THE INFLUENCE or REAPPORTIDNMENT '3 fiijf. 37:13} ON THE DISTRIBUTION 0F , I. ‘ STATE AID T0 EDUcATION j Them for. the DegEee of Ph D MICHIGAN STATE UNIVERSITY STEPHEN WHEELER BURKS LIBRARY Michigan State University . .___E ”Wag—I“ 1-“...- .'-.'-' ‘- This is to certify that the thesis entitled THE INFLUENCE OF REAPPORTIONMENT 0N THE DISTRIBUTION OF STATE AID T0 EDUCATION presented by Stephen Wheeler Burks \ has been accepted towards fulfillment of the requirements for Doctorate degree in Philosophy Major professor Date W1 0-7839 mm: a 35%") BOOK BINDERY MEI - I 8W" amen: “i lI This study ewes the th'IIution of state aid to 10‘ ate. the central hYPOtheSis ishte aid between metropolit Emmet}! 1962 and 1969 3-! retial voting power of metrc 3'Ilting from reapportionment Mlitan legislative powe Eislature that are located in Esra is refined to account Elation and socio-economi Wail analysis of post-re The results are gene: ABSTRACT THE INFLUENCE OF REAPPORTIONMENT ON THE DISTRIBUTION OF STATE AID TO EDUCATION BY Stephen Wheeler Burks This study examines the influence of reapportionment on the distribution of state aid to local school districts in twenty—six states. The central hypothesis is that changes in the distribution of state aid between metropolitan and non—metropolitan school dis— tricts between 1962 and 1969 are directly related to changes in the potential voting power of metropolitan counties in state legislatures resulting from reapportionment. The major independent variable is ll "metropolitan legislative power, or the proportion of seats in the legislature that are located within SMSA counties. The basic power measure is refined to account for variations in central city—suburban populations and socio—economic heterogeneity, and used in a cross- sectional analysis of post—reapportionment aid patterns. The results are generally inconclusive. While several of the most malapportioned states show dramatic shifts in aid toward metro— politan school districts, the overall pattern is much more dispersed, Showing both significant changes in states with minor shifts in legis— lative power and minor changes in states with large shifts in power. 1"- : of a Clear re] increased metro] album: 1) the unavaila um school districts ”I ‘ suiting imrecision of the J giiul. incremntal, and admin' smallish reduces the influ :s‘ute aid appropriations; and mm the impact on the (1: finally been assumed. Stephen Wheeler Burks The absence of a clear relationship between increased metro— politan power and increased metropolitan state aid are attributable to several factors: 1) the unavailability of data showing the correspon— dence between school districts and state legislative districts, and the resulting imprecision of the legislative power measure; 2) the cyclical, incremental, and administrative character of the budgetary process, which reduces the influence of legislative power arrangements on state aid appropriations; and 3) the fact that apportionment systems do not have the impact on the distribution of state aid that has tra— ditionally been assumed . Mich: in partial in DC Depart THE INFLUENCE OF REAPPORTIONMENT ON THE DISTRIBUTION OF STATE AID TO EDUCATION BY Stephen Wheeler Burks A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Political Science 1973 amnpflfimOfU can he mudged I flhllw are to the PeoPle th‘ members of my disser‘ In. and genes H5554” thess clarified much of “Y ’ Edible large quantities of 5' buffered frequent reassura Haven," and his detailed Elude for a much improved ihlryan. As my major pro Wing and conscientious. cc MOI. More importantly. a: diminities for professiona ialttlht account for much oi Wail science. “to extraordinary pe ‘d . 5“ Lansmg much more tr aft 64?“ ACKNOWLEDGMENTS AND DEDICATION Only a small portion of the many debts one incurs in writing a dissertation can be acknowledged in print. The four sets of thanks that follow are to the people who helped the most along the way. The members of my dissertation committee—~Bryan Downes, Frank Pinner, and Charles Press——have been uniformly supportive and helpful. Mr. Press clarified much of my fuzzy thinking on apportionment and made available large quantities of source material from his own study. Frank offered frequent reassurances during the project that "all would yet be well," and his detailed and incisive criticisms of the first draft made for a much improved final product. I am particularly grate- ful to Bryan. As my major professor, he provided the good—natured prodding and conscientious, constructive criticism needed to complete the job. More importantly, as a friend he gave me encouragement and Opportunities for professional development while still a graduate stu— dent that account for much of the good fortune that I have had in political science. Two extraordinary people, John and Martha Seaman, made our stay in East Lansing much more than the typical graduate school “layover.“ Their warmth and hospitality provided us with the proverbial "home away from home," and their contagious zest for life frequently prevented an ii mine. from taking himself ITS remember their kindness. suite haircaret had the “I W3 i0: tom and a half year intends to ponder her dads-1c ihghqqood humor, patience, J New complete the refinement noedit for a degree than She intist ever had a better part1 Finally, my family. MY ETIEIIlth a rush typing job 01 :ever mentioning that she wa: Ration. Without her help I Escript to send to the commi Gist friend and sharpest cr {W the years have shaped hunts. Paul and Edith Bu: them they gave me the f and to do, even when they MI father placed grea he ' IIII life was a standard I ilugh to see me complete th aqfiring Ph.D. from taking himself and his work too seriously. We will aways remember their kindness. My wife Margaret had the misfortune of marrying an academic "gfindfl' For four and a half years now she has had many lonely nights amiweekends to ponder her decision. Throughout, she has maintained the unflagging good humor, patience, and understanding that has made it pos- flble to complete the requirements for the degree. No wife ever shared nbre credit for a degree than she does for this one, and no political scimnfist ever had a better partner with whom to ply his trade. Finally, my family. My aunt, Patricia Beuter, came to the reanm with a rush typing job of a large part of the initial draft with— omzever mentioning that she was in the midst of recovering from an eye cmeration. Without her help I would probably still be waiting for a manuscript to send to the committee. My brother Alan has always been Hy best friend and sharpest critic, and our heated political discussions fluough the years have shaped many of my interests in political science. My parents, Paul and Edith Burks, taught me to think and care and crit— icize, and they gave me the freedom to make mistakes and to do what I rmeded to do, even when they didn‘t understand. Few sons are so lucky. My father placed great value on integrity and excellence, and hksown life was a standard of the highest order. He did not live long mmugh to see me complete the degree, but I think that he would have beaipleased and proud. It is to his memory that the dissertation is dedicated. iii 1. REVIEW AND CRITIQUE ‘ II. THE muomncu CONTEX III. minimum SYSTEMS W. IIETRDPDLITAN panama I- hummus EXPLANATI VI- THE cmmmnve exam I'11- CONCLUDING earners. TABLE OF CONTENTS LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . Chapter I. A REVIEW AND CRITIQUE OF THE LITERATURE . . . . . o II. THE THEORETICAL CONTEXT . . . . . . . . . . . . . III. APPORTIONMENT SYSTEMS AND STATE AID TO EDUCATION. IV. METROPOLITAN DEMOGRAPHY AND STATE AID TO EDUCATION. V. ALTERNATIVE EXPLANATIONS OF STATE AID PATTERNS. . VI. THE CUMULATIVE EXPLANATORY POWER OF THE MODEL . . VII. CONCLUDING REMARKS. . . . . . . . . . . . . . . . BIBLIOGRAPHY...................... APPENDIX........................ iv Page viii 25 49 80 101 132 156 171 189 H. State Percent of Putin-cl III-l. NIIII mm“ PM mpfltan Lesislal mlitan Metre] 31-1. 1960 retropolitan 9°?“ Ketmpolitan Legisla retropolitan RePreS‘ :I-I. letropolitan Legislat: the Percent Increas EH. Changes in 116112090155t from High to Low . III. 1962 State per Pupil lid. Ranked. . . . 31-6. 1962 Metropolitan RE 3H. Percent Increase in 1962-1969, Ranked 11'3- Percent Increase in 1962-1969, Ranked 3H. 1960 Large City Ne- Ranked . taco 1y. 2- Percent Non-Whites 1960, Ranked . 3 y. I 3. Correlation Coeff Variables and S I'M. Correlation Coefi Variables and I Table II—l. III-l. III—2. III—3. III-4. III-5. III—6. III-7. III—8. IV-1. IV—3. LIST OF TABLES Page State Percent of Public School Expenditures, 1970. . . 35 1960 Metropolitan Population (percent), 1962 Metropolitan Legislative Power, and 1962 Metropolitan Underrepresentation . . . . . . . . . . 50 1960 Metropolitan Population (percent), 1967 Metropolitan Legislative Power, and 1967 Metropolitan Representation. . . . . . . . . . . . . 53 Metropolitan Legislative Power in 1962 and 1967, and the Percent Increase following Reapportionment . . . 58 Changes in Metropolitan Legislative Power, Ranked from High to Low . . . . .,. . . . . . . . . . . . . 61 1962 State per Pupil Aid and Metropolitan per Pupil Aid, Ranked. . . . . . . . . . . . . . . . . . . . . 64 1962 Metropolitan Relative Advantage, Ranked . . . . . 66 Percent Increase in Metropolitan per Pupil Aid, 1962—1969, Ranked. . . . . . . . . . . . . . . . . . 72 Percent Increase in Metropolitan Relative Advantage, 1962—1969, Ranked. . . . . . . . . . . . . . . . . . 76 1960 Large City Metropolitan Population (percent), Ranked . . . . . . . . . . . . . . . . . . . . . . . 83 Percent Non-Whites in Large MetrOpolitan Cities in 1960, Ranked . . . . . . . . . . . . . . . . . . . . 86 Correlation Coefficients for 1962 Apportionment Variables and State Aid Variables. . . . . . . . . . 91 Correlation Coefficients for 1967 Apportionment Variables and 1969 State Aid Variables . . . . . . . - ' fl 1' “913551011 Statistics 1 } mwlitan Rel“ H. Percent of Students “'1 1967, med - ' ‘ ' H. Ranney Index of Party ‘ H. 1961 Party Competition Ranked . . . . . . . H. Regression Statistics 1969 Metropolitan p 1'5. Regression Statistics 1969 Metropolitan l H- Regression Statistic: Changes in Metropo ‘H. Regression Statistic Changes in Metropt 1962-1969. . . . 1H. Correlation Matrix Regression Model “~2- Coeffieients of mi Independent Vari per Pupil Aid Re “‘3‘ Backward Eliminati Variable Change “‘4‘ Cumulative Explan Statistics for in Metropolitar \'1~5, c("aff'lcierxts of 1 Independent Va Relative Adv“ LIST or TABLES (cont.) Table Page IV-5. Correlation Coefficients for Changes in Apportionment Variables and Changes in State Aid Variables . . . . 92 IV-6. Regression Statistics for Apportionment Variables and Metr0politan per Pupil Aid . . . . . . . . . . . 96 IV—7. Regression Statistics for Apportionment Variables and Metropolitan Relative Advantage. . . . . . . 98 V-1. Percent of Students Attending Public Schools in 1967, Ranked . . . . . . . . . . . . . . . . . . . . 112 V-2. Ranney Index of Party Competition, Ranked. . . . . . 116 V-3. 1967 Party Competition in State Legislatures, Ranked . . . . . . . . . . . . . . . . . . . . . . 119 V-4. Regression Statistics for Control Variables and 1969 Metropolitan per Pupil Aid. . . . . . . . 125 V'5- Regression Statistics for Control Variables and . 126 1969 Metropolitan Relative Advantage . . . . V‘6- Regression Statistics for Control Variables and Changes in Metropolitan per Pupil Aid, 1962—1969 . . 127 V‘7- Regression Statistics for Control Variables and Changes in Metropolitan Relative Advantage, 1962—1969. . . . . . . . . . . . . . VI‘l- COrrelation Matrix for Variables in State Aid Change 134 Regression Model . . . . . . . - 128 . n o o . VI‘Z- Coefficients of Multiple Determination for ' Independent Variables in Change in Metropolitan 135 per Pupil Aid Regression Model . . . . . VI—3. Backward Elimination Procedure for Dependent . 36 Variable Change in Metropolitan per Pupil Ald. . . . 1 Regression bles in Change gression Model . . V174- Cumulative Explanatory Power and Statistics for Independent Varia in Metropolitan per Pupil Aid Re 138 VI‘5. Coefficients of Multiple Determination for . Metropolitan Independent Variables in Change in 139 Relative Advantage Regression Model. . . . . . . . . vi Regression Model . .1 H. Coefficients of “‘11-‘51? Independent Variable Aid Regression Fade] R-lR. mm Elimination 1 Variable Metropolitt‘ fi-ll. Cmulative Explanator Statistics for Inde Metropolitan per Pr 71-12. Coefficients of Mult: Independent Variab? Advantage Regressi 11-13. Radnard Elimination Variable Metropoll' fl'14- Cumulative Explanatl Statistics for In Metropolitan Rela LIST OF TABLES (cont.) Tafle VI-6. VI-7. VI-B. VI-9. VI—lO. VI-ll. VI-lZ. VI-13. VI—l4. *Backward Elimination Procedure for Dependent Variable Change in Metropolitan Relative Advantage. . . . . . . . . . . . . . . . . . . . . . Cumulative Explanatory Power and Regression Statistics for Independent Variables in Change in Metropolitan Relative Advantage Regression Model . . Correlation Matrix for Variables in State Aid Regression Model . . . . . . . . . . . . . . . . . . Coefficients of Multiple Determination for Independent Variables in Metropolitan per Pupil Aid Regression Model . . . . . . . . . . . . . . . . Backward Elimination Procedure for Dependent Variable Metropolitan per Pupil Aid. . . . . . . . . Cumulative Explanatory Power and Regression Statistics for Independent Variables in Metropolitan per Pupil Aid Regression Model. . . . . Coefficients of Multiple Determination for Independent Variables in Metropolitan Relative Advantage Regression Model . . . . . . . . . . . Backward Elimination Procedure for Dependent Variable MetrOpolitan Relative Advantage . . . . . . Cumulative Explanatory Power and Regression Statistics for Independent Variables in Metropolitan Relative Advantage Regression Model . . vii Page 140 141 144 146 148 149 151 152 154 LIST the 31-1. 1960 Metropolitan P0pul. Legislative Power. . 3-2. 1960 Metropolitan Popul Legislative Power. . E-l. Metropolitan Underrepr 214. Metropolitan Underrepl 21-5. 1962 Metropolitan Leg per Pupil Aid. . . 11-6. 1962 Metropolitan Leg Relative Advantage 34. 1967 Metropolitan Le Metropolitan per P 3'3. 1967 Metropolitan Le Metropolitan Relal 3‘9» Change in Metropoli and Change in Met 31-10. Change in Metropolf and Change in Me‘ 1962—1969. . . . “3'11. Change in Metrop01 1n Metropolitan Il~ 112. Change in MetrOPO 1n Metropolitan RH. 1967 Power 2 and IV. 2. 1967 Rover 2 and IV. 3. 