial. So in the model design, two variables are related to other public agencies. One is a support from other public agencies: 1 means support and 0 means other situations; the other is opposition from other public agencies: 1 means opposition and 0 means other situations. The independent variables in this binary response mo dels are as follows:

PAGE 116

116 Use The type of variance. 1= use variance and 0= area variance OPSupport Support from the Office of Planning. 1= Support and 0= other situations. OPNoaction There was no action or no objection from the Office of Planning, or no data recorded in the document. 1= No action/ No objection/ No data, and 0=other situations. ANCSupport Support from the Advisory Neighborhood Commissions. 1= Support and 0= other situations. ANCnoaction There was no action or no objection fr om the Advisory Neighborhood Commissions, or no data recorded in the document. 1= No action/ No objection/ No data, and 0=other situations. HearingOppose Opponents from neighborhood present at the public hearing. 1= Opponents showed up and 0=No opponen ts showed up. CRSupport Support from communities. 1= Support and 0= other situations. CRNoaction There was no action from communities. 1= No action and 0= other situations. OASupport Support from other public agencies. 1= Support and 0= other s ituations. OAOppose Opposition from other public agencies. 1= Opposition and 0= other situations. The dependent variable is as follows: BZADecision The applications. 1= Denial and 0= Grant.

PAGE 117

11 7 Model re sults and testing The author ran the crude model using Wald method which eliminated the variable CRSupport. Alt Hosmer Lemeshow test shows that the model does not fit data. Multicollinearity was found in the model. It was found that OASupport is correlated with OPSupport, and CRNoaction is negatively correlated with HearingOppose. The abov e correlations are reasonable. Alt hough report s from different public agencies focus on different aspects all of them con sider the benefits of the whol e community. In all likelihood, t hey would proba bly come to the same opinion. Accordingly support from the OP is correlated with support from other public agencies. The negative correlation between no action f rom community and opposition in public hearing might imply that if no opponents appear at a public hearing, usually no action emerges from the community either. The author eliminated OASupport and CRNoaction from the model and ran the adjusted model again Appendix A demonstrates the detailed process of model assessment and adjustment. The adjusted model includes seven independent variable s. Ta ble 5 4 shows the significance, parameters of each variable, Wald test an d the odds ratio (Exp(B)) of the model. The sixth column Sig. provides the sig nificance of each variable. At a 95% confidence level, all the variables included in the adjusted model satisfied the Wald test and Hosmer Lemeshow test. The equation of this m odel is as follows:

PAGE 118

118 In which Since the function is monotonically increas ing, we know that the higher x is, the higher f is. In our model, the higher z is, the higher probability that a variance application is denied by the BZA. The parameters for the three variables Use, HearingOppose, and OAOppose have the positive value ( Ta ble 5 4 the second column B), which mean s that when they have the value of 1, the probability that a variance applica ti on is denied is higher than if they are given the value of 0. The degree of influence by opponents p resent at a public hearing is slightly higher than that of use variance and opposition from other public agencies. Since the variables of opposition from the OP and opposition from the ANCs were set as reference groups, we cannot directly see their influences in Ta ble 5 4 However, from their counterparts we see how those counterparts decrease the probability of variance applicati ons denied by the BZA. The parameters of the categorical variables in this model are negative: the parameter of OPSupport is 3.202; the parameter of OPNoaction is 2.408, the parameter of ANCSupport is 1.715; and the parameter of ANCnoaction is 0.854. The absolute value of the parameter of OPSupport is higher than th e absolute value of parameter for ANCSupport, which impl ies that the influence of the OP to the than that of the ANC s The nega tive parameters of the variables that no action from the OP and no action from the ANCs also imply that when there is no clear stance from either body or the lack of recommendation to the BZA, the effect is not neutral. The probability of granting v ariance application increases in this situat ion The results of

PAGE 119

119 significance for OPNoaction, ANCnoaction and their negative parameter s should be paid attention. The re might be two explanations: since there were a large amount of ca ses in which there was no involvement from the OP, or the ANCs and th e granting ratio remained high during these 30 years, the model related the two and emphasized the significance. Indeed, there may be no relationship. Another explanation is that it might suggest to some extent no action/no objection is a kind of acquiesce nce, which sends a signal to the BZA to incline towards granting the application. It is hard to fathom which explanation is more reasonable according to the existing literature. More information is needed to further study this question. The results of the model also provide us with the following information: (1) Type of variance application The odds ratio for Use show s that when holding all other variables constant, use variance application is 1.978 times more li kely to be denied by the BZA than area vari ance application. As shown i n the section on Zoning Variance in Washington D.C. in Chapter 4 the test of practical difficulties is applied to area variance s and undue hardship is applied to use variance s The condition s to gra nt a use variance applica tion are already stricter than the condition s to grant an area variance from the legal perspective. This result may indicate that the BZA is more prudent in granting a use variance which might more negative ly impact to the community (2) Opponents appeare d at the public hearing The odds ratio for HearingOppose indicates the variance application is 2.916 times more likely to be denied when there is an opponent present at the public hearing

PAGE 120

120 than when no opponents is present In terms of probability, it pre dicts 74.5% of cases would be denied by the BZA when an opponent is present at the public hearing. (3) Recommendation from other public agencies Alt hough in the crude model OASupport is statistically significant, it was eliminated in the adjusted model du e to its correlation with OPSupport. On the other hand, the odds ratio of OAOppose reflects that the variance application is 1.927 times more likely to be denied by the BZA than when there is opposition fr om other public agencies than in other situations The model predicts that 65.8% of variance applications would be denied by the BZA when there are other public agenc ies which oppos e the applications. (4) Other forms of community involvement In the crude model, the variable CRSupport was excluded from the model according to Wald method, which means that support from residents in the community does not affect the decision significantly. CRNoaction was significant, which means other forms of opposition from the community exert a significant influenc e on the the community. However, it was negatively correlated with opponents present at the public hearing. This correlation implies that in the cases where opponents were present other forms of community oppositi on might also occur (5) Recommendation from the OP The odds ra tio of the categorical variable compare s each scenario except opposition to the opposition scenario. The inverted odds ratio compares the opposition scenario to an other scenario. The result sh ows that the variance application is 24.4

PAGE 121

121 (1/0.041) times more likely to be denied when the OP opposes the application than when the OP supports the application. This vast difference reveals that the influence of the recommendations from that OP is essenti al to the The inverted odds ratio shows the variance application is 11.1(1/0.090) times more likely to be denied when the OP opposed the application than when the OP exhibit s no action/no objection. The above results are reasonable a nd c an be explained simply When the OP has a clear stance about the application, the relatively consistent with the and support and the likelihood that the BZA denies the app lication st ands between the likelihood of the OP s oppos ition and he OP support (6) Recommendation from the ANCs The inverted odds ratio indicates that the variance application is 5.5 (1/0.180) times more likely to be denied by the BZA when the ANCs op pose the application than when the ANCs support it The inverted odds ratio indicates that the variance application is 2.3 (1/0.426) times more likely to be denied by the BZA when the ANCs oppose the application than when there is no action/ no objection f rom the ANC s According to the above results, we reject the null hypothesis in hypothesis 2 which states that opponents from affected neighborhoods do not influence the zoning A t least in Washington D.C., the null hyp othesis is rejected The opponents from the neighborhood affect the board s final decision. The forms of opposi tion include opponents appearing in public hearing s opponents writ ing letters to the board, and opponents petitioning etc. The participants cou ld be individual s

PAGE 122

122 several persons who join as one group, a local community association, or the representatives of a stakeholder. holding all other variables constant, use va riance application is 1.978 times more likely to be denied by the BZA than an ar ea variance application. If we consider this one variable in the model, it predicts that 8.5% of area variance will be denied by the BZA and 32.2% of use variance will be denie d by the BZA. Thus we reject the null hypothesis in hypothesis 3 which states that type of variance does not affect the decision. For hypothesis 4, we partially reject the null hypothesis that the zoning administrative board does not place significan t weight on suggestions from other public agencies in its decision making. As the result shows, the consistent with the OP and the recommendations for both granted cases and denied cases. Supports from the OP or the ANCs t ended to influence the BZA decision to grant the variance applications and oppositions from the OP or the ANCs tended to influence the BZA in den ying the variance applications. The variance application is 24.4 (1/0.041) times more likely to be denied when the OP opposes the application than when the OP supports the application. And the variance application is 2.3 (1/0.426) times more likely to be denied by the BZA when the ANC s oppose the application than when there is no action/ no objection from the ANC s In the second column of Ta ble 5 4 which shows the parameter of each variable, the parameter for OPSupport is 3.202 and the parameter for ANCSupport is 1.715. The above values mean that the influence of support from the OP was also higher than the influence of

PAGE 123

123 support from the ANC s Besides the OP and the ANCs, other public agencies also play important roles in the process and their recommendations significantly affect the decision when they oppose the app lication. Opposition from other public agencies increases the probability that the BZA denies a variance application. However, statistical analysis indicates that support from public agencies does not make any difference in the ultimate decision mak ing. Overall, the OP exhibited the strongest influence in affecting the BZA s decision. The second strongest factor is weight of the ANCs. This finding is in accordan ce with the zoning ordinance s in Washington D.C. As two major public agencies involved in the procedure of zoning variance application, the input of O P and the ANCs assist the BZA in making a fair decision. The influence of opponents present at the public hearing is weaker than the recommendations from the OP and the ANCs, yet it is stronger th an type of variance and recommendation from other public agencies (the OP and the ANCs are excluded from the other public agencies). Assessment of the m odel Table 5 5 shows the 2 log likelihood statistic and the coefficient of d etermination R Square. R square in OLS regression measures to what extent a model can explain the dependent variable. In logistic regression, there is no true R 2 value as there is in OLS regression ( Newsom, 2010 p. 1 ) The Cox & Snell R square and Nagelkerke R square in the logit binary response model are two methods to represent the function of R square in OLS regression. The larger the value is, the more the model explains t he dependent variable. However, the maximum value of The Cox & Snell R square is less than1. Since Nagelkerke R square ran ges from 0 to 1, it is prefer ably used in model explanation. In the adjusted model, the Cox & Snell R square is 0.232 and Nagelkerke

PAGE 124

124 R square is 0.406. It indicates that the independent variables in our model can explain 40.6% of the The Hosmer Lemeshow test is used to test the overall m odel fit. The null hypothesis shows that the model fits the data well. As Table 5 6 shows, the value of Sig is 0.122, which is larger than 0.05. It indicates that the data fit the model well. Classification Table ( Table 5 7 ) shows the cross table of the observed amount of dichot omous dependents and the predicted amount of dichotomous dependents. As Table 5 7 shows, the 1 501 cases granted and the 252 cases denied by the BZA were predicted correctly by the model; the 72 cases denied by the BZA were predic ted to be granted by the BZA; and 315 cases granted by the BZA were predicted being denied by the BZA. In terms of granting, the correct rate is 82.7% and in terms of denial, the correct rate is 77.8%. The overall correct rate in this model is 81.9% 11 The ROC curve is one method to evaluate the performance of the classification table. A ROC curve is a graphical representation of the tradeoff between the false negative and false positive r ates for every possible cut off ( Abrahams & Zhang, 2008 p. 316 ) The value of th e area under the curve is between 0.5 and 1. The closer the area is to 1, the better the classification table performs. Figure 5 26 shows the ROC curve for this binary response model. As Table 5 8 shows, the area under the ROC curve is 0.869. This value is close to the value 1 and much larger than the value of 0.5. This value could be considered as an ideal value in this test. The binary response model is quite sufficient model to predict t he 11 The author decided to use 0.150 as a cut value in this model in order to balance the correct rate between grant and denial. It m eans that we assume that when the probability of denial in this model is above 0.150, the BZA would deny the variance application. And when the probability of denial is under 0.150, the BZA grants the variance application.

PAGE 125

125 the OP, the ANC, Other public agencies, type of variance, and whether there is an opponent present at a public hearing. Further Investigation Other F actors The basic binary response model above is applied to testify Hypothesis 2, 3 and 4. As show n in the above section, the results are convincing and reasonable. The decisions are affected by recommendations from the OP, the ANC s an d other public agencies. Moreover opponents present at a public hearing d id influence the decisions compared to no opponents attending a public hearing. Alt hough it should not be one of the conditions required by law to grant/deny variance application, opponents do impact the ted out. Type of variance also affects the Use variance is almost tw ice as lik ely to be denied than area variance. Beside s the above factors which were tested in the basic binary response model, the author conducted further investigation o n other possible factors that might also affect the 12 existing type of property (residential, commercial, mixed use, other), and type of variance applicant (individual, firm, other). These variables were added into the basic binary response model and tested by statistic methods. Besides statistic reasonability, the author also verified reasonability in reality. Since the data for land value did not cover all 2 140 cases, in this step 1 301 cases whic h have the data of land value were used for 12 Property area and land value in this model were transformed using the logarithm function for statistic reason.

PAGE 126

126 statistical analysis. Due to the limitation of data availability, the results in this section might not fully represent the facts. So the author focused more on general characters rather than detailed numbers. T he analysis here may provide guidance for further research. The statistic results showed that with the except ion of land value, other variables do not affect the garithm function for land value is negat ive, which means the land value decreases the probability that the BZA denies the application. The la r ger the land value, the lower the probability that the BZA den ies the application This finding might imply that economic power affects the s potentially. Clustering of Zoning Variance Spatial analysis was applied to test whether variances are clustered as a result of the literature review, researchers found evidence of clustering of variances which might lead to further granting of variances or rezoning ( "Zoning Variance and," 2005 ) From Figure 5 17 variance applications are shown to be concentrated in Ward 1, Ward 2, and Ward 6 Whether these variance applications are spatially clustered which has statistical significance is the question in this section. Hot sp ot analysis in ArcGIS was applied to test the distribution of variance applications. Hot spots refers to ( "A Spatial Statistics," 2010 Chapter 6 p.1 ) Hot spot analysis is widely applied in the examination of crime locations to identify the areas where crime occurs with high frequency. It is a lso widely applied in some emergency preparedness programs ( "A Spatial Statistics," 2010 ) ArcGIS provides hot spot analysis which can