1967 Power 3 and Fimme III-l. III-2. III—3. III-4. III-5. III-6. III-7. III-8. III—9. III-10. III-ll. III-12. IV—1. IV—2. IV—3. LIST OF FIGURES Page 1960 Metropolitan Population and 1962 Metropolitan Legislative Power. . . . . . . . . . . . . . . . . . 189 1960 Metropolitan Population and 1967 Metropolitan Legislative Power. . . . . . . . . . . . . . . . . . 190 Metropolitan Underrepresentation, 1962 . . . . . . . . 191 Metropolitan Underrepresentation, 1967 . . . . . . . . 192 1962 Metropolitan Legislative Power and Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . 193 1962 Metropolitan Legislative Power and Metropolitan Relative Advantage . . . . . . . . . . . . . . . . . 194 1967 Metropolitan Legislative Power and 1969 Metropolitan per Pupil Aid . . . . . . . . . . . . . 195 1967 Metropolitan Legislative Power and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 196 Change in Metropolitan Legislative Power, 1962—1967, and Change in Metropolitan per Pupil Aid, 1962—1969. 197 Change in Metropolitan Legislative Power, 1962—1967, and Change in Metropolitan Relative Advantage, 1962—1969...................... 198 Change in Metropolitan Legislative Status and Change in Metropolitan per Pupil Aid. . . . . . . . . . , . 199 Change in Metropolitan Legislative Status and Change in MetrOpolitan Relative Advantage . . . . . , . _ . 200 1967 Power 2 and 1969 Metropolitan per Pupil Aid . . . 201 1967 Power 2 and 1969 Metropolitan Relative Advantage. 202 1967 Power 3 and 1969 Metropolitan per Pupil Aid . . , 203 viii n OR FIGURES (cont .1 we ..4.. V4. 1967 Rover 3 and 1969 M: Advantage. . . . . . M. 1966 per Capita Persona Metropolitan per Pup] 1-2. 1966 per Capita Person; MetroPolitan Relativ 1-1. 1967 Educational Tax B per Pupil Aid. . . . H. 1967 Educational Tax 1 Relative Advantage 16. Change in Personal In per Pupil Aid. . . H. Change in Income and Advantage. . . . 7-7. 1966 Metropolitan P0 MetrOpolitan Relat 1'8. 1966 Metropolitan Pc Metropolitan Relat 7'9. 1960 Large City Met: thousands) and 19 "10. 1960 Large City Met thousands) and Ch Aid. l962-l969 . g. 11. 1960 Large City Mo thousands) and C‘ Advantage, 1962- ‘H 2. 1960 Large City Me and 1969 Metmpr Percent Non~White 1969 and 1969 M H 4. Percent Non-White 1960 and 1969 h LIST OF FIGURES (cont.) Fimme IV-4. V-l. V-2. V-3. V-4. V-5. V-6. V-7. V-8. V-9. V-lO. V-ll. V—12. V—l3. V-l4. 1967 Power 3 and 1969 Metropolitan Relative Advantage. . . . . . . . . . . . . . . . . . . . . . 204 1966 per Capita Personal Income and 1969 Metropolitan per Pupil Aid . . . . . . . . . . . . . 205 1966 per Capita Personal Income and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 206 1967 Educational Tax Burden and 1969 Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . 207 1967 Educational Tax Burden and 1969 Metropolitan Relative Advantage . . . . . . . . . . . . . . . . . 208 Change in Personal Income and Change in Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . 209 Change in Income and Change in Metropolitan Relative Advantage. . . . . . . . . . . . . . . . . . . . . . 210 1966 Metropolitan Population (in thousands) and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 211 1966 Metropolitan Population (percent) and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 212 1960 Large City Metropolitan Population (in thousands) and 1969 Metropolitan per Pupil Aid . . . 213 1960 Large City Metropolitan Population (in thousands) and Change in Metropolitan per Pupil Aid, 1962—1969 . . . . . . . . . . . . . . . . . . . 214 1960 Large City Metropolitan Population (in thousands) and Change in Metropolitan Relative Advantage, 1962—1969 . . . . . . . . . . . . . . , , 215 1960 Large City Metropolitan Population (percent) and 1969 Metropolitan Relative Advantage . . . _ , ~ 216 Percent Non—Whites in Large Metropolitan Cities in 1969 and 1969 Metropolitan per Pupil Aid . . . . . . 217 Percent Non-Whites in Large Metropolitan Cities in 1960 and 1969 Metropolitan Relative Advantage. , . . 218 ix Relative Advantage . +18. Percent Students in P1 Metropolitan per Pu] H9. Percent Students in P Metropolitan Relati HI. Ranney Index of Party Metropolitan 'per Pt ”1- Ranney Index of Part} Metropolitan Relat: 1-22, Ranney Index of Part Metropolitan per P “3' We}! Index of Part Metropolitan Relat HI. 1967 Party Competit: Pupil Aid. . . . LIST OF FIGURES (cont.) Fimue V-15. V—l6. V-l7. V-18. V-l9. V-20. V-21. V—22. V-23. V-24. V—25. V-26. V—27. V—28. V-29. Page Percent Non—Whites in Large Metropolitan Cities in 1960 and Change in Metropolitan per Pupil Aid, 1962—1969. . . . . . . . . . . . . . . . . . . . . . 219 Change in Population and Change in Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . . . 220 Change in Population and Change in Metropolitan Relative Advantage . . . . . . . . . . . . . . . . . 221 Percent Students in Public Schools in 1967 and 1969 Metropolitan per Pupil Aid . . . . . . . . . . . . 222 Percent Students in Public Schools in 1967 and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 223 Ranney Index of Party Competition and 1969 Metropolitan per Pupil Aid . . . . . . . . . . . . . 224 Ranney Index of Party Competition and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 225 Ranney Index of Party Competition and Change in Metropolitan per Pupil Aid, 1962—1969. . . . . . . . 226 Ranney Index of Party Competition and Change in Metropolitan Relative Advantage, 1962—1969 . . . . . 227 1967 Party Competition and 1969 Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . . . 228 1967 Party Competition and 1969 Metropolitan Relative Advantage . . . . . . . . . . . . . . . . . 229 Milbrath Index of Participation and 1969 Metropolitan per Pupil Aid . . . . . . . . . . . . . 230 Milbrath Index of Participation and 1969 Metropolitan Relative Advantage. . . . . . . . . . . 231 Milbrath Index of Participation and Change in Metropolitan per Pupil Aid, 1962—1969. . . . , _ . . 232 Milbrath Index of Participation and Change in Metropolitan Relative Advantage, 1962—1969 . . . , _ 233 ve Advantage - m m Percent of Total 1969 and 1969 Metropi HI. State Percent of Total 1969 and 1969 Kemp +11. State Percent of Total 1969 and Change in 1 1962-1969. . . . . 195. State Percent of Tota 1969 and Change in 1962-1969. . . . . Page gleaner Vote Nearest 1968 and 1969 Metropolitan per Pupil Aid. . . . . . . . . . . . . . . . . . . . . 234 Governor Vote Nearest 1968 and 1969 Metropolitan Relative Advantage . . . . . . . . . . . . . . . . . . 235 State Percent of Total Public School Revenues in 1969 and 1969 Metropolitan per Pupil Aid . . . . . . . 236 State Percent of Total Public School Revenues in 1969 and 1969 Metropolitan Relative Advantage. . . . . 237 State Percent of Total Public School Revenues in 1969 and Change in Metropolitan per Pupil Aid, 1962-1969. . . . . . . . . . . . . . . . . . . . . . . 238 State Percent of Total Public School Revenues in 1969 and Change in Metropolitan Relative Advantage, 1962-1969. . . . . . . . . . . . . . . . . . . . . . . 239 xi a-j Of the states 11 not the lid and late 19‘ mm the level of seven” windy and effectively a iralrevenue sharing and “9‘ Eris whether the states ha itsponsibility for dealins Belated essentially to 1'1 9316a. The evolution of An :‘iIshas been one of expansf 7lifestation being an incre I‘m Swemmental agencies. rial‘JO‘Iemment as the inev: “logical society .3 0t INT RODUCT ION The role of the states in the American federal system is a subject of growing interest to both political leaders and political scientists. The uncertain success of the massive federal anti—poverty programs of the mid and late 1960's has raised a number of questions concerning the level of government at which such programs are most appropriately and effectively administered.l Current controversy over federal revenue sharing and welfare reform is of a similar order.2 At issue is whether the states have the capability and/or desire to assume the responsibility for dealing with the above types of problems, prob— lems related essentially to the development of urban and metropolitan America. The evolution of American federalism during the past fifty years has been one of expansion in the federal role, its major domestic manifestation being an increasingly strong tie between federal and local governmental agencies. Some View the expanded role of the fed- eral government as the inevitable consequence of an industrialized, technological society.3 Others see it as the development of positiVe and constructive interaction and cooperation among different levels of government.4 Still others explain it in terms of accumulated federal functions and responsibilities that the states have been unwilling to assume themselves.5 All of these explanations are in part correct, but the evidence on the latter is the most persuasive; the expanded l “.1900de °f the weal «may. These reOO‘ line. are traceable to the 1 weir citizens. Often id intakes for state govern his!!! man that the ideolO Inhelnlden is not apmerj which to judge 910905“ PM are it is appropriate for Mistrial society rather U society.6 ignition is shared and has minus book, m : three overriding defi Eulmi and the mythology v 4151: 1s in orientation--m< acMince with the rural 1 st“second is in timeliness es are anachronistic; ‘ Etudes of the modern wo [Tine leaders are by Ion-bound, which ill oi“idem Slovemment . . domestic role of the federal government is primarily the result of the failure of the states to respond to changing demographic patterns and political needs. State problems have been characterized by Press and Adrian as stemming mainly from a lack of timeliness: . a good many of the weaknesses of state government are unnecessary. These recognized shortcomings, we be— lieve, are traceable to the failure of state governments to reflect the modern viewpoints held by a large majority of their citizens. Often ideas dominant among the deci- sion makers for state government lack timeliness. By this we mean that the ideology to which decision makers are beholden is not appropriate as a yardstick against which to judge proposed public policies for today be— cause it is appropriate for a rural, small town, pre— industrial society rather than our contemporary urban society.6 This position is shared and has been elaborated upon at length by Roscoe Martin in his book, The Cities and the Federal System. He notes, . . three overriding deficiencies flow from the state of mind and the mythology which grip the states. The first is in orientation—~most states are governed in ‘ accordance with the rural traditions of an earlier day. \ The second is in timeliness——the governments of most ‘ states are anachronistic; they lack relevance to the urgencies of the modern world. The third is in leader— ship—-state leaders are by confession cautious and tradition—bound, which ill equips them for the tasks 0f modern government . . .7 A major factor perpetuating this lack of timeliness has been the rural domination of state legislatures through policies of malap— POrtionment. Until 1962, when the Supreme Court ruled malapportionment unconstitutional, states had ignored changes in population to the point Where "the average value of the vote in the big city was less than half the average vote in the open country, so far as electing members of the . ,8 State legislature [was] concerned.’ . I' in state 1egislat an states reapportimt 1 am has shifted from non holofthe legislature give indirecting state resource! as. To the extent that the the punters in the federal The purpose of the pr! El'iortionment on the distri‘ thPOthesis to be test em between 1962 and 1s Metre of representation Hi“ the literature on real it. In Chapter Two, these A prec0ndition to any re-establishment by the states of their constitutional role in the federal system is a greater degree of time— liness and responsiveness in their legislative policy making. In most states, this means greater attention to the problems and needs of urban and metropolitan areas. The reapportionment decisions have pro- vided the basis for such a shift by essentially eliminating rural over- representation in state legislatures in the course of only five years. In many states reapportionment has meant that the balance of legisla— tive power has shifted from non—metropolitan to metropolitan areas. Control of the legislature gives metropolitan legislators the potential for redirecting state resources that have traditionally favored rural areas. To the extent that they do so, states are likely to become more active partners in the federal system than they have been to date. The purpose of the present study is to examine the impact of reaPPOrtionment on the distribution of state aid to education. The central hypothesis to be tested is that changes in the distribution of state aid between 1962 and 1969 are directly related to changes in the structure of representation in state legislatures. The study will first review the literature on malapportionment and state policy in Chapter One. In Chapter Two, these findings will be placed in a theoretical context and an alternative model presented for examining the relation— Ship between apportionment systems and state policy. Chapters Three thrOugh Six Will present the findings of the data analysis, and Chapter Seven will relate these findings to the hypotheses, draw appropriate conclusions, and make suggestions for future research. line. for M19! the a‘ 1.1969). em 99- 224‘239' 1 5mm units of sow—me] tnofbemcracy.“ MEL “Manon. 953-970- 2lreeent analysis of 1 'nsinstate goverrment befo: intents Sharing: Crutcl filled York: Praeger, 197 ismus grants from the fed itl'urlleller, New Dimensions 361 . 3 Two arguments of thi: ?: B. E. Schattschneider, ‘ Jim and Wmston, 1960) . mum of Power in America 4 “I The famous metaphor :Mf State and Federal A f7 the best example of thi éhmmmental Relations“ . “theme and Douglas St lsl qs in American Federal ‘5 Inn. 1969) . 5 3 The CA major proponent ities and the Bede INTRODUCT ION Notes lSee, for example, the arguments presented by Peter Marris and Martin Rein in their book, Dilemmas of Social Reform (New York: Ather- ton, 1969) , esp. pp. 224—239. Robert Dahl has dealt with the question of governable units of government in his article, “The City in the Future of Democracy," American Political Science Review, Vol. LXI, No. 4 (December 1967) , 953—970. 2A recent analysis of this issue that calls for significant re- forms in state government before revenue sharing is enacted by Henry S. Reuss, Revenue Sharing: Crutch or Catalyst for State and Local Govern— ment? (New York: Praeger, 1970) . The alternative View, calling for no strings grants from the federal government to states is found in Walter Heller, New Dimensions of Political Economy (New York: Norton, 1966) . 