PAGE 127

127 identify and map clusters of incidents. It calculates the Getis Ord Gi statistic which measures spatial clustering. The author ap plied this tool to identify whether cluster of variance applications and decisions existed over th e 30 year period and where these variances were concentrated. Appendix C shows the detailed process of hot spot analysis in this research. Figure 5 27 shows the results of hot spots analysis. The red color in thi s map covers the area where variance applications were spatially clustered with statistical significance. The area with red color is located in the center of th e District, which includes the south ern part of Ward 1, the northeast part of Ward 2, and the north ern part of Ward 6. It is the bu siest and most active area in the city. It includes central business districts, has a high population density, and is undergo ing development activities. The dark blue color covered areas are also variance applications clustered areas. The d ifference between the red and dark blue is that the in the red areas the concentration of applications was more intense and some variance ap plications occurred in the same location. The light blue color means that in its areas variance applications were not clustering in a statistically significant manner It is apparent that in Ward 3 the variance ap plications were not clustered. Alt hough in Ward 4, Ward 5, Ward 7 and Ward 8 the total number of variance applications was less than that in Ward 3 ; the map shows that in th e se four wards the applications were concentrated in some certain areas. This map implies that the activities of variance app lications in Washington D.C. were not evenly distributed. The coincidence that some location s had variance local economic activities and development activities

PAGE 128

128 may reflect some inner causal relationship between economic activit ies and variance applications. This is a question which needs more material for further research. As mentioned above, the red color also means that the locations of variance applications are closely connected to each other, includ ing the situation that several applications occurred in the same location. The data shows that there were 137 locations which had more than 2 variance applications during 1980 to 2009, in which 8 locations had 4 variance applications and 9 locations had 3 variance applicati ons. Figure 5 28 shows the locations which have more than two variance applications. They are mainly concentrated in the middle of the District. Within these 137 locations, 56 (or 40.9%) locations are for commercia l use. Compared to the 25% average rate of properties for commercial use in zoning variances, this number is much larger. This fact reflects that residential properties are relatively more stable than commercial properties in terms of conforming to zoning ordinance s under the condition that the location had previous zoning variance application. Figure 5 29 shows the approved variance applications hot spots It is very similar to the variance application hot spots ( Figure 5 27 ). The comparable hot spots distribution implies that there was no apparent bias from the the location of property. The spatial concentration of granted variance stems from th e spatial concentration of variance applications.

PAGE 129

129 Table 5 1 Cross tabulation of a pproval and d enial r ate under different r ecommendations from the Office of Planning, the ANCs, and the c ommunity OP Oppositi on OP Support ANC s Opposition ANC s Support Community Opposition Community Support Approval 38% 93% 57% 93% 67% 80% Denial 62% 7% 43% 7% 33% 20% Table 5 2 Cross tabulation of a pproval r ate by p roperty u se and t ype of a pplicant Total Approval Rate Residential Approval Rate Commercial Approval Rate Mixed Use Approval Rate Other Approval Rate Firm 88.9% 91.1% 88.0% 97.3% 76.9% Individual 79.8% 82.9% 68.3% 100.0% 65.2% Other 93.9% 93.1% 100.0% NA 93.7% Ta ble 5 3 Binomial t est Category N Observed Prop. Test Prop. Asymp. Sig. (2 tailed) decision Group 1 1 324 .15 .50 .000(a) Group 2 0 1816 .85 Total 2140 1.00 a Based on Z Approximation. Ta ble 5 4 Significance and p arameters of v ariable in b inary r esponse m odel B S.E. Wald df Sig. Exp(B) Step 1(a) Use .682 .182 14.098 1 .000 1.978 OPSupport 3.202 .221 209.108 1 .000 .041 OPNoaction 2 .408 .178 182.321 1 .000 .090 ANCSupport 1.715 .227 56.842 1 .000 .180 ANCnoaction .854 .211 16.351 1 .000 .426 HearingOppose 1.070 .180 35.396 1 .000 2.916 OAOppose .656 .319 4.222 1 .040 1.927 Constant .880 .241 13.378 1 .000 2.411 a V ariable(s) entered on step 1: Use, OPSupport, OPNoaction, ANCSupport, ANCnoaction, HearingOppose, OAOppose.

PAGE 130

130 Table 5 5 2 L og l ikelihood s tatistic and the c oefficient of d etermination R Step 2 Log likelihoo d Cox & Snell R Square Nagelkerke R Square 1253.647(a) .232 .406 a Estimation terminated at iteration number 6 because parameter estimates changed by less than .001. Table 5 6 Hosmer and l emeshow t es t Step Chi square df Sig. 1 11.407 7 .122 Table 5 7 Classification t able of b inary r esponse m odel Observed Predicted BZA's Decision Percentage Correct Grant Denial Step 1 BZA's Decision Grant 1501 315 82.7 Denial 72 252 77.8 Overall Percentage 81.9 Note: t he cut value is .150 Table 5 8 Area u nder the c urve Test Result Variable(s): Predicted probability Area Std. E rror(a) A symptotic Sig.(b) Asymptotic 95% Confidence Interval Lower Bound Upper Bound .869 .011 .000 .847 .892 The test result variable(s): Predicted probability has at least one tie between the positive actual state group and the negative actual state group. Statistics may be biased. a Under the nonparametric assumption b Null hypothesis: true area = 0.5

PAGE 131

131 Figure 5 1 Number of v ariance a pplications by t ype and y ear 13 Figure 5 2 Approval r ate by t ype of v ariance and y ear 13 Note: Data for 1985 for four months from May 15 to Sep. 25 are not available form Office of Zoning website.

PAGE 132

132 Figure 5 3 Area v ariance a pproval r ate by a pproval t ype in a pproved a pplications by e very f ive y ear s from 1980 to 2009 F igure 5 4 Use v ariance a pproval r ate by a pproval t ype in a pproved a pplications by e very f ive y ear s from 1980 to 2009

PAGE 133

133 Figure 5 5 the f inal d ecisions from the BZA Figure 5 6 the f inal d ecisions from the BZA by t ype of v ariance

PAGE 134

134 Figure 5 7 The Office of Planning i nvolvement by e v ery f ive y ear s Figure 5 8 The ANC s i nvolvement by e very f ive y ear s

PAGE 135

135 Figure 5 9 The c ommunity i nvolvement by e very f ive y ear s Figure 5 10 Community o pposition r ate s in c ases where the c ommunity was involved by y ear

PAGE 136

136 Figure 5 11 Comparisons between d enial r ate, the Office of Planning o pposition r ate, the ANCs o pposition r ate and c ommunity o pposition r ate by y ear from 2000 to 2009 Figure 5 12 Use of p roperties r equested for z oning r elief

PAGE 137

137 Figure 5 13 Approval r ate by p roperty u se Figure 5 14 Number of v ariance a pplications by p roperty u se by y ear

PAGE 138

138 Figure 5 15 Distributions of v ariance a pplications from 1980 to 2009 by Ward

PAGE 139

139 Figure 5 16 Distribution of a pproved and d enied v ariance a pplications from 1980 to 2009

PAGE 140

140 Figure 5 17 Distributions of v ariance a pplication s and d ecisions in Ward 1 from 1980 to 2009

PAGE 141

141 Figure 5 18 Distributions of v ariance a pplications and d ecisions in Ward 2 from 1980 to 2009

PAGE 142

142 Figure 5 19 Di stributions of v ariance a pplications and d ecisions in Ward 2 by d ecennial p eriods

PAGE 143

143 Figure 5 20 Distributions of v ariance a pplications and d ecisions in Ward 3 from 1980 to 2009

PAGE 144

144 Figure 5 21 Distributions of v ariance a pplications and d ecisions in Ward 4 from 1980 to 2009

PAGE 145

145 Figure 5 22 Distributions of v ariance a pplications and d ecisions in Ward 5 from 1980 to 2009

PAGE 146

146 Figure 5 23 Distributions of v ariance a pplications and d ecisions in Ward 6 from 1980 to 2009

PAGE 147

147 Figure 5 24 Distributions of v ariance a pplications an d d ecisions in Ward 7 from 1980 to 2009

PAGE 148

148 Figure 5 25 Distributions of v ariance a pplications and d ecisions in Ward 8 from 1980 to 2009

PAGE 149

149 Figure 5 26 ROC c u rve

PAGE 150

150 Figure 5 27 Map of v ariance a pplication h ot s pots

PAGE 151

151 Figure 5 28 Map of l ocations w hich h as m ore t han t wo v ariance a pplications

PAGE 152

152 Figure 5 29 Map of a pproved v ariance c ases h ot s pots

PAGE 153

153 CHAPTER 6 CONCLUSIONS AND RECOMMENDATION F OR FURTHER RESEARCH Summary Zoning variance is a widely appl ied tool in the United States for land use ( Owens & Brueggemann, 2004 p. 283 ) Guided by the general criteria for granting zoning variance stated by state/local legislature, the board of zoning adjustment plays an important role making final decision s based on the exact situation of property. Since sparing exercised and only in rare instances and under exceptional circumstances peculiar in their nature, and with due regard to the main purpose of a zoning ordinance to preserve the property rights of others 1 However, some researchers have noticed that in practice the board is too lenient to the applicants and the rate of approval remain s very high ( "Building Size, Shape ," 1951 ; Re ynolds, 1999 ) Moreover, discrepanc ies between theory and practice appear in the variance application is thought to be a very important factor affecting the deci sion making, especially regarding the reaction of opponents ( "Administrative Discretion in Zoning," 1969 ; Leary, 1958 ; Madry, 2007 ; Shapiro, 1969 ; "Zoning Variances and Exceptions: The Philadelphia Experience," 1955 ) So me researchers indicate that the BZA is reluctant to grant use variance compared to area variance ( Burke & Snoe, 2004 ; Salkin, 2008 ) Some states requ ire the BZA to place great weight 1 Hammond v. Bd. of Appeal, 154 N.E. 82, 83 (Mass. 1926)

PAGE 154

154 reaching a deci sion. However, findings show th at public agencies have a limited influence on the BZA ( "Administrative Discretion in Zoning," 1969 ; Dukeminier & Stapleton, 1962 ; Owens, 2004 ; Shapiro, 1969 ) This dissertation has attempt to examine the factors which affect zoning o n variance applications, as well as to analyze the trends of variance application and determination from the perspective of time dimension. Based on existing literature, four hypotheses were proposed by the author. The first hypothesis was about the general situation of zoning variance decisions : Hypothesis 1 Granting of variance applications is signif icantly higher than denial. The other three hypotheses are concerned with the possible factors which were identified according to existing literature. They are: Hypothesis 2 Opponents from affected decision making; Hypothesis 3 Type of variance application affects the Hypothesis 4 The zoning administrative board does not place significant weight on suggestions from other public agencies during decision making. To assist in answering the se research questions and testing the four hypotheses statistically, a case study was conducted. The author collected and compiled 2140 variance cases decided by the Board of Zoning Adjustment in Washington D C from 1980 to 2009. The author c onducted a comprehensive analysis of variance applications and decisions from the BZA in Washington D.C. in the following aspects: variance applications, decisions from the BZA, area variance approval, use variance approval,

PAGE 155

155 the Office of Planning involvem ents, the the community involvements, use of properties requested for zoning relief, and variances by ward. Hypothesis 1 was tested using the Binomial Test. Hypothesis 2 to 4 were tested in the basic binary response model developed by the author. The results provide d us with comprehensive information and as sisted the author to come to conclusion s about the above four hypotheses. The results of the Binomial Test showed the distribution of dichotomous (approval and denial) data is not th e same as the expected distribution 50% each. Considering that the average approval rate is much higher than the average denial rate, the author came to the conclusion that the granting of variance application is significantly higher than the denial. Th e results of the ba sic binary response model show that the type of variance, recommendations from the Office of Planning, recommendations from the ANCs, opponents present at public hearing s and opposition from other public agencies did affect the ision significantly. The model indicate s use variance application is 1.978 times more likely to be denied by the BZA tha n area variance application; the variance application is 2.916 times more likely to be denied when there is an opponent present at the public hearing when no opponents present at the hearing; the variance application is 24.4 times more likely to be denied when the OP opposes the application than when the OP supports the application; the variance application is 5.5 times more likely to be denied by the BZA when the ANC s oppose the application than when the ANCs support the application. The model predicts that 65.8% of variance applications would be denied by the BZA when there are other public agenc ies opposing the

PAGE 156

156 applications. However, st atistical analysis shows that support from other public agencies does not make any difference in the In summary the OP exerted the strongest influence in affecting the BZA s decision s on zoning variance applications The second stro ngest factor influencing these decisions is the support or opposition offered by the ANCs. This finding is in accordance with the zoning ordinance in Washington D.C. The input of O P and the ANCs assist the BZA in making a fair decision. The influence of o p ponents present at the public hearing is weaker than recommendations from the OP and the ANCs, yet it is stronger than type of variance and recommendation from other public agencies (the OP and the ANCs are excluded from the other public agencies). The se tting of ANCs is peculiar in Washington D.C. N ot many cities in U.S. have similar public agencies in the process of zoning variance. Compared to individuals opposition, the opinions of ANCs which are composed of local residents are less biased. Their stan ce is not concerned with personal benefits. Rather, they are more interested in the cost and benefit of the whole community. Besides the above factors which were tested in the basic binary response model, the author conducted further investigation on othe r possible factors that might also affect the includ ing property area, land value, existing type of property (residential, commercial, mixed use, other), and type of variance applicant (individual, firm, other). These variables were added into the basic binary response model and tested by statistic methods. The statistic results showed that with the except ion of land value, other variables d id not affect the The larger the land value is, the less likely that th e BZA denies the application.

PAGE 157

157 T he c lustering of zoning variance is a spatial representation of the increment of variance in time dimension. Researchers have found evidence for the clustering of variances which might lead to further granting of variances or rezoning. O ne of the major concerns expressed by some scholars is that it would alter the characteristic s of the community and lead to the invalidation of zoning ordinance s On the other hand, it might also signify that the community is no longer suitabl e for existing zoning function. It is necessary to determine whether variance applications are spatially clustered as a the fundamental questions in this research. The author applied hot spot anal ysis in ArcGIS to investigate whether the concentration of variance applications and approved variance applications exist in Wa shington D.C. The results show that for variance applications, cluster exists in Washington D.C. T he concentration of application s was more intense and some variance applications occurred in the same location especially in the center of the District (where CBD is located) The results also reveal that the residential properties are relatively more stable than commercial properties in terms of conforming to zoning ordinance under the condition that the location had previous zoning variance application. As mentioned above, the clustering of variances may signify that the functions of the zone may no longer be suitable for the ar ea and rezoning is needed for the area. Planning measure s need to be taken in the center of Washington D.C. to address the increasing needs for functional change. Especially in commercial areas, where applications for variance concentrated, other land use tools may be applied instead of zoning variance to insert more flexibility and vitality, essential for local economic