3Two arguments of this type from very different perspectives are: E. E. Schattschneider, The Semi—Sovereign People (New York: Holt, Rinehart and Winston, 1960) , Chapter 5, and Talcott Parsons, "The Dis- tribution of Power in American Society," World Politics (October 1957). 4 . The famous metaphor of the "marble—cake" of inter—related func— tions of State and Federal Agencies developed by Daniel Elazar is prob— ably the best example of this interpretation' See his "Federalism and Intergovernmental Relations" in Daniel Elazar, R. Bruce Carroll, E. Lester Levine and Douglas St. Angelo (eds.) / W W (Itasca' “limis‘ F‘ E‘ PeaCOCk Nb" lishers Inc., 1969) . 5A major proponent of this position is Roscoe C. Martin. See his The Cities and the Federal System (New York: Atherton, 1965) . 6Charles Press and Charles Adrian, "Why out State Governments are Sick," in Charles Press and Charles Adrian (eds.) , Democracy in the Fifty States (Chicago: Rand McNally, 1966), p. 347. 7Martin, op. cit., p. 79. 8Paul T. David and Ralph Eisenberg, Devaluation of the Urban and Suburban Vote, Volume 1 (Charlottesville, Va.: University of Virginia Bureau of Public Administration, 1961), p. l. O .mmm an influence of “1391’" an federalism not-.wi'chstam ruins unclear. The sta1 stat it sorks to the disadva as. Such an argument carrii first is that the meth011 "Lilith values and interests 3W is that if metropo] in in state legislatures, ashihution of state resourc migrate: share of state inProblems that metropolit: "shelves. These assumptions ax fitment on state pouci. “his 0! research. The ii i x 2 i \ CHAPTER I A REVIEW AND CRITIQUE OF THE LITERATURE The influence of malapportionment on the overall character of American federalism notwithstanding, its impact on specific state pol—- icies remains unclear. The standard argument against malapportionment is that it works to the disadvantage of central city and suburban areas. Such an argument carries with it several implicit assumptions. The first is that the metropolis is somehow a cohesive, distinct en— tity with values and interests in opposition to the rest of the state. The second is that if metropolitan areas were given greater represen— tation in state legislatures, they would act as a bloc to shift the distribution of state resources more in their favor. The third is that a greater share of state resources for urban areas would alleviate the problems that metropolitan areas are unable to effectively solve themselves. These assumptions and others relating to the impact of reap— Portionment on state policies have been challenged by two separate bodies or research. The first deals with legislative behavior, par- ticularly roll call voting; the second, with the correlates of expen— diture and income redistribution policies. The findings of both sets Will first be summarized and then critiqued. .m in Wing relevant 1. “ants in terms of whes intropolitan and non-IIIem an unified voting blocs - 'aislahms, David Derge conc .. . the traditional beli wimplitan and non-metrc legislature must be rejec1 at least at the roll-call support such a conclusion 1. Non-metropolitan together with hig politan legislatc 2- Metropolitan legi together with hi1 3- Metropolitan leg wailing side whe high (:Ohesion.l “Similar study of the or Wider“? both the H01 the.validity 0f the met] “Wide-hie that urban—rm although “man-rural CO Wt °°Casions. nth tha ’ Legislative Behavior A prerequisite to metropolitan effectiveness in state legisla- UHesis a degree of unity and cooperation among metropolitan legis- lators in supporting relevant legislation. Several researchers have neasured this in terms of cohesion on roll call votes and have found 'flmt metropolitan and non-metropolitan legislators seldom oppose each oflmr in unified voting blocs. In a study of the Illinois and Missouri legislatures, David Derge concluded, . . the traditional belief in bitter c0nflict between metropolitan and non-metropolitan areas in the state legislature must be rejected for Illinois and Missouri, at least at the roll-call stage. The following findings support such a conclusion: 1. Non—metropolitan legislators seldom vote together with high cohesion against metro- politan legislators. 2. Metropolitan legislators usually do not vote together with high cohesion. i 3. Metropolitan legislators are usually on the pre— ‘ vailing side when they do vote together with high cohesion.l haa similar study of the Ohio legislature, Flinn concluded: Considering both the House and the Senate and assuming the validity of the method used, the conclusion is un— avoidable that urban—rural factionalism does not exist although urban-rural conflict may occur on very infre— quent occasions. , . . This conclusion is consistent with that reached in a recent study of Illinois and Missouri. It may be inferred from the similarity of these results that urban—rural factionalism is unimpor— tant in the operations of two—party legislatures although the conclusion must be tentative pending further study.2 These findings are generally consistent with those of Hamilton and others in Indiana, Friedman in Tennessee, and Becker and others inMichigan.3 All of these studies, but particularly the latter, m...‘ n. comes “dime of 189151ative a film“ constituency pr ‘ fining their behavior.6 '1'. :‘iqispolitical parth alt“ namiety of factors and a In srm, this research : ‘ Ishtar and urban legislator Fehgether as a bloc on mar iWise in state legislatuJ hues-within metropol ita: According to this arg asShrevent legislators frc ~15 the heterogeneous, fra‘ .. *urfl “.Ww__wr ._ suggest that whatever cohesion exists among metropolitan legislators is related more to party than to constituency factors. Other research on the factors associated with legislative voting suggests that the urban-rural issue is dominant in only a few, rural, predominantly southern states,4 that constituency characteristics are relatively poor predictors of legislative voting,5 and that state legislators gen— erally perceive constituency pressures as being relatively minor factors influencing their behavior.6 The single best predictor of legislative voting is political party, although even its influence is contingent upon a variety of factors and appears to be diminishing. In sum, this research suggests, with Jewell, that, ". . . Met— ropolitan and urban legislators cannot realistically be expected to vote together as a bloc on many issues. The most controversial issues that arise in state legislatures usually cause divisions—-often very deep ones——within metropolitan areas."8 According to this argument, if divisions within metropolitan areas prevent legislators from voting cohesively on most issues, then it is the heterogeneous, fragmented nature of the metropolis itself rather than metropolitan underrepresentation in the state legislature that accounts for whatever non—metropolitan "biases" may exist in state pol icy . W Income Redistribution A second body of research challenging the influence of malap- portionment on state policy is a series of studies using input—output Thomasbj ...systel characteristic: mt effect on colic: Mic developnent shapes pliq nutmeg, and most 0: Ms between system charat M can be attributed to momentm L‘Mclusion on the influence millewhole, the policy ch legislatures are not notice whey choices of well-appc ht of the policy differer fitto be a product of socz' Mme states rather the “Nihilist practices .11 he semnd stage of th models, aggregate data, and correlation-regression statistical tech— niques. These studies have evolved through three fairly distinct stages. The first stage examined the relative impact of “economic deveIOpment" and "political system" variables on expenditure levels in the fifty states.9 Thomas Dye's conclusions are representative: . system characteristics have relatively little independent effect on policy Outcomes in the states. Economic development shapes both political systems and policy outcomes, and most of the association that occurs between system characteristics and policy out— comes can be attributed to the influence of economic development . 10 His conclusion on the influence of malapportionment is similar: On the whole, the policy choices of malapportioned legislatures are not noticeably different from the policy choices of well—apportioned legislatures. Most of ~the policy differences which do occur turn out to be a product of socio—economic differences among the states rather than a direct product of apportionment practices . 11 The second stage of this research consisted of increasing the number of variables initially considered, factor analyzing them to iso— Using this technique, Sharkansky and Hofferbert conclude that: While these findings add to the inquiry into polit- ical and economic determinants of public policies, they offer little encouragement to those [who] would seek to expand the level and scope of public serv- ices by manipulating one political or structural characteristic of state government (g;g,, voter turnout, party competition, or apportionment). It is apparent only that certain aspects of politics haVing to do with voter turnout and interparty com- Petition are related to certain public policies.12 \ late the underlying dimensions, and then correlating the factor scores. i In the third stage of this research, the dependent variable, PUblic policy, has been redefined in terms other than expenditure levels. lithium! despite a 5. suinlitical variables and ‘l Ehibself is found to be W use three sets of fin ssomlement the roll call 'Eshucture of representatim isofmalapportionment has Ethiopt. The techniques i’iiations themselves, howew 3‘1“! any conclusions regar fifigge of Roll-Call and « lemon-Regression Studi1 The absence of an ur 10 In their study of the determinants of variation in the net redistribu- tive impact of revenues and expenditures, Fry and Winters conclude: Not only do the political variables have an independent impact on redistributive policies in the states; they also account for considerably more of the variance in redistribution than do socioeconomic variables.l3 Once again, however, despite a significant relationship between the set of political variables and redistributive policies, malapportion— ment by itself is found to be unimportant. These three sets of findings using aggregate data for all fifty states complement the roll call studies noted above in suggesting that the structure of representation in state legislatures related to pol— icies of malapportionment has a minimal impact on the types of policies states adopt. The techniques used in these studies are not without limitations themselves, however, and deserve closer examination before drawing any conclusions regarding malapportionment and policy. A Critique of Roll—Call and Correlation—Regression Studies The absence of an urban—rural dimension in roll call voting does not also mean that these forces are not present in the legislature. As Jewell has noted: Although urban and rural legislators do not instinctively man opposing barricades when the roll is called, their attitudes have been different in most legislatures on many issues of importance. Voting records probably min— imize these differences because the outnumbered urban legislators have compromised in order to win concessions from the rural majority . . .14 In his study of urban’r“! on techniques of merge and he bloc-identification methr that each substantive issue : more points in the legisl. votes. . . [Ilt can be said ut safe in Kansas. To be s passage is required by the F mte is rarely the decisive usures, even many of the < inqreen light in the com: record vote and the final v hyabase constitutional ma udnot voting . . .15 In their study of educe tiers and his colleagues com ...it seems fair to say ally never unite to oppos litter, there is a kind of inference which must frec horde: to enact urban 16 Several other case st ~ in legislatures have sh as are more significant ti in L L sun types of cleavages ‘ The attitudes and no Wile . mum of the urban-n “he he . k of "timeliness“ ll In his study of urban-rural conflict in Kansas, Page concludes that the techniques of Derge and others are not applicable in his state: The bloc—identification method depends on the assumption that each substantive issue is tested decisively at one or more points in the legislative process by recorded votes . . . [I]t can be said that such an assumption is not safe in Kansas. To be sure, a record vote on final passage is required by the Kansas constitution, but this vote is rarely the decisive test of the process. Many measures, even many of the controversial measures, get the green light in the committee of the whole without a record vote and the final vote is perfunctory approval by a base constitutional majority . . . with many absent and not voting . . .15 In their study of educational politics in the Midwest, Nicholas Masters and his colleagues concluded: . . . it seems fair to say that rural interests practi— cally never unite to oppose united urban interests. Rather, there is a kind of presumptive urban—rural difference which must frequently be negotiated away in order to enact urban legislation.16 Several other case studies of the nature of urban—rural differ— ences in legislatures have shown that the processes preceding roll calls are more significant than the votes themselves in revealing the dominant types of cleavages that influence legislative behavior. 17 The attitudes and norms of the legislature are another aspect or dimension of the urban—rural conflict. One manifestation of this is the lack of "timeliness" in the thinking of many rural legislators noted earlier. In his Kansas study, Page refers to it as "symbolic localism"; - . . the formal representation of rural areas in the l<“‘-91islature have developed a symbolic localism, a type 0f institutionalized identification of their consti— tuents' interests with the tax interests of rural local government . . .18 . ofu & 3 , {rib flanges .