PAGE 158

158 development. The clustering of appro ved variance applications is similar to that of variance applications, which implies that there was no appar ent bias from the decisions The spatial concentration of granted variance stems from the spatial concentration of variance applications. This research provides a new method to researchers and practitioners from both the legal field and urban planni ng field to better understand the situation of zoning variance applications and decisions in the United States. Compared to previous research, this study is innovative mainly in the following aspects: (1) point of view concerning the issue of zoning varianc e. Instead of focusing on how practice is diverg es from theory, this study focuses more on the affecting factors. How do these factors affect the decision? What role do urban planners play in this proces s? How zoning variances are spatially represented and what is the inne r meaning of local community? (2) It is t he first time t hat the statistical model is applied to the research area of zoning variance base d on solid and reliable long range data Applyi ng statistical analysis in the social science s has the following advantages : scientific, objective, and re plicable Besides, the model also shows how and to what extent the identified factors affect the s decision, which was not discussed in the prev ious studies. ( 3 ) The spatial ana lysis on the distribution of zoning variance applications and decisions could be a tool for urban planners to monitor the spatial demands of variances and elaborate on how those variances change the local character istic s. I n addition, the concentration of variances might be an indicator that the area needs more flexibility under existing land use regulation

PAGE 159

159 Implications to Planning and Law Previous studies on zoning v ariance mainly came from the legal field, most of which focused on how practic e in zoning variance diver ges from theory. However, as an indispensable part in zoning and comprehensive planning, zoning variance is considered by urban planning researchers. Th is missing part of research from the planning perspectiv e, which is supposed to be the connection between legislation and social practice, might lead to an even la r ger divergence between theory and practice in zoning variance. It is not sufficient to simply judge whether practice is divergent from theory. The e ssential questions are how to make regulations fit in with the rational needs of society The rationality and reality of the factors which affect the decision should be noted and reviewed It is a long range and complicated project in identifying th e factors affect ing the s decision and understanding why they become factors. H ow are these zoning variance s spatially distributed and whether these changes influence the local community? If the influence exists, is it positive or negative? The load is too heavy for planning practitioners to assume this project. Urban researchers should take the responsibility in answering the above questions under the assistance of planning practitioners to provide regulation makers with solid ground. Regulation make rs take the responsibility to establish reasonable regulation s and ensure that they are implemented effectively. This research is the first step to setting up a replicable method to identify the facts and situation s in zoning variances. The statistical met hod could be applied to different areas to identify the situation of zoning variance administration and the affecting factors in the drawn from the case study in Washington

PAGE 160

160 D.C., the author proposes the following qu estions to law makers and urban planners for further discussion s : (1) How do we view the high approval rate of zoning variance applications? Many studies criticize the BZA as too lenient in accepting variance applicants. This stance views the BZA as ey ing f zoning variance. However, th is high approval rate is common across the United States. If it is a problem, the BZA should not be the only agency to be criticized Profound ineffectiveness of policies and regulations may exist within this body Or there is another possibility T he BZA may be is doing exactly what society needs. (2) How to view the opinions? Public hearing is a channel to allow cross examinations between variance applicants and other parties includ ing residents The results in this study show that the BZA is significantly affected by the opponents appearing in public hearing. The variance application is 2.916 times more likely to be denied when there is an opponent present at a public hearing than no opponents being present at a public hearing. However, there is no explicit requirement by the zoning ordinance that the BZA should give credence to opponents from the community in its decision making Though most ordinances require that the variance must not be cont rary to public interest, safety, and welfare, it is hard to say whether the present opponents indicate that the variance application is contrary to public interest, safety, and welfare. On the other hand, neglect of opinions is inadvisable eithe r. It might depend on the better judgment. The main issue is that there is no leg opinions and how to handle the

PAGE 161

161 the question s for both planners and law ma kers. (3) How to maintain the independency and fairness of the BZA ? This study shows that the land value might be another factor which affects the It is unknown what inner reason lead s to this result. It might be that the owners of properties with high land value are capable of hiring experienced attorneys. Since the burden of proof is partially on applicant applications with complete prepar ation might be more likely granted. However, if the cause comes from economic power affecting the judgment measures should be taken to preserve the independency and fairness of the BZA The complication of land use decides that the theory and practice in zoning and zoning variance are not simple. In order to have zoning variance ordinance and the BZA better execute their social functions, practitioners and researcher from both planning and l egal fields should cooperate with each other. Recommendation for Further Research Zoning variance research is an interdisciplinary research project which requires knowledge of urban planning, land use law, public administration and economic development for a complete comprehension. T his dissertation is a beginning to the autho r rather than a finale Lack of complete data experience in practice have limited the depth and scope of this research. More information needs to be acquired and more efforts are need ed for a co mprehensive understandin g of the In the future improvements c an be made in the following aspects:

PAGE 162

162 First, as shown in Table 5 5 the Nagelkerke R Square is 0.406 in the basic binary respon se model. 59.4% of the BZA s fi nal decisions could not be explained by this model. In the section o n Further Investigation the author analyzed some other possible factors that might affect the BZA s decisions. The land value was identified as another factor. However, due to the limit ation of data availability and the method conducted in this research, some possible factors were not tested. As indicated previously in Chapter 3 two methods can be used to collect data regarding the determination zoning variance. The first method is to s ummarize zoning variance cases during a period based on the official documents. Using the date, researchers could investigate the relation between the protestors and letters of opposition from neighborhood type of application The second method is via sur vey. The surveyors dispatch questions to the administrative officials and related agencies. Using this method, some information that cannot be collected by knowledge of planning, engineering, and architecture, attitudes towards protestors, and other public agencies). The ideal approach is to combine the aforementioned methods. This research applied the first method and did not conduct a survey. More factors affect ing the be found if a survey were present. Second, the author selected Washington D.C. as the study area to investigate the It has been explained in Chapter 3 why the author selecte d this city. The four hypotheses were tested using the variance data in Washington D .C., and the author came to conclusions based on th e se data. In order to

PAGE 163

163 know whether those conclusions are prevalent across the United States, th e author needs more time t o study other areas. Third, in Chapter 5 the map ( Figure 5 27 ) shows the areas where variance applications concentrated The occurrence of locations where variance applications spatial ly concentrated coincide with high economic activities and development activities may reflect some inner causal relationship between economic activities and variance applications. It is a question which needs more information for further research. Fourth, in the further investigation section in Chapter 5 the author found that land value is another factor which affects the ince the data for l and value did not cover all 2 140 cases, 1 301 cases which have data o n land value were used for statistical analysis. D ue to limitation of data availability, the results in th at section might not fully represent the facts. In order to get more reliable results, the author s hould use complete data and examine this factor again. Fifth, this research utilized a single case st udy approach The f i nding s o n Washington D.C. cannot be directly applied to other cities. The boards decisions in different jurisdictions may be affected by different factors. Therefore, the conclusion is not intended to be regarded as a universal finding In order to determine whether some factors have the same impact on the board s decision in different cities, the binary response model needs to be applied to other cities in future studies Apart from the aforementioned issues which should be addressed in the further research, the author will d esign a conceptual framework to integrate existing and newly developed land use tools effectively in land use system. The discrepancies be tween the theory and practice of zoning variance indicates that the traditiona l land use

PAGE 164

164 tools/regulations no longer satisfy the social needs i n this rapid changing world. Presently it is difficult to create a brand new land use tool, which can address land use issues better than the traditional zoning system. The integration of ex isting and newly developed land use tools may allow more flexibility without losing the essence of land use planning.

PAGE 165

165 APPENDIX A C RUDE BINARY RESPONSE M ODEL The software SPSS was applied to run the crude binary respons e model, which included all ten independent vari ables. The method of forward stepwise (Wald) was applied to eliminate the variables not significant at 95% confidence level. This method runs 9 steps to get the variables which affect dependent variables statistically significant. In each step a new variab le is added 1 Table A 1 shows the significance, parameters of each variable, Wald test and the odds ratio (Exp(B)) of the model. The sixth column Sig. provides the significance of each variable. At the 95% confi dence level, except the variable of support from communities all other variables passed the Wald test and were included in this model. Omnibus tests verify whether the model with a new added variable is significantly better than the model without a new added variable. Table A 2 shows the results for each step of the binary response model. The Sig column shows whether the model with new added variable in each step is better than the model in the last step. As in dicated from step 1 to step 9, each model with a new added variable is better than the model in last step. Step 9 has included all the variables except CRSupport The variable for the sig nificantly and it is excluded f rom the binary response model. Table A 3 shows the 2 log likelihood statistic and the coefficient of determination R Square. The first column of 2 log likelihood decreases from s tep 1 to step 9 and the R Square in both the second column of Cox & Snell R Square and Nagelkerke R 1 The sequence entered by the variables does not affect the final results. In this model, the sequence is as follows: step 1: CRNoation, Step 2: OPSupport, Step 3: OPNoaction, Step 4: ANCSupport, Step 5: ANCnoaction, Step 6: Use, Step 7: HearingOppose, Step 8: OASupport, Step 9: OAOppose

PAGE 166

166 Square increases from step 1 to step 9. These indicate that the model with new added variable in each step is improving at predicting the on than the model in last step. R square in OLS regression measures to what extent a model can explain the dependent variable. In logistic regression, there is no true R 2 value as there is in OLS regression ( Newsom, 2010 p. 1 ) The Cox & Snell R square and Nagelkerke R square are two methods to represent the function of R square in OLS regression. The larger the value is, the more the model explains the dependent variable. Howev er, the maximum value of The Cox & Snell R square is less than 1. Since Nagelkerke R square ranges from 0 to 1, it is preferred in model explanation. In step 9, the Cox & Snell R square is 0.238 and Nagelkerke R square is 0.416. This indicates that the in dependent variables in our model can explain 41.6% of the Hosmer Lemeshow test is used to test the overall model fit. The null hypothesis is employed so that the model fits the data well. As show n in Table A 2 in step 9 the Sig<0.05, which means the null hypothesis is rejected. Alt hough in the Omnibus tests nine variables were included in thi s model, the model does not pas s the Hosmer Lemeshow test, which means the model should be adjusted. Notice that in Step 5, Step 7, Step 8 and Step 9 the Sig<0.5, multicollinearity might exist in this model. Since all the independent variables are dummy variables, multicollinearity is common in this kind of model. By applying the Pearson test, the author obtained the correlation significance and eliminated two more variables: OASupport and CRNoaction. Both variables are significant in the original model, which means that they affect the significantly. However, since they are correlated to o ther variables, they include both the

PAGE 167

167 causes that bias prediction. On the other hand, the correlation shows that the views between other public agencies are consistent with that of the Office of Planning. The output also shows that no other form of action from the community is negatively correlated with opposition in public hearing. Thi s outcome is also reasonable, implying that the number of cases which had support from the community was so small that it c an be safely neglected in the model. Furthermore f or the cases in which opponents appeared at a public hearing, other forms of opposition from the community might also take place.

PAGE 168

168 Table A 1 Significance and p arameter of c rude b inary r esponse m odel Variable B S.E. Wald Sig. Exp (B) Use 0.657 0.183 12.962 0 1.929 OPSupport 3.091 0.222 193.072 0 0.045 OPNoaction 2.298 0.181 160.419 0 0.101 ANCSupport 1.652 0.229 52.217 0 0.192 ANCnoaction 0.805 0.212 14.479 0 0.447 HearingOppose 0.667 0.215 9.664 0.002 1.948 CRNoaction 0.663 0.193 11.846 0.001 0.515 OASupport 0.925 0.378 5.971 0.015 0.397 OAOppose 0.645 0.323 3.983 0.046 1.906 Constant 1.257 0.261 23.207 0 3.515

PAGE 169

169 Table A 2 Omnibus t ests of m odel c oefficients by s tep Chi square df Sig. Step 1 Step 141.754 1 .000 Block 141.754 1 .000 Model 141.754 1 .000 Step 2 Step 103.927 1 .000 Block 245.680 2 .000 Model 245.680 2 .000 Step 3 Step 222.476 1 .000 Block 468.156 3 .000 Model 468.156 3 .000 Step 4 Step 53.951 1 .000 Block 522.108 4 .000 Model 522.108 4 .000 Step 5 Step 26.656 1 .000 Block 548.764 5 .000 Model 548.764 5 .000 Step 6 Step 13.107 1 .000 Block 561.872 6 .000 Model 561.872 6 .000 Step 7 Step 10.236 1 .001 Block 572.107 7 .000 Model 572.107 7 .000 Step 8 Step 6.864 1 .009 Block 578.971 8 .000 Model 578.971 8 .000 Step 9 Step 3.875 1 .049 Block 582.846 9 .000 Model 582.846 9 .000

PAGE 170

170 Table A 3 2 Log l ikelihood and R s quare by s tep Step 2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 1677.816(a) .064 .112 2 1573.889(b) .108 .189 3 1351.413(a) .196 .343 4 1297.462(b) .216 .378 5 1270.805(b) .226 .395 6 1257.698(b) .231 .403 7 1247.462(b) .235 .410 8 1240.598(b) .237 .414 9 1 236.723(b) .238 .416 a Estimation terminated at iteration number 5 because parameter estimates changed by less than .001. b Estimation terminated at iteration number 6 because parameter estimates changed by less than .001.

PAGE 171

171 APPENDIX B C ATEGORIES OF VARIANC E REQU ESTED Table B 1 Categories of v ariance r equested Category Category 1 Category 2 Lot lot occupa nc y lot width lot area rear yard front yard side yard lot subdivision theoretical lot second principle stru cture on a single lot number of buildings in the same lot Building Principal Building height story floor area ratio gross floor area roof structure setback number of roof structures arcade penthouse nonconforming building nonconf orming structure addition enlarge nonconforming building Accessory Building structure height story accessory structure floor area ratio Court open court width area closed court width rear yard area Parking parking parking space location size of parking spaces off street parking number of off street parking acc essor y parking space paving materials

PAGE 172

172 Table B 1. Continued Category Category 1 Category 2 Street street frontage street wall Alley alley wid th alley structure height alley set back building on alley lot Location district boundary location of gasoline service location of accessory use and building S pace residential recreation space residential open space ground level public space maneuvering and accessibility space retail space L oading loading berth loading space off street loading facility Other antenna structural change for community service center number of persons in rehabilitation home t iming of combined lot development landscaping entrance access aisle transferable development right driveway

PAGE 173

173 APPENDIX C HOT SPOT ANALYSIS This part presents the theory of hot spot analysis and its application on variance applications and decisions c luster analysis. Figure C 1 shows the statistic model of hot spot analysis For each location the model returns a value called Z score. value (probability) for a feature indicates a spatial clustering of high values. A low nega tive Z score and small p value indicates a spatial clustering of low values. The higher (or lower) the Z score, the more intense the clustering. A Z score near zero indicates no apparent spatial ( ESRI 2009 ) The purpose of using hot spot analy sis in this research is to determine whether cluster of variance applications and decisions exist. To identify cluster, we focus on a relatively larger scale rather than individual cases. By combining the nearby cases occurr ing in a certain distance we ge t the aggregated data which we assume happened in the same location In this study, 200 feet 1 was used as the threshold distance. The second step wa s to count the number of cases which occupied the same location s by under Spati al Statistics Tools Utilities Toolset Then the author applied hot spot analysis with rendering under ArcGIS Spatial Statistic Tools Rendering Toolset Figure C 2 shows the map of Z score of aggregated variance application. For a better view of cluster the author applied Inverse Distance Weighted tool under Spatial Analyst Toolbox and obtained Figure 5 27 The interpretation of the 1 In zoning variance application it is required to inform owners of property within 200 feet of the site.