-" u: in tile 1011 can mil nuns findings Dame's i'mteeted' roll calls for l tile entire crow 0f ““5' intantvariation in voting 1 than distinguish the nat‘ its. Some types of votes It bticularly on a vote such a: mount support for passac In a malapportioned 1 12 Independent of the factors just mentioned that raise some questions as to the utility of using roll call votes as the basis for measuring urban—rural cleavages in a state, there are also some limi— tations in the roll call technique itself that cast doubt on the val- idity of its findings. Derge's study, the most often cited, aggregated all "contested" roll calls for a given period and examined the patterns for the entire group of votes. By doing this, it ignored potentially important variation in voting alignments across issues, as well as failing to distinguish the nature of the contested XE- uncontested votes. Some types of votes reveal valid coalitions better than others. Particularly on a vote such as an educational appropriations bill, there are so many sub-appropriations and riders" subsumed under one heading that the final vote is likely to be merely a formal endorsement 0f a series of bargains and trades that have gone on beforehand to gain sufficient support for passage. In a malapportioned legislature, any majority will necessarily include a disproportionate number of rural legislators. The costs of attaining that majority are likely to be a distribution of aid favorable to rural districts. Even if only a minimum number of rural legislators are included in the majority, the distribution of aid will tend to advantage their districts and others they choose to benefit because of thair pivotal position in creating that majority. The correlation—regression studies of policy outputs, outcomes, and impacts in the fifty states is a relatively new and very different :iuqsof these studies at diff one that the findings are idiots, measures. and ‘18ij than variables can be SW" artisan socio-economic variab shone redistribution, and a bias per capita expendituri and variables is greater separate variables, are exa flatly shapes their findings ‘ than from this research a- irllfferent governmental se ifimnonic development, whil 13 income redistribution p03 Eiicsof the political syst iii mre money, independer 3W: is significant and < ESnotions about the centr an . ‘Di' b“ It is only an i :5 faCi. . 0 rs influencing st; “my e . Starch wall he give- I n sun, although t bless Ion studies challer Etdin ' ytbe influence of Ilmus‘ Ions b ased 0 n this 13 type of analysis from the roll call studies. The variation in the findings of these studies at different stages in their development indicates that the findings are at least in part artifacts of the techniques, measures, and definitions that are used. The fact that political variables can be shown to have a greater independent influ— ence than socio-economic variables on policy when policy is defined as income redistribution, and a lesser influence when policy is de— fined as per capita expenditures, and that the relative impact of political variables is greater when presented as factor scores than as separate variables, are examples of how the design of such studies clearly shapes their findings. The only conclusion that can validly be drawn from this research at present is that state expenditure levels for different governmental services are primarily related to the level of economic development, while the distribution of its resources and its income redistribution policies are primarily related to character— istics of the political system. To find that states having more money Spend more money, independent of the characteristics of the political SYstem, is significant and challenges a number of traditionally held assumptions about the centrality of the political system in policy making, but it is only an initial step toward a precise explanation of the factors influencing state policy. A more detailed criticism of this research will be given in the next chapter. In sum, although the findings of roll call and correlation— regression studies challenge some of the traditional assumptions re— garding the influence of apportionment systems on state policies, any conclusions based on this research are likely to be premature. intact melv t0 be “1‘ I. II“In of plicy has been “filaments in local finite: lbst of the analYSe‘ of several dimensions [0f 1 venturing an awareness t involved. Most frequently level of expenditure. Pro? list. In addition, we car iinansion: the distributf Lion. . . Thus to under: level one must understand hated to school districts in unfortunately are ra (in interesting political °5 data cannot deter poli liiing what may be the m: h’qnts.19 in a more recent art «I . .--.1c policy, Robert Salish l4 TheConceptualization of Policy A major limitation of the studies cited above is their failure tospecify the dimensions of legislative structure, process, or policy amt are most likely to be related. Concern over the lack of concep— Umlization of policy has been the strongest. In an important article rmflewing developments in local and state politics research, Jacob and Lipsky note: .. . Most of the analyses we have cited use measures of several dimensions [of policy] indiscriminately with— out showing an awareness that more than one dimension is involved. Most frequently used are measures of the level of expenditure, program quality, and program im— pact. In addition, we can identify at least one other dimension: the distribution of benefits among a popula— tion . . . Thus to understand the politics at the state level one must understand how grants—in~aid are distri— buted to school districts . . . Measures of distribu— tion unfortunately are rarely available in public records (an interesting political fact in itself). But the lack Of data cannot deter political scientists from investi— gating what may be the most important dimension of policy Outputs.19 In a more recent article also dealing with the dimensions of Pmflic policy, Robert Salisbury builds on the policy typology of flmodore Lowi and distinguishes among "distributive," "redistributive," "regulatory," and "self—regulatory" policies.20 He defines "distri— butive" policy as follows: .. . those perceived to confer direct benefits upon one or more groups. Typically such policies are determined with little or no conflict over the passage of the legislation, but only over the size and specific distri— bution of the shares. Rafistributive policy is characterized as me indication of the? udstatepolicv is clear: Emmmtation in til am in policies involvin‘. 'zin the other types cited n It follows from the ab Ely related to rural favori ‘Jlilited evidence on the 51 illitems of apportionment, Etalkelations concluded: helm and 1962 reports renal that its observers Present apportionment of State grants-in-aid or ti ”Went, and labor an ntvons are in accord wi itrons cements made to PNParation of this repc $3Sandlldrian come to a a cause legislation and es, constitutional with its approval, 15 .. . also confer[ring] benefits, but also [is] perceived to take benefits away from other groups. They therefore involve more intense conflict over passage itself, over the legitimacy of the action as well as the specific content . . .21 The implication of these typologies for research on apportion— nmnt and state policy is clear: since malapportionment is essentially nual over-representation in the legislature, its impact should be semvnwre in policies involving the distribution of state resources thmiin the other types cited above. It follows from the above that malapportionment shOuld be di— redfly related to rural favoritism in the distribution of state aid. Thelimited evidence on the subject is inconsistent. In a 1962 study of patterns of apportionment, the Advisory Commission on Intergovern— nental Relations concluded: The 1960 and 1962 reports of the National Municipal League I reveal that its observers find the greatest effect of l present apportionment of state legislatures involves ‘ state grants—in-aid or the allocation of funds to local government, and labor and welfare matters. These obser— vations are in accord with other studies as well as nu— merous comments made to the Commission staff during the preparation of this report.22 Press and Adrian come to a similar conclusion: Because legislation and, in more than two—thirds of the states, constitutional amendments as well can be proposed only with its approval, the small—town bloc will often levy a special price when it agrees to act. This is the most obvious result of its control. A study in Connec— ticut, for example, described a state aid formula con— structed so that towns with less than 500 population re- ceived $27.19 per student while cities of over 100,000 received $4.95 per student. In Colorado, Denver schools with an enrollment of 90,000 received $2,300,000 under the state—aid formula while the schools of nearby Jeffer- son County with 7,200 fewer children enrolled received $100,000 more in state aid. The same pattern is 3m ' .33, Jewell c‘ wherein discemjhl Wt effect of halal)?d when the mad: °n Sta” hm used for diSd ”1%,th: and Othel. m this proportionate W light to population.24 0n the other hand: in :thse described above. Brad .. . all the heavily P091 less than their fair sharl situation is just as com as in poorly apportioned to argue that malapportio We share of revenue t receive.25 One explanation for l 16 frequently repeated in state aid for local roads, wel— fare grants, police protection, library facilities, and almost every other purpose.23 More recently, Jewell concluded: The widespread, discernible, and probably the most important effect of malapportionment on legislation has been the impact on state aid to local government. The formulas used for distributing state funds for roads, education, and other purposes have frequently given this proportionate weight to area and little weight to population.24 On the other hand, in a correlation—regression study similar to those described above, Brady and Edmonds found that: . all the heavily populated counties are getting less than their fair share of state revenue. But this situation is just as common in well apportioned states as in poorly apportioned ones. . . . [W]e are inclined to argue that malapportionment has little or no effect on the share of revenue that counties of different size receive.25 One explanation for the inconsistency between the findings of the studies in specific states and those using aggregate data for all fifty states is the relative "crudeness" of the summary measures uSed in the latter. The appropriateness of using a summary measure of mal— apportionment based essentially on variation in the size of legislative districts to examine its impact on policy is open to question, as is a measure of state aid distribution which distinguishes simply between C0unties with over 250,000 population and those with less. Brady and Edmonds' conclusions should therefore be treated with the same Sk6pti_ CiSm given other correlation—regression studies. new and P011 Chan“ the reapportiOnment dec: quYOfDOII'iIIVO1V8ment in 5 ”we” in what Justice Fra that“ of deciding what const stueutation. Although the < :05 variation in the size 0 in questions such as the s than composition, how ofte anionthen, and other issx Tiélpon. The background, mt WW the decisions are Site these complexities, a trashed OppositiOn to reap hells were enacted with su: kidnap-e u ' - - - [Bletwee will significant malappc states had virtually di The full impact of 3 :5 liars to cene A - - 5 With l7 Reapportionment and Policy Change The reapportionment decisions of the Supreme Court reversed a history of non—involvement in state apportionment systems and placed it squarely in what Justice Frankfurter referred to as "the political thicket" of deciding what constitutes equitable systems of legislative representation. Although the decisions dealt directly with the prob— lem of variation in the size of legislative districts, they left unre— solved questions such as the size of districts, their demographic and partisan composition, how often they should be changed, who should re— apportion them, and other issues that the courts must now ultimately rule upon. The background, merits, implications and expected conse— quences of the decisions are subjects of an extensive literature.26 Despite these complexities, and despite extensive and politically well— entrenched opposition to reapportionment in many states, the Court rulings were enacted with surprising ease, permitting Frederickson and Cho to note, ". . . [B]etween the Supreme Court's decision in 1962 and 1968 all significant malapportionment in the legislatures of the Amer—- ican states had virtually disappeared.”27 The full impact of the decisions will not be known for a number Of years to come. As with the assessment of malapportionment, "experts" differ as to the likely consequences of reapportionment on state pOlicy. Jewell and Patterson foresee significant changes in taxing and spending Policies: The most direct effect [of reapportionment on policy] is likely to be on taxing and spending policies. Changes are likely in both the types of taxes levied and in the authority given to cities to levy taxes. More important soasto overcome the gross of sparsely pepulated connti zen Jacob, 0n the other ham evasion researchers in his c It is improbable that it [1 stantially envigorate state thestalemates which sap pl So few studies of the simulated that it is diff outsmre correct. In On: the Georgia Legislature, I E‘Iooolita l ' ‘ n egislators incr “sluncing policy more in t1 who study of the same le .. .The Georgia study opoortionment do affect ls detectable when intr' out are studied. TTef Systems may n 0 m . t be dete an envi much in the da varlublromnental fact: hell h e .111 Georgia_ lib: t in more urban rel. and Who are l8 and more certain changes can be anticipated in the formulas adopted for the distribution of state aid, so as to overcome the gross discrimination in favor of sparsely populated counties in some states.28 Hamert Jacob, on the other hand, is typical of the correlation- regression researchers in his conclusion that: It is improbable that it [reapportionment] will sub— stantially envigorate state governments or dissolve the stalemates which sap public confidence in them.29 So few studies of the actual impact of reapportionment have bemncompleted that it is difficult to know which of the two predic— trxm is more correct. In one study of changes in voting behavior nlthe Georgia Legislature, Ira Sharkansky found that cohesion among metropolitan legislators increased as they gained the potential for hfiluencing policy more in their favor through reapportionment.3O In anther study of the same legislature, Brett Hawkins concluded . The Georgia study suggests that variations in apportionment do affect policy, and that this influence is detectable when intrastate variations in apportion— ment are studied. The true impact of apportionment systems may not be detectable when only interstate variations in appOrtionment are studied . . . [W]e cannot confidently conclude that reapportionment is the direct or majOr cause of the increased success rate of municipal association measures of increased urban successes in urban—relevant policy areas. Some of these changes began before reapportionment; and all seem partially the result of such other variables as urbanization and the growing saliency of urban needs. Indeed, much in the data Suggests that urbanization, an environmental factor, is an important explanatory variable in Georgia. But because reapportionment has brought in more urban representatives, who are more liberal, and who are voting together more often, and winning more often, we conclude that to some unknown degree reapportionment has been a factor in observed Policy changes since reapportionment . . . Changes observable now, in addition, suggest that reapportion— mmtwfllbeanimmmmm:mdmrinfimmepmiw ChOices.31 hHasifnifimt mac: d mm [is] eased“j h resulted in a distinct state aid formulae againSl luly in duration. It 9 a, argued that reapPOIti inthe way a state aPPOrt My were particularlY cc state fiscal discriminati mold decline with reaPP‘ Although provocative ml] is no more conceptual] hpmdecessors. Even thov Missing the question . it! rather than making i mapportionment, the li stout; large numbers of 1 ielevation as to their in ills between various inde 19 The only study of reapportionment using correlation—regression techniques and aggregate data from all the states has been conducted by Frederickson and Cho. Their findings suggest that reapportionment has had a significant impact on policy: The evidence [is] especially strong that reapportionment has resulted in a distinct lessening in the disparity in state aid formulae against metropolitan areas——particu— larly in education. It seems clear, then, that those who argued that reapportionment would make a difference in the way a state apportions its funds were correct. They were particularly correct in their prediction that state fiscal discrimination against urban interests would decline with reapportionment.32 Although provocative in its findings, the Frederickson and Cho study is no more conceptually precise or theoretically informed than its predecessors. Even though it improves upon the earlier research by addressing the question of reapportionment from a perspective of change rather than making inferences based on cross—sectional measures of malapportionment, the limitations of the earlier studies are still Present; large numbers of independent variables are still used without explanation as to their inclusion in the model; the basic relation— ships between various independent and dependent variables are never Presented, the only statistics being coefficients of multiple determi— nation (R2) and significance levels for different clusters of indepen— dent variables; and the measures of apportionment are, again, indirect rather than direct. mm to swift the 9°? difference. we recent res? W ihfindings that apportim titles in accounting f0! I 'mistency is at least pal h1iuprecision of the latb iapportionment and the dim ltd, the nature of that re lmhxt of actual change, Hessian techniques can 0 Reef the issues existing 20 Conclusion The evidence on the impact of apportionment on state policy remains both inconsistent and inconclusive, despite extensive research. While the "conventional wisdom," based primarily on case study find- ings, tends to support the position that apportionment systems do make a difference, more recent research has challenged this interpretation with findings that apportionment is much less important than other variables in accounting for policies in all the American states. The inconsistency is at least partially due to the conceptual and theoret— ical imprecision of the latter studies. By specifying the dimensions of apportionment and the dimensions of policy most likely to be re— lated, the nature of that relationship, and then testing the model in a context of actual change, research using aggregate data and correlation— regression techniques can combine the best of both approaches and address some of the issues existing studies have left unresolved. 1manage. “Hemp sum Legislative Dela BE. n1. LIII (December I 21mins A. Flinn! "“1 Line,‘ in Charles Press an! 3110mm D. Hamilton. 'Etin Reapportionment in In Rebate Law Review, Vol. 2fi."illet1rban-Rural Confli MN (June 1961): R. W 33H “Correlates of Legi: eeutives,“ Midwest J our $1496 4Wayne Francis, Let ative Anal sis (Chic: 5 Hugh L. LeBlanc, “Midwest J 6John C. Wahlke, j .iegisiative Behavior ( CHAPTER ONE Notes lDavid Derge, "Metropolitan and Outstate Alignments in Illinois and Missouri Legislative Delegations," American Political Science Imview, Vol. LIII (December 1958), 1065. Thomas A. Flinn, "Urban—Rural FactiOnalism in the Ohio Legis— lature," in Charles Press and Charles A. Adrian, op. cit., p. 366. 3Howard D. Hamilton, J. E. Beardsley, and C. C. Coats, "Legis— lative Reapportionment in Indiana: Some Observations and a Suggestion," Notre Dame Law Review, Vol. XXXV (May 1960), 368—404; Robert S. Fried— mnh "The Urban-Rural Conflict Revisited," Western Political Quarterly, Vol.XIV (June 1961): R. W. Becker, F. E. Foote, M. Lubega, and S. V. Mnmma, "Correlates of Legislative Voting: The Michigan House of Rep— resentatives," Midwest Journal of Political Science, Vol. VI (1962), 384-396. 4 . . . . Wayne Francis, Legislative ISSues in the Fifty States: A Comparative Analysis (Chicago: Rand McNally, 1967). 5 . Hugh L. LeBlanc, "Voting in State Senates: Party and Consti- tuency Factors," Midwest Journal of Political Science, Vol. XIII (February 1969). 6 John C. Wahlke, et_al,, The Legislative System: Explorations niLegislative Behavior (Glencoe, 111.: Free Press, 1962). 7 , . Julius Turner, Party and Constituency: Pressures on Congress (Baltimore: The John Hopkins Press, 1971. Revised Edition by Edward v_ Schneier, Jr.). 8Malcolm Jewell, The State Legislature: Politics and Practice _______.___.___._a.a__.___~_s.~____..*~_~____ (New York: Random House, l969), P- 21- 21 11 ' "ualapport mot Politics. Vol- xxv 12In snarkansky and ‘ httolitics, Economics and 39mm, v01. LX111 (Sepi 3 . . anan R. F1? and R1 thtion,“ American Politic 370]. 521. 14 talcum Jewell . “T1 ‘31-.) State Legislatures '1 my: Prentice-Hall, 196 15Thomas Page, Legi___s 2‘51: BUreau of Governuu 16 it Nicholas H. Masts . H’01itics. and the Put flWP 28. 22 9The two initial studies of this type were Richard E. Dawson andJames A. Robinson, "Inter—Party Competition, Economic Variables am Welfare Policies in the American States," Journal of Politics, Vol.XXIII, No. 2 (May 1963), 265—289; and Herbert Jacob, "The Conse— mmnces of Malapportionment: A Note of Caution," Secial Forces, Vol. XLIII (December 1964), 256-261. 10Thomas R. Dye, Politics, Economics, and the Public: Policy Outcomes in the American States (Chicago: Rand McNally, 1966). 11 . . . . , "Malapportionment and Public Policy in the States," Journal of Politics, Vol. XXVII (February 1965), 586. l . . 2Ira Sharkansky and Richard I. Hofferbert, "DimenSions of State Politics, Economics and Public Policy," American Political Sci— ence Review, Vol. LXIII (September 1969), 878. 13 . . . . Q Brian R. Fry and Richard F. Winters, "The Politics of Redis— tribution," American Political Science Review, Vol. LXII, No. 2 (June 1970), 521. 14 . . . . Malcolm Jewell, "The Political Setting," in Alexander Heard (edJ, State Legislatures in American Politics (Englewood Cliffs, New Jersey: Prentice—Hall, 1966), p. 72. 15 . . . . Thomas Page, Legislative Apportionment in Kansas (Lawrence, Kansas: Bureau of Government Research, University of Kansas, 1952), p. 149. l6 . . Nicholas H. Masters, Robert H. Salisbury, Thomas H. Eliot, State Politics and the Public Schools (New York: Alfred A. Knon, 1964) , p. 28. 17 See in particular Richard T. Frost, "On Derge's Metropolitan and Outstate Legislative Delegations," American Political Science Eflflgyj and William C. Havard and Loren P. Beth, The Politics of Mis- RePresentation: Rural—Urban Conflict in the Florida Legislature (Baton Rouge: Louisiana State University Press, 1962). 18 Page, op. cit., p. 104. 19 . Herbert Jacob and Michael Lipsky, "Outputs, Structure, and Power: An Assessment of Changes in the Study of State and Local Poli— tics," Journal of Politics, Vol. xxx (May 1968), 515. comission G W “"4 am; and Adrian: 92; 24Jeuell,196€n ere—Ci 25vivid Brady and Doug inaction, Vol. IV (March 1 26A major proponent 0 hidpai League. Its argum tenement can be found 1 lltgislative Apportionment “553); Reamrtionment: A hateliational Municipal L1 Asitionwas taken by the C1 Mi its positions on th aJyLeqislative ApErtio @1964). A more scholar “MG. Dixon, Jr., Demo< 3i and Politics (New York 27 4 3. George Frederi Want and Public Policy Melive 1315 ‘ 23 20Robert H. Salisbury, "The Analysis of Public Policy: A Search for Theories and Roles," in Austin Ranney (ed.) , Political Science and Public Policy (Chicago: Markham, 1968), p. 158. 2lIbid. 22 . . . . . AdVisory ComIniSSion on Intergovernmental Relations, Apportion— ment of State Legislatures (Washington, D.C.: ACIR, 1962), p. 28. 2 . 3Press and Adrian, op. c1t., p. 365. 24Jewell, 1966, op. cit., p. 73. 25David Brady and Douglas Edmonds, "One Man, One Vote——So What?," Transaction, Vol. IV (March 1967) , 43. 26A major proponent of reapportionment has been the National Municipal League. Its arguments and documentation of the case for re— apportionment can be found in the following publications: Compendium on Legislative Apportionment (1962); Reapportionment: A Year in Review (1963); Reapportionment: A Second Year in Review (1964); all published by the National Municipal League, New York. The anti-reapportionment position was taken by the Council of State Governments. Excerpts of some of its positions on the issue can be found in Howard D. Hamilton (ed.) , Legislative Apportionment: Key to Power (New York: Harper and Row, 1964). A more scholarly, comprehensive treatment of the issue is Robert G. Dixon, Jr., Democratic Representation: Reapportionment in Law and Politics (New York: Oxford University Press, 1968) . 27 . . H. George Frederickson and Yong Hyo Cho, "Legislative Appor— tiOnment and Public Policy in the American States,“ a paper prepared for delivery at the sixty-Sixth Annual Meeting of the American Polit- ical Science Association, Biltmore Hotel, Los Angeles, California, September 8-12, 1970, p. l- 28 Malcolm E. Jewell and Samuel C. Patterson, The Legislative Process in the United States (New York: Random House, 1966) , P. 66. 29 . Herbert Jacob, "The Consequences of Malapportionment: A Note of Caution," Social Forces, Vol. XLIII (December 1964) , 256~261. 3O Ira Sharkansky, "Reapportionment and Roll Call Voting: The Case of the Georgia Legislature," WI VOL LI (June 1970), 129—137. at. 1971), p. 297. l . 3'Brett W. Hawkins , "C01 32Frederickson and Cho. hthasubsequent draft 0f huthe text. When asked whj eiithat a re-analysis of th mount on state aid to e finally been thOught. This I gression model that was use itiata for several of the \ Sedhthis research, all 51 inidbe treated cautiously 24 31 . . . . Brett W. Hawkins, "Consequences of Reapportionment in Georgia," in Richard I. Hofferbert and Ira Sharkansky (eds.) , State and Urban Politics: Readings in Comparative Public Policy (Boston: Little Brown, 1971), p. 297. 32Frederickson and Cho, op. cit., p. 44. It should be noted that in a subsequent draft of the same paper, this passage was deleted from the text. When asked why in a phone conversation, Frederickson said that a re—analysis of the data had shown that the impact of reap— portionment on state aid to education was not as strong as had ori— ginally been thought. This revision, plus the omnibus nature of the regression model that was used, plus some errors that were found in the data for several of the variables in their study that were also used in this research, all suggest that their initial conclusions should be treated cautiously. Heard! on State 901: 53 of perspectives and met inset in the last chaptfl :me theoretical contex‘ fining the influence of a imative Approaches to late Politics Research Research on state P iiierent from, that focusi Mimicking research is l itial~psychological or th CHAPTER II THE THEORETICAL CONTEXT Research on state policy and policy making includes a diverse range of perspectives and methods, many of which have been reviewed and discussed in the last chapter. This chapter will place these findings in a more theoretical context and present an alternative model for examining the influence of apportionment systems on state policy. Alternative Approaches to State Politics Research Research on state policy making is related to, yet also very different from, that focusing on the determinants of state policy. Policy-making research is behaviorally oriented, using either the social-psychological or the rational—calculus perspective to explain some aspect of the political, usually legislative, process.1 Studies using the social—psychological approach have dealt primarily with the role orientations and behavioral cues of legislators, but have not related such predispositions either to actual behavior or to policy outcomes.2 Those employing a form of rational—calculus to explain behavior have often used roll call votes as the dependent variable, examining it with such varied techniques as cluster analysis, axiomatic . . 3 theory, and regreSSion equations. 25 hill“! cases a varietY: itunins and cunpromises 1 niuelf a formality °f °“ minis llexplained,“ as in aims and assxmptions are ‘ Airflmjority party °°he stifled range of issues, 1' thiively narrow range of 3 Hitch predicting or inte melts. Third, roll call Ahlicy outcomes. Finai “it Variation in the dete 26 Roll call analysis, although more theoretically advanced than other approaches to legislative research, has several clear limita— tions. First, as was noted earlier, the significance of roll call (wting as an expression of individual policy preference is unclear, since in many cases a variety of sub-issues which serve as the basis for bargains and compromises are subsumed within a vote, making the vote itself a formality of only secondary importance. Second, where voting is "explained," as in the Meltz axiomatic model,4 the qualifi— cations and assumptions are so numerous that the phenomena being ex— plained. (majority party cohesion on contested roll calls, over an un- specified range of issues, in a limited number of states) represent a relatively narrow range of legislative behavior and are of questionable value in predicting or interpreting legislative behavior in specific contexts. Third, roll call analysis provides no linkage between voting and POlicy outcomes. Finally, it aggregates roll calls in such a way that variation in the determinants of voting across different types of issues are either completely obscured (as in the Meltz, Derge, and other roll call studies that lump together all "contested" roll calls in one or several legislatures Over a given period of time) or are distinguished only by general categories (as in the Clausen and Cheney article which found that voting on economic issues was related to party variables and voting on welfare issues was related to constituency variables.5 In sum, social—psychological studies contain insights as to the attitudes and role orientations of legislators, but relate these gully die mumship beta theoretical linkage bet“ pm my be established. 