PAGE 174

174 color on the map was explained under the section Clustering of Zoning Variance i n Chapter 5 Hot spot analysis for approved variance applications also appl ied the same process as that of variance application.

PAGE 175

175 Figure C 1 T he s tatistic m odel of h ot s pot a nalysis 2 2 Source: ( ESRI 2009 )

PAGE 176

176 Figure C 2 M ap of Z s core of a ggregated v ariance a pplication s

PAGE 177

177 LIST OF REFERENCE S Abrahams, C., & Zhang, M. (2008). Fair lending compliance: intelligence and implications for credit risk Hoboken: John Wi ley & Sons, Inc. Administrative d iscretion in z oning. (1969). Harvard Law Review, 82 (3), 668 685. Babcock, R. F. (1966). The z oning g ame m unicipal p ractices and p olicies Madison: University of Wisconsin Press. Bailey, M. J. (1959). Note on the e conomi cs of r esidential z oning and u rban r enewal. Land Economics 35 (3), 288 292. Barry, C. J. (1993). The p olitical c ulture of p lanning a merican land use planning in comparative perspective New York: Routledge. Bassett, E. M. (1922). Zoning New York Natio nal Municipal League. Benjaminson, P., & Anderson, D. (1990). Investigative r eporting (2nd ed.). Ames: Iowa State University Press. Building s ize, s hape, and p lacement r egulations: b ulk c ontrol z oning r eexamined. (1951). The Yale Law Journal,, 60 (3), 506 528. Burke, B., & Snoe, J. (2004). Property (2nd ed.). New York: Aspen Publishers. Byrne, D., & Ragin, C. C. (2009). The SAGE h andbook of c ase based m ethods : SAGE Publications. Caemmerer, H. P. (1939). A m anual on the o rigin and d evelopment of w ashingt on Washington: United States Government Printing Office. Cohen, J. E. (1995). A c onstitutional s afety v alve: t he v ariance in z oning and l and use b ased e nvironmental c ontrols. Boston College Environmental Affairs Law Review, 22 (2), 307 364. Dennis, R. (2 000). 'Zoning' before z oning: the r egulation of a partment h ousing in e arly t wentieth c entury w innipeg and t oronto. Planning Perspectives, 15 267 299. District of Columbia C ourt of A ppeals. ( 1972) P almer v. board of zoning adjustment. Retrieved May 15, 2 010, from http://174.123.24.242/leagle/xmlResult.aspx?xmldoc=1972822287A2d535_1814. xml&docbase=CSLWAR1 1950 1985 District of Columbia O ffice of P lanning. (2007). District of Columbia s tatistical h andbook. (2000 2005) Washington D.C.: Author

PAGE 178

178 District of Columbia O ffice of P lanning. (2007). The 2006 Comprehensive p lan. Retrieved April 27 2011 from http://planning.dc.gov/DC/Planning/Across+the+City/Comprehensive+Plan/2006+ Comprehensive+Plan District of Columbia O ffice of the C hief T echnology O fficer (2009). Square, s uffix, l ot (SSL) Retrieved April 10, 2011 from http://dcaddresscoordinates.blogspot.com/2009/08/square suffix lot ssl.html District of Columbia O ffice of the C hief T echnology O ffi cer (2008). Vector p roperty m ap m odernization. Retrieved April 10, 2011 from http://vpm.dc.gov/ District of Columbia O ffice of Z oning (1974). Application n o. 11000 of the c lerics of s aint v iator, i nc. Retrieved February 2 0 201 1 from http://dcoz.dc.gov/orders/11000_3926 819.pdf Distr ict of Columbia O ffice of Z oning (1974). Application n o. 11511 of e leanor a hrens Retrieved February 2 0 201 1 from http://dcoz.dc.gov/orders/11511_1242 120.pdf District of Columbia O ffice of Z oning (1975 ). Application n o. 11925 of d iscalced c armelite f athers, i nc. Retri eved February 2 0 201 1 from http://dcoz.dc.gov/orders/11925_3539 801.pdf District of Columbia O ffice of Z oning (1981 ). Application n o. 13405 of r uth s and s amuel w illiams. Retrieved February 2 0 201 1 from http://dcoz.dc.gov/orders/13405_891 77.pdf District of Colum bia O ffice of Z oning (1983 ). Application n o. 14019 of a my g oldstein and r uth s mall. Retrieved February 2 0 201 1 from http://dcoz.dc.gov/orders/14019_247 852.pdf District of Columbia O ffice of Z oning (1992 ). Application n o. 15605 of j ohn and v innie b est. Retrieved February 2 0 201 1 from http://dcoz.dc.gov/ord ers/15605_9203 204.pdf District of Columbia O ffice of Z oning (2001 ). Application n o. 16654 of the l ucy w ebb h ayes t raining s chool for d eaconesses and m issionaries d/b/a s ibley m emorial h ospital. Retrieved February 2 0 201 1 from http://dcoz.dc.gov/orders/16654_N 1448 25.p df District of Columbia O ffice of Z oning ( 2008 ). Form 121 applicant s burden of proof for variance and variance special exception applications. Retrieved February 25, 201 1 from http://dcoz.dc.gov/services/pdf/BZA/Form121.pdf

PAGE 179

179 District of Columbia O ffice of Z oning ( 2011 ) F orm 140 party status request Retrieved May 25, 201 1 from http://dcoz.dc.gov/services/pdf/BZA/Form140.pdf Donovan, T. B. (1962). Zoning: v ariance a dministration in a lameda c ounty. California Law Review, 50 (1), 101 120. Dukeminier, J. (2002). Property (5th ed.): Aspen. Dukeminier, J., & Stapleton, C. L. (1962). The z oning b oard of a djustment: a c ase s tudy in m isrule. Kentucky Law Journal, 50 273 350. Durkin, C. (2006). The e xclusionary e ffect of m ansionization": a rea v ariances u ndermine e fforts to a chieve h ou sing a ffordability. Catholic University Law Review, 55 439 472. ESRI. (2009). ArcGIS d esktop h elp 9.3. Retrieved February 25, 2010, from http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?TopicName=Hot_Spot_Anal ysis_%28Getis Ord_Gi*%29_%28Spatial_Statistics%29 Fischel, W. A. (2004). An e conomic h istory of z oning and a c ure for its e xclusionary e ffects. Urban Studi es, 41 317 340. Gardner, E. (2004). To d efer or n ot to d efer: j udicial r eview of z oning b oard d ecisions in n ew y ork. Cardozo Public Law, Policy and Ethics Journal, 2 421 468. Gerring, J. (2007). Case s tudy r esearch p rinciples and p ractices New York: Cambridge University Press. Groves, J. R., & Helland, E. (2002). Zoning and the d istribution of l ocation r ents: a n e mpirical a nalysis of h arris c ounty, Texas. Land Economics, 78 (1), 28 44. Heffley, D. R., & Hewitt, D. P. (1988). Land u se z oning in a l oc al e conomy with o ptimal p roperty t axes and p ublic e xpenditures. Journal of Real Es tate Finance and Economics 1 373 391. Helpman, E., & Pines, D. (1977). Land and z oning in an u rban e conomy: f urther r esults. The American Economic Review, 67 (5), 982 986. Horowitz, J. L., & Savin, N. E. (2001). Binary r esponse m odels: l ogits, p robits and s emiparametrics. Journal of Economic Perspectives, 15 43 56. Jacobs, F. D. (1958). Municipal c orporations: z oning: t he g ranting of a v ariance b ased on u nnecessary h ardsh ip. Michigan Law Review, 56 (5), 820 823. Juergensmeryer, J. C., & Roberts, T. E. (2003). Land u se p lanning and d evelopment r egulation l aw St. Paul: West Group.

PAGE 180

180 Kent, D. L. (1993). The p resumption in f avor of g ranting z oning v ariances. New Hampshire Bar Journal, 34 29 34. Kress, J. R. (2008 ). Zoning in the d istrict of c olumbia Retrieved May 15, 2010, from http://dcoz.dcgov.org/Brochureforweb2008.pdf Leary, R. M. (1958). A m odel p rocedure fo r the a dministration of z oning r egulations Washington D.C.: Urban Land Institute. Levine, N. & Associates. (2010 ). A s patial s tatistics p rogram for the a nalysis of c rime i ncident l ocations Washington, D.C.: The National Institute of Justice. Retrieved N ovember 15, 2010, from http://www.icpsr.umich.edu/CrimeStat/ Madry, A. R. (2007). Judging z iervogel: the t wisted p ath of r ecent z oning v ariance d ecisions in w isconsin. Marquette Law Review, 91 485 53 4. Mandelker, D. R. (1997). Land u se l aw (Fourth ed.). Charlottesville: LEXIS Law Publishing. Mandelker, D. R. (2003). Land u se l aw : Mattew Bender & Company, Inc. Mandelker, D. R., Payne, J. M., Salsich, P. W., & Stroud, N. E. (2005). Planning and c ontr ol of l and d evelopment: c ases and m aterials (6th ed.): LexisNexis. Newsom, J. (2010). More on m odel f it and s ignificance of p redictors with l ogistic r egression. Retrieved November 15, 2010, from http://www.upa.pdx.edu/IOA/newsom/da2/ho_logistic3.pdf Office of the General Counsel (1973). District of C olumbia h ome r ule a ct Retrieved November 20, 2010, from http://www.abfa.com/ogc/hrta ll.htm Owens, D. W. (2004). The z oning v ariance: r eappraisal and r ecommendations for r eform of a m uch m aligned t ool. Columbia Journal of Environmental Law, 29 (2), 279 322. Owens, D. W., & Brueggemann, A. (2004). A s urvey of e xperience with z oning v arian ces. UNC School of Government 40. Retrieved January 20, 2011, from http://sogpubs.unc.edu/electronicversions/pdfs/zonvar.pdf Peel, R. V. (1939). L ewis, h arold m aclean c ity p lanni ng w hy and h ow. Pp. xxi, 257. New York: Longmans, Green and Co., 1939. $2.50. The Annals of the American Academy of Political and Social Science 158 159. Perin, C. (1977). Everything in i ts p lace: s ocial o rder and l and u se in a merica Princeton: Prince ton University Press.

PAGE 181

181 Pogodzinski, J. M., & Sass, T. R. (1990). The e conomic t heory of z oning: a c ritical r eview. Land Economics, 66 (3), 294 314. Psota, K. (2005). To b uild or n ot to b uild, t hat is the q uestion d ecided by the p ennsylvania s upreme c ourt i n n ettleton V. z oning b oard of a djustment. W idener Law Journa I 14 537 550. Replacing the h ardship d octrine: a w orkable, e quitable t est for z oning v ariance (1987). Connecticut Law Review, 20 669 721. Reynolds, O. M. (1999). The u nique c ircumstances" r ule in z oning v ariances a n a id in a chieving g reater p rudence and l ess l eniency. The Urban Lawyer, 31 (1), 127 148. Rice, T. (2006). 2004 2005 s urvey of n ew y ork l aw: z oning and l and u se. Syracuse Law Review, 56 1101 1150. Rohan, P. J. (1990). Zoning an d l and u se c ontrols New York: Matthew Bender. Salkin, P. E. (2008). American l aw of z oning (5th ed. Vol. 13). Thomson Reuters Sampson, R. W. (2007). Theory and p ractice in the g ranting of d imensional l and u se v ariances: i s the l egal s tandard c onscienti ously a pplied, c onsciously i gnored, or s omething in b etween? The Urban Lawyer, 39 877 958. Schmutz, G. L. (1931). Economic e ffects of z oning. Annals of the American Academy of Political and Social Science 155 Seeking a v ariance as a p rerequisite to c h allenging a z oning o rdinance. (2000). The Wayne Law Review, 45 2011 2038. Shapiro, R. M. (1969). The z oning v ariance p ower c onstructive in t heory, d estructive in p ractice. Maryland Law Review, XXIX 3 24. Shlay, A. B., & Rossi, P. H. (1981). Keeping u p the n eighborhood: e stimating n et e ffects of z oning. American Sociological Review, 46 (6), 703 719. State of New York Legislative Bill Drafting Commission (2004). New y ork c ity c harter Retrieved January 15, 2011, from http://www.nyc.gov/html/charter/downloads/pdf/citycharter2004.pdf Steele, E. H. (1986). Participation and r ules The f unctions of z oning. American Bar Foundation Research Journal, 11 (4), 709 755. Swan, H. S. (1949). Economic and s ocial a spects of z oning and c ity p lanning. American Journal of Economics and Sociology, 9 (1), 45 56.

PAGE 182

182 Wolf, M. A. (2008). The z oning of a merica: e uclid v. a mbler : University Press of Kansas. Wooldridge, J. M. (2002). Econometric a nal ysis of c ross s ection and p anel d ata Cambridge: The MIT Press. Wuensch, K. L. (2011). Binary l ogistic r egression with PASW/SPSS. 2011, Retrieved August 20 20 11 from http://c ore.ecu.edu/psyc/wuenschk/MV/Multreg/Logistic SPSS.pdf Yin, R. K. (1984). Case s tudy r esearch d esign and m ethods Beverly Hills: SAGE Publications. Zoning for the r egional w elfare. (1980). The Yale Law Journal, 89 (4), 748 768. Zoning v ariance and the n ew y ork c ity b oard of s tandards and a ppeals. (2005). Columbia Journal of Environmental Law, 30 (1), 193 247. Zoning v ariances. (1961). Harvard Law Review, 74 (7), 1396 1407. Zoning v ariances and e xceptions: t he p hiladelphia e xperience. (1955). University of Pennsylvania Law Review, 103 (4), 516 555.

PAGE 183

183 BIOGRAPHICAL SKETCH Jun Zhao was born in Taiyuan China in 198 3 She received her BA in International Economics &Trade from Nankai University in 2005. In 2007, she earned her MA in Regional Economics from Nan kai University The same year Jun began her doctoral study in the United States in the Department of Urban and Regional Planning at the University of Florida. She received her Ph.D. from the University of Florida in the fall of 2011 Her scholarly research interests are theory, policy and practice in land use planning. She is also interested in economic development, GIS and statistical application in planning, housing studies, and environmental planning.