1 brie: mist suffice. Rese Iipolicy has as much inhel iii the policy-making proct Slips and the measures used Emblem with such resea “itemized the relatior fine 0i precision. 27 neither to structure nor policy, nor provide any theory. Roll call studies as a grOup~are the most theoretically advanced, but the phenomena they explain are conceptually undifferentiated and fail to specify the relationship between roll calls and policy. Ultimately, the theoretical linkage between legislative structure, process and policy may be established. Until that time, partial, incomplete theories must suffice. Research on the relationship between structure and policy has as much inherent validity at present as that dealing with the policy—making process, so long as the hypothesized relation— ships and the measures used to test them are theoretically informed. The problem with such research to date is that it generally has not conceptualized the relationship between structure and policy with any degree of precision . Legislative Structure and Public Policy The structure of state political systems in correlation- regression policy studies has for the most part been represented by Summary measures such as the frequency of turnover in party control of the houses of the legislature, variation in the size of legislative districts, degree of legislative "professionalization," level of legis— lative "conflict," and the formal powers of the governor. A number of these measures are then more or less arbitrarily brought together and subsumed under the heading "political system" and correlated with var- ious policy measures. Since they are not related to each other or to the policies they are being correlated with in any systematic fashion, rill. Imustially Signi uh]. me Schlesinger we“ qie of an filament val identical context for a I habitat and applied to . ism logical relationship. hitch for not controlling search that has indiscrim litres of state political manly associating large inniical basis for assu: The indiscriminate iterates the additional 28 however, the only basis for evaluating the findings is the significance of the correlations. Where no assumptions or hypotheses have been made at the outset about the nature of the relationships among the variables, statistically significant correlations have no theoretical meaning. The Schlesinger index of formal powers of the governor is an example of an independent variable initially developed and tested in a theoretical context for a specific purpose, and then taken out of that context and applied to a variety of policy variables to which it has no logical relationship.6 The Schlesinger study can validly be faulted for not controlling for other variables, but the subsequent research that has indiscriminately used the index and other summary measures of state political systems is guilty of the greater error of randomly associating large numbers of variables without having any theoretical basis for assuming a relationship in the first place. The indiscriminate use of independent variables in this fash— ion creates the additional problem of multicollinearity, or high correlations among independent variables that make the regression and correlation coefficients highly unreliable (large confidence intervals, high P(HO)) . Another problem with the correlation—regression policy research is its static, cross—sectional design. The theory of incrementalism suggests that policy making occurs in a matrix of formal and informal constraints that permit only a limited amount of change to occur from Year to year.7 The interaction between policy and the factors that influence it become so complex during periods of incremental change .‘ifl ”Wes cmss-swtio .m interaction betved “m preceding the anal Finally, the policy 1‘ flies are usully expenditt hisof expediency more than gImithey are mst likely 1 Mwnelations that are inflation. l1- . £ng rt10nment and Policy These conceptual and 29 that it is difficult to analytically sort out the nature and direction of the relationships. By measuring the linkage between structure and policy variables cross—sectionally, no control is possible over the cumulative interaction between independent and dependent variables in the period preceding the analysis. Finally, the policy measures used in correlation—regression studies are usually expenditure levels, once again selected on the basis of expediency more than for the characteristics of the political system they are most likely to reflect. The result is the same; hap- hazard correlations that are made "theoretical" through p93; fl interpretation . Malapportionment and Policy These conceptual and methodological shortcomings account for at least some of the inconsistency between the policy study findings that malapportionment has no impact on state policy, and the case study findings that it does. The policy studies have invariably used one or more of three indexes of malapportionment: the David—Eisenberg, Bauer-— Kelsay, and Schubert—Press. Each index provides a summary malapportion— ment score for each state. Although different in conceptual emphasis and computational technique, all three are based essentially on the decJree of variation in the size of legislative districts. The David— Eisenberg Index of malapportionment is calculated by computing the average population of a single member district in each state and then COmparing the population of actual constituencies with average consti- tuencies, the "value" of a vote representing the ratio of an actual l l l “a, Index measures en's Wm“ that can m—Ptess Index ”bin! idistribution of distric Ennis in this distributi eel state.8 Variation in distr: lithe structure of repres' an of malapportionment ml legislators disprOPC Tithe legislature, and t1 lIllesenl: measures of ma Fete policy may be relat 30 constituency to an average constituency in each state. The "value“ of a vote in the largest category of county in each state is computed for eadihouse and then the measures for both houses are averaged. The Dauer—Kelsay Index measures the theoretical minimum percentage of a state's population that can elect a majority of each house. The Schubert—Press Index combines inverted coefficients of variation in the distribution of district populations with measures of skewness and kurtosis in this distribution to produce an "apportionment score for each state.8 Variation in district size is at best only an indirect measure of the structure of representation in the legislature. The signifi- cance of malapportionment from a policy standpoint was that it gave nual legislators disproportionate power in the decision—making bodies of the legislature, and therefore control over policy. The inability 0f present measures of malapportionment to account for variation in state policy may be related to their emphasis on indirect rather than direct measures of legislative power. The level of malapportionment in states in 1962 represented the gradual accumulation over many years of disproportionate rural power in legislatures. For reasons noted above, the incremental nature of both structural and policy change during this period obscures the specific influence of malapportionment on policy. Reapportionment, on the other hand, represents a significant disruption of the incremental chain. In the course of five years, the demographic structure of legislatures was altered to conform to the ”one man, one vote" principle. In most states this meant a significant increase in metropolitan representation, sihsone, a major Shift in nut that rural-oriented Pc lleqlslatureS. and to the | hislative power to metropo main types of policies. wide a more valid basis 1 stems on state POliCY tha In sum, althou‘lh th hisysteus have a signifi ushte policy making and tiled to adequately conce 'ilhships they have examir Lisleading conclusions. < knot significantly rela litreapportionment is r Llfinal conclusions are Edllt‘mnal research is n time“ Systems and of l «filed, and then tests tl 31 andin some, a major shift in the legislative power structure. To the extent that rural-oriented policies are the result of rural domination oflegislatures, and to the extent that reapportionment has shifted legislative power to metropolitan areas, changes can be expected in certain types of policies. Reapportionment would therefore seem to provide a more valid basis for examining the influence of apportionment systems on state policy than malapportionment. In sum, although the structural characteristics of state polit— ical systems have a significant and theoretically valid relationship to state policy making and policy outputs, recent policy studies have failed to adequately conceptualize the structural variables and rela— tionships they have examined, resulting in premature and potentially IMsleading conclusions. One such conclusion is that malapportionment is not significantly related to state policy, the implication being that reapportionment is unlikely to produce much policy change. Before any final conclusions are drawn as to the impact of reapportionment, additional research is needed that specifies the dimensions of appor— tionment systems and of state policy that are most likely to be re— lated, and then tests them in a context of actual change. An Alternative Model The structure of the legislature can be seen as a framework or constraint that shapes the legislative process. It represents the DECessary but insufficient prerequisite for a demographic or partisan coalition to form, and as such, influences the probability of party or constituency related policies being adopted. lull mm“ and won 9 W mm} Ford we vote. The ultimte 1 m0! interest is its aal byislation. Legislative votes C insists are more stable fluent in the legislaturl aloof state policy are though party is the best isilence varies under dil ifil‘ltaphic structure, or Me less important than its states and on those Mal minority to be! Studies of the c fie Partisan and demogra are“ the potential vc ii is °f legislators, a: 32 State policy outcomes are largely determined by the configura— tion of power in a legislature at a given point in time. Legislative power is of both a formal and informal nature. Informal power is largely intangible and involves the exchange of what Coleman calls "political currency."9 Formal power, in contrast, inheres primarily tithe vote. The ultimate test of the legislative power of a given group or interest is its aability to muster sufficient votes to enact legislation. Legislative votes cluster along a variety of dimensions. Some dimensions are more stable than others. Two that are structurally humrent in the legislature and exert an ongoing influence on the char— acter of state policy are the partisan and the demographic dimensions. Although party is the best single predictor of legislative voting, its influence varies under different conditions and across issues.10 The demographic structure, or rural—urban cleavage, is generally considered to be less important than party, but its influence is significant in those states and on those issues where malapportionment has permitted the rural minority to benefit at the expense of the urban majority. Studies of the correlation of legislative voting suggest that Um partisan and demographic structures of state legislatures, defined here as the potential voting power of partisan and/or demographic—based blocs of legislators, are directly related to specific types of state Policy. While the underlying conceptual dimensions of these relation- Ships have yet to be tested, it would seem plausible that party structure is more likely to influence conflictual, ideologically—based hling mt so much with “hi hell in a given area. but Woe is to be allocated sion to the “distributiv ilythebest exauples of < minus forms of state aid fate aid formilae is well ’fhmtionment and State aid to Education Just as apportion alime than other types 5More clearly “POlitil at to variations in Is 33 issues which relate to the appropriate role and responsibility of gov- ernment in regulating individual behavior and in apportioning burdens and benefits among the population. Using the Lowi typology, these would be considered "regulative" and “redistributive" types of issues. Demographic structure is most likely to influence policies dealing not so much with whether or how much the state should involve itself in a given area, but rather, how an already established policy resource is to be allocated among competing interests. Such policies conform to the "distributive" dimension of the Lowi typology. Prob— ably the best examples of distributive types of state policies are various forms of state aid. The influence of malapportionment on state aid formulae is well documented and was discussed in Chapter I. Reapportionment and State Aid to Education Just as apportionment systems are likely to influence state aid more than other types of policy, so also are some types of state aid more clearly ”political" in their content, and therefore more sub— ject to variations in legislative power, than others. General purpose grants to local communities, for example, are usually distributed on a strict per capita basis, making them relatively immune to legislative tampering. Various types of categorical grants, however, reflect the values and priorities of the dominant power bloc in the legislature and sPeCify how the aid monies are to be distributed. A major type of categorical aid that has traditionally reflected the rural domination Of the legislatures is state aid to education. l i l l I It whim the Signifil liq “Emma” Educal malmm‘semdm ‘ nlyeigllt percent in 1967- comment for 19'70'71"1 nstmns “3mm“ mtry has meant that Sta. Educational finance and he functions and powers '1th to perform themsei ill.llmever, growing dis; Situational costs have pr lute: share of the res: 111968-69, they contrib‘ ihpublic elementary an this ranged from a high in in New Hampshire ~11 State support P1 :1th since their beg N . 10 education total] 34 Formal responsibility for public education rests with the sep— arate states under the implied powers clause of the Tenth Amendment to theConstitution. The federal government has traditionally become involved in educational policy making only in times of clear national need. Even with the significant increases in aid related to the Ele— mentary and Secondary Education Act of 1965, the federal portion of total elementary—secondary educational expenditures reached a peak of only eight percent in 1967—68 and has since dropped back to less than seven percent for 1970—71.11 A strong tradition of Jeffersonian localism in much of the country has meant that states as a group have had a subordinate role n1educational finance and administration and have tended to assume those functions and powers that local communities were unable or un— willing to perform themselves. Particularly since the Second World War, however, growing disparities in local resources and rapidly rising educational costs have provided the impetus for states to assume a greater share of the responsibility for financing public education. h11968—69, they contributed an average of 39.9 percent of expenditures for Public elementary and secondary schools. As shown in Table II—l, this ranged from a high of 85.1 percent in Hawaii to a low of 9.3 per— cent in New Hampshire.12 State support programs have grown in both magnitude and com— plexity since their beginnings in the 1930's.13 Expenditures for Public education totalled over $28 billion in l966~67, ranking it second only to national defense in policy priorities. Of the $19.1 billion in total state aid to local governments during the same year, STATE PERCENT OF ___________,__,_.__—-————-—-'__,_.__.—— 1. Hawaii 85 1. Delaware 71 3. North Carolina 65 4. Georgia 5‘ New Mexico 5 6. Washington 5 an 1. South Carolina 8. Alabama E 3. Louisiana '21 Florida Kentucky 11- Mississippi '3- Utah 14. New York L Tennessee :5- Texas 17' Arizona '5‘ Minnesota 19‘ Arkansas N. West Virginia 21“ Michigan 22' PermsYlwania 23- Alaska 10. ll. l2. l3. 14. 15. l6. l7. 18. 19. 20. 21. 22. 23. b u N H . . . 0 (D \l O" . n . o 35 Table II—l STATE PERCENT OF PUBLIC SCHOOL EXPENDITURES, 1969 Hawaii Delaware North Carolina Georgia New Mexico Washington South Carolina Alabama Louisiana Florida Kentucky Mississippi Utah New York Tennessee Texas Arizona Minnesota Arkansas West Virginia Michigan Pennsylvania Alaska United States 85.1% 71.4 65.2 59.7 59.7 59.4 59.3 59.0 57.8 56.7 52.6 51.6 50.1 48.4 47.6 47.1 46.8 46.1 45.0 44.7 44.4 43.7 40.7 39.9 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. Idaho Oklahoma Maryland Virginia Nevada Maine Indiana California Rhode Island Missouri Iowa Connecticut Ohio Vermont Montana New Jersey Kansas Illinois North Dakota Wisconsin Colorado Wyoming Massachusetts Oregon Nebraska South Dakota New Hampshire 39.8 38.6 37.3 37.0 35.4 34.7 34.0 33.9 32.9 32.7 32.6 31.3 29.5 28.6 27.6 27.3 27.1 26.6 26.2 25.2 24.0 20.6 18.9 17.7 16.0 11.4 9.3 _~_~___________________*____.~______________________~______~_~_*_______ me formulae by “hm lull! we! and “mm” mm in each stai m four basic types Of an and special purpose- sicategory determines t1 loofits to local district: national program depend this contributed by the s Mthough the tren: a: greater Equalization meg“ 810“. uneven, a El states. Resistance “ties 0f the apportiw 36 $11.6 billion, or 62 percent, was for elementary and secondary educa- timL Public welfare, the second largest category of state aid, _ . l4 accounted for only 15.2 percent by comparison. The formulae by which state aid funds are distributed are hidfly complex and understood in detail by only a few educational fhmnce experts in each state. Generally, funds are distributed 'Umough four basic types of grants: flat, equalizing, general pur— pose and special purpose. The proportion of total aid allocated to eadicategory determines the program priorities and the patterns of benefits to local districts. The impact of state aid on the overall educational program depends upon the proportion of total educational . 15 funds contributed by the state. Although the trend has been toward greater state responsibility and greater equalization in the distribution of state aid, such changes have been slow, uneven, and have met with considerable resistance in nbst states. Resistance to equalization is based on the political realities of the apportionment process. A number of concessions to the equalization principle have to be made in order to make the aid formula acceptable to a majority of legislators in each house. In a 1969 study of state aid, the Advisory Commission on Intergovernmental Relations noted some of the ways this can occur: There are . . . many points where slippage between the Goal of equalization and the actual distribution of State aid may occur. In some states, for example, equalization relates to a relatively small portion of tOtal State funds provided. . . . While a portion of State aid may equalize, it may have only a slight im— Pact on local service levels if the total funds for this purpose are small, while the totality of State education aid may, in fact, work against equalization. legislators are 3“ idols in their district- mtion sends each 13915] are. Interviews with i inofficials. and member Mtituency benefits are in the aid package-17 RC dually: legislators view-01' i not so much as a mean: ltles than as an inst money for their own c tion of cutting the b for one's home distri educational opportuni State. Thus, the gre lnations has always ionization but to t school districts thn T0 the extent th 37 Even where equalization governs the distribution of a large portion of State education assistance, such formulas may be based only in part on local ability, with additional measures also used. These additional factors may, in fact, turn out to work against equali— zation.l6 Legislators are aware of the amount of aid allocated to the schools in their district. In Michigan, the State Department of Education sends each legislator a quarterly statement listing these figures. Interviews with a number of lobbyists, department of educa— tion officials, and members of the governor's staff also suggest that constituency benefits are a major determinant of legislator support for the aid package.17 Roger A. Freeman states the case more empha— tically: Legislators view——or are expected to view——state aid not so much as a means of helping low—income commun— ities than as an instrument of getting the most state money for their own constituency. This is a proposi— tion of cutting the biggest possible slice of the pie for one's home district, rather than of improving educational opportunities in other sections of the State. Thus, the greater part of state school appro— priations has always been used not for the purpose of equalization but to channel state funds back to local school districts throughout the state.18 To the extent that the Michigan interviews and the Freeman nmerpretation are valid, the distribution of state aid should reflect the relative power of different types of constituencies in the legis- lature. The relationship between rural over—representation in the legislature and rural favoritism in state aid has been discussed. The evidence, although inconsistent, supports such a relationship. Reapportionment has increased metropolitan representation in moSt state legislatures. In some, the metropolitan delegation has of slate aid is Eels siof these types of chan lblislrlbution of state a amphically-based la and of reapportionment o in The purpose of the tnoased voting power of lilures resulting from rea listribution of state aid Sletropolitan legislati lithe uehopolitan dele‘i mine the relationship loficy both longitudinal ‘119 distribution of stat effects of metropolitan lfimortionment . The f( i3Imtheses The central ass tionship between the di legislatures , and the fore, 38 remained a minority, in others a majority, and in others it has moved frmna minority to a majority status. To the extent that the distri— bution of state aid is related to constituency-based voting blocs, eadiof these types of change should produce predictable changes in the distribution of state aid. This study will develop the concept of demographically—based legislative coalitions in explaining the impact of reapportionment on the distribution of state aid. The Study The purpose of the study is to examine the extent to which the increased voting power of the metropolitan delegations in state legis— latures resulting from reapportionment has produced a more favorable distribution of state aid to metropolitan school districts. A measure of metropolitan legislative power based on the proportiOnal strength of the metrOpolitan delegation in state legislatures will be used to examine the relationship between legislative structure and educational Policy both longitudinally, showing the effects of reapportionment on the distribution of state aid, and cross—sectionally, showing the effects of metropolitan legislative power on state aid patterns after reapportionment. The following hypotheses will be tested. @3632 The central assumption of the study is that there is a rela— tionship between the demographic structure of representation in state leQiSlatures, and the distribution of state aid to education. There- fOre, In the present stud lidbo ednmtion depends u] M and the cohesiveness webetween central city nase as the size and sou shunts increases . There H3: The gr of met greats polite relat: distr aid. An alternative 2 s-017aula for its distribu that such expenditures . Mdeungraphic charact hBlanca or decision—nah iiucational problems an Sonic-economic charact Shh as income, educat useful, indicators of 39 The relative advantage of metropolitan school districts in the distribution of state aid is directly related to the voting power of the metropolitan delegation in the state legislature. Changes in the relative advantage of metrOpolitan school districts between 1962 and 1969 are directly related to the type and amount of change in metro- politan power resulting from reappor- tionment between 1962 and 1967. In the present study, the impact of reapportionment on state aid to education depends upon both the amount of change in metropolitan power and the cohesiveness of the metropolitan delegation. Differ— ences between central city and suburban legislators are likely to in— crease as the size and socio-economic differential between cities and suburbs increases. Therefore, H3: The greater the demographic homogeneity of metropolitan constituencies, the greater the cohesiveness of the metro— politan delegation, and the greater the relative advantage of metrOpolitan school districts in the distribution of state aid. An alternative explanation for the level of state aid and the fOrmula for its distribution is the so—called “cybernetic" argument that such expenditures are more a reflection of the socio-economic and demographic characteristics of states than of their political power balance or decision—making processes. Metropolitan areas have distinct educational problems and needs. A general indication of need is the secio—economic characteristics of the metr0politan areas. Factors such as income, education, and racial composition are indirect, yet useful, indicators of metropolitan educational need. According to the Ndistricts. The £91. I! : The 1e and the itan sch of state socio-ea politan (he of the central mention promotes progr ht organized opposition :ease welfare and other 1: mm political power . Mirectly related to th flatly, electoral partici miveness of policy ma welfare-oriented expendii fidRobinson, Dye, and 0+ fliparty competition the mmuasiveness of th wire that the relatic be examined. The follo‘ H : The i and itan tion to ‘t stat 4O 'rybernetic" argument, these factors are an important determinant of the amount and proportion of state aid distributed to metropolitan school districts. The following hypothesis will therefore be tested: H4: The level of metropolitan per pupil aid and the relative advantage of metropol- itan school districts in the distribution of state aid are directly related to the socio-econOmic characteristics of metro- politan populations. One of the central tenets of political science is that party competition promotes progressive, liberal policies. The argument is that organized opposition creates incentives for policy makers to in- crease welfare and other types of benefits in order to acquire or maintain political power. It follows that party competition should be directly related to the amount and distribution of state aid. Sim— ilarly, electoral participation is also thought to increase the re— sponsiveness of policy makers to public pressure for increases in welfare—oriented expenditures. Despite the recent findings of Dawson and Robinson, Dye, and others that cast some doubt on the validity of the party competition model, the inconsistency of these findings and the persuasiveness of the theory in the state politics literature require that the relationship between party competition and state aid be examined. The following hypotheses will therefore be tested: H5: The level of metropolitan per pupil aid and the relative advantage of metropol— itan school districts in the distribu— tion of state aid are directly related to the level of party competition in states. Since the study ass ally'political' in nature milbles and the state ai nhtionshjp between the s more. Therefore, H7: The la and th itan 5 tion 0 party cipati econon polite A second assumpt: mition of the legis 1in and distribution 0 Therefore, H : The i and ‘ itan tion met: part A final assumpi “Sure based on metro; direct measure based c inExplaining state p< 41 H z The level of metropolitan per pupil aid and the relative advantage of metropol— itan school districts in the distribu— tion of state aid are directly related to the level of electoral participation in states. Since the study assumes that the state aid formula is basic- ally "political" in nature, the relationship between the political variables and the state aid measures should be stronger than the relationship between the socio—economic variables and the state aid measures. Therefore, H7: The level of metropolitan per pupil aid and the relative advantage of metrOpol— itan school districts in the distribu— tion of state aid are more related to party competition and electoral parti— cipation than they are to the socio— economic characteristics of the metro— politan population. A second assumption of the study is that the demographic composition of the legislature is a more important determinant of the level and distribution of metropolitan aid than party competition. Therefore, H : The level of metropolitan per pupil aid and the relative advantage of metropol— itan school districts in the distribu— tion of state aid are more related to metropolitan legislative power than to party competition. A final assumption of the study is that a direct apportionment measure based on metrOpolitan legislative power is superior to an in— direct measure based on variation in the size of legislative districts in exPlaining state policy choices. Therefore, Hg: The new measure tioment patterns! rifles and Research Pros The central concept in. pmr being defined lure of metropolitan po amnion of legislators ‘J’lncounties, plus the p] asserting metropolitan mounts for differences bases of the same legisl in both houses have r0\ ‘é‘estate aid formula. 1 hforty-three states f Estate legislative mar legislative power resul1 i1”like basic power meas i‘scussed in Chapters '1 The major dePe‘ Education. Three spec 1] total per pupil aid ilvided by the number 2’ metropolitan per 13 al'stricts in metropol 42 H : The metropolitan legislative power measure is superior to other appor- tionment measures in explaining patterns of state aid to education. Variables and Research Procedures The central concept of the study is metropolitan legislative power, power being defined as potential voting strength. The basic measure of metropolitan power (PWR) , the independent variable, is the proportion of legislators in the lower house representing metropol— itan counties, plus the proportion of legislators in the upper house . . . . . 19 . representing metropolitan counties, diVided by two. This measure accounts for differences in rural over—representation across different houses of the same legislature under malapportionment, and it assumes that both houses have roughly equal influence in the develoPment of the state aid formula. Metropolitan legislative power was calculated for forty—three states for both 1961—62 and 1967—68 from information in state legislative manuals and bluebooks.2o Measures of change in legislative power resulting from reapportionment and two refinements in the basic power measure for 1967-68 were also calculated. They are discussed in Chapters Three and Four. The major dependent variable of the study is state aid to Education. Three specific measures of state aid will be used: 1) total per pupil aid (PPA), or the total state aid appropriation, divided by the number of students in average daily attendance; 2) metropolitan per pupil aid (MPPA), or the PPA received by school dlStinCts in metrOpolitan counties; and 3) metropolitan relative W M): 01 the rat plaid. These measures pfimt, and for 1968-6 WWMMmm thud formula. Changei apportiomt were also 4 mespondence with indivi huh forty-nine departme Mata, the data they fa tent formats, and were u .EWided sufficient data The alternative e Ethe hypotheses will be SOcio-economic Varia 1. Personal Inc 1968 . 2. Change in PE Personal In: 3. Educational for Public Income , 196 4. 1960 Metro; population 5. Percentage Areas (PCT living in 6. Large Cit: the 1960 1 100,000. 43 advantage (MRA), or the ratio of metropolitan to non—metropolitan per pupil aid. These measures were calculated for 1961-62, before reap— portionment, and for 1968-69, at least one year after reapportioned legislatures had had an opportunity to make their will felt on the state aid formula. Changes in these measures during the period of requrtionment were also calculated. The data were obtained through correspondence with individual state departments of education. Al— though forty-nine departments eventually responded to the requests for data, the data they forwarded were of uneven quality, had dif— ferent formats, and were often incomplete. Only twenty-six states 21 provided sufficient data to be included in the study. The alternative explanations for state aid patterns discussed hithe hypotheses will be tested with the following control variables: Socio-economic variables 1. Personal Income (PERSINC)——Per Capita Personal Income, 1968. 2. Change in Personal Income (CGINC)——Percent Increase in Personal Income, 1958—68. 3. Educational Tax Burden (BURDEN)~—Local and State Revenue for Public Schools in 1968—69 as Percent of Personal Income, 1968. 4. 1960 Metropolitan Population (METPOP)——Size of the 1960 population living in SMSA counties (in thousands). 5. Percentage of the 1960 Population Living in Metropolitan Areas (PCTMET)——Pr0portion of total population in 1960 living in SMSA counties. 6. Large City Metropolitan P0pulation (LGCTYPOP)—~Size of the 1960 population living in metropolitan cities over 100,000. 1. tenet of hetn (ment-La: as a proportion Percent of hon-1 mm)--The population in 1 9. Percent Student stmlents attenc‘ 10. change in Metr: in population, Political Variables 11. Long—term Part Party Competiu 12. Short-term Par tition in the 13. long—term Ele Index of Part 14. Short-term El for Gubernatc 15. State role it State and L01 state in 196‘ 16. Apportionmen of variation 17. Apportionmer the state 9: legislature 18. Change in It and 1967. 19. Change in I: and 1.967. The sources fo: We to the analysis , “'11 be discussed in t 44 7. Percent of Metropolitan Population in Large Cities (PCTLGCTY)-—Large city metropolitan population in 1960 as a preportion of the total metr0politan population. 8. Percent of Non—whites in Large Metropolitan Cities (NONWHITE)~—The percent of the large city metropolitan population in 1960 that was non—white. 9. Percent Students in Public Schools (PCTPUB)—-Percent of students attending public schools in 1967. 10. Change in Metropolitan Population (CGPOP)——Percent change in population, 1960—69. Political Variables ll. Long-term Party Competition (RANNEY)—-Ranney Index of Party Competition. 12. Short—term Party Competition (PTYCOMP)——l967 Party Compe— tition in the State Legislature. 13. Long—term Electoral Participation (MILBRATH)——Milbrath Index of Participation. l4. Short—term Electoral Participation (GOV)-—Percent turnout for Gubernatorial election nearest to 1968. 15. State role in Educational Finance (STPCT)—~Percent of State and Local Public School Revenues contributed by the state in 1969. 16. Apportionment Variable #l (ICV)-—An inverted coefficient of variation in legislative district size, 1962 and 1967. 17. Apportionment Variable #2 (DK)--The minimum percentage of the state population necessary to elect a majority of the legislature, 1962 and 1967. 18. Change in ICV (CGICV)——Percent change in ICV between 1962 and 1967. 19. Change in DK (CGDK)——Percent change in DK between 1962 and 1967. The sources for each of these measures, their theoretical rele— ‘ane to the analysis, and their relationship to the state aid measures, ‘nll be discussed in the following chapters. It should be noted that minuapters. Their t1 nnmnuonsnip to the Elysis that follows. The research methq “We statistics and bonnlation and regress “in" the extent to whi< can. War. and the re: '1 45 the above measures are calculated at the state level, unless otherwise stated. The sources for each of these measures, and where appropriate, a description of the derivation, are presented in footnotes in suc— ceeding chapters. Their theoretical relevance to the analysis and their relationship to the state aid measures will be discussed in the analysis that follows. The research methods used in the study will range from simple descriptive statistics and graphic plots of bivariate relationships, to correlation and regression techniques.22 The study will first examine the extent to which reapportionment produced changes in metro— politan power, and the relationship between changes in power and changes in state aid, in Chapter III. In Chapter IV, the basic power measure will be refined and its increased explanatory power vis-a—vis the state aid measures assessed. In Chapter V, the relationship be— tween the two sets of control variables and the state aid measures will be examined. In Chapter VI, the relative explanatory power of different sub-sets of variables, as well as their cumulative explana— tory power, will be examined with multiple regression techniques. In Chapter VII, the findings will be related to the hypotheses guiding the study, appropriate conclusions drawn, and suggestions made for future research . 1MB“! succinc mgsychological and t “mph 11. semesinger' 9 untheSocial Sciences (1" WSCJ‘ 110-911. 2pm major exampl haleSysten (New York: J M (New Haven: Cluster analysii ’alaan, “The Systematic A1 tumor,“ Western Polit. lunatic theory, cutlin tithe Uses of Function a1. LXII, to. 2 (June 1 Pfluence of competition WDavid Meltz in an unp instituency Influences. 11 (February 1969) . 3 4 Meltz, M11 5Aug R : e . Claus la5enate~1iouse Voting g=er1c - - 32' onPohtlcal Sci 6 3511a . The Index of s . and Re:“fir's article. _. th V Me Bro Ines (ed: CHAPTER II Notes 1A useful, succinct discussion of the differences between the social—psychological and the rational-calculus perspectives is found hlJoseph A. Schlesinger's review of Seymour Martin Lipset, Politics and the Social Sciences (New York: Oxford University Press, 1969) in the American Political Science Review, Vol. LXIV (September 1970), 910-911. Two major examples are John C. Wahlke, et al., The Legisla— 2 tive System (New York: John Wiley and Son, 19657—535 James Barber, The Lawmakers (New Haven: Yale University Press, 1966). 3 . . . Cluster analysis of roll call votes is explained in John ermm "The Systematic Analysis of Blocs in the Study of Legislative Behavior," Western Political Quarterly, Vol. XVIII (1965), 350—362. Axiomatic theory, outlined by A. James Gregor in "Political Science and the Uses of Functional Analysis," American Political Science Review, Vol. LXII, No. 2 (June 1968), 428—440, is applied to an analysis of the influence of competition on cohesion in majority party roll call voting by David Meltz in an unpublished monograph, “Legislative Party Cohesion: A Model of the Bargaining Process in State Legislatures." A regression analysis of party and constituency influences in state senate roll call votes is found in Hugh L. LeBlanc, "Voting in State Senates: Party and Constituency Influences," Midwest Journal of Political Science, Vol. XIII (February 1969), 33—57. 4 . Meltz, op. Cit. 5Aage R. Clausen and Richard B. Cheney, "A Comparative Analysis of Senate-House Voting on Economic and Welfare Policy, l953—l964," American Political Science Review, Vol. LXIV, No. 1 (March 1970), I38— 152. 6 . . The Index of Formal Governor Powers originally appeared in Schlesinger's article, "The Politics of the Executive" in Herbert Jacob and Kenneth Vines (eds.), Politics and the American States (Boston; Little, Brown and Company, 1965). 46 7The theory of incr as originally developed by lildavsky in their article, lexican Political Science 519-541. 88cc Paul '1‘. David oiPirginia Bureau of Publ nilobert G. Kelsay, "Unr P11351101. XLIV (Decembe tales Press, "Measuring egg, V01. LVIII (June 1 9 lanes S. Coleman use Review, Vol. LXIV, N 10 See Julius Turns 3% (Revised Edition by inns Hopkins Press, 197( 11 Rankings of the heathen, 1971). P- 4 12 w” p. 45. 13 For a good dis. PUMP systems, see Robe: onion: Machillan, 19 47 7The theory of incrementalism in the federal budgetary process was originally develOped by Otto A. Davis, M. A. H. Dempster and Aaron Wildavsky in their article, "A Theory of the Budgetary Pr0cess," American Political Science Review, Vol. LX, No. 3 (September 1966), 529-547. 8See Paul T. David and Ralph Eisenberg, Devaluation of the Urban and Suburban Vote, Vols. I and II (Charlottesville: University of Virginia Bureau of Public Administration, 1961); Manning J. Bauer and Robert G. Kelsay, "Unrepresentative States," National Municipal Review, Vol. XLIV (December 1955), 571—575; Glendon Schubert and Charles Press, "Measuring Malapportionment," American Political Science Review, Vol. LVIII (June 1964), 302-327. James S. Coleman, "Political Money, American Political Sci— ence Review, Vol. LXIV, No. 4 (December 1970), 1074—1087. 1 . OSee Julius Turner, Party and Constituency: Pressures on Con— gress (Revised Edition by Edward V. Schneier, Jr.). (Baltimore: The Johns Hopkins Press, 1970). l . . . lRankings of the States, 1970 (Washington: National Education Association, 1971), p. 46. 12 Ibid., p. 45. 13For a good discussion of the development of educational fi— nance systems, see Robert J. Garvue, Modern Public School Finance (London: MacMillan, 1969), particularly Chapter 9. 14Advisory Commission on Intergovernmental Relations, State Aid to Local Government (Washington, D.C.: ACIR, 1969), pp. 4, 31—35. 15 A thorough description of state revenue programs can be found in Albert R. Munse, Revenue Programs for the Public Schools in the United States, 1959—60 (Washington, D.C.: U.S. Department of Health, Education and Welfare, Office of Education, 1961). ”ACIR, W, op.-__ci.t-: p- 46- l7These interviews were held during the summer of 1971 as part Of some exploratory research on educational reform in Michigan. The research project was funded by the Urban Institute and was under the direction of Professor Frank A. Pinner of the MSU Political Science Department. "‘-IIIFF'-H l9In cases where le| mum counties, unmet aunties was ‘1 Willa“: then the the amt! 0f residencfl antics, or if there W31" mcpolitan counties. th‘ election) to make the 56* airtit igmres consider: Inaldifferences within ‘ agaccmnn criterion co aalysis. 20 The seven states are: Alaska and Hawaii. Potion, Colorado, Vera Included in their legisli in lack metropolitan p 21 The states incl Alabama Arizona Arkansas California Connecticut Florida 48 18Roger A. Freeman, Taxes for the Schools (washington, D.C.: The Institute for Social Science Research, 1960), p. 249; see also, Laurence Iannaccone, Politics in Education (New York: The Center for Applied Research in Education, Inc., 1967). 19In cases where legislators represented both metropolitan and non—metropolitan counties, either the dominant characteristic of the constituent counties was used (i.e., if two of three counties were non—metropolitan, then the legislator was labelled non—metropolitan), or the county of residence was used (i.e., if there were only two counties, or if there were an equal number of metropolitan and non- metropolitan counties, then the legislator's home county guided the selection) to make the designation. This measure admittedly is crude, since it ignores considerable variation in the urban, suburban, and rural differences within the same county, but it was seen as the only way a common criterion could be applied to all the states in the analysis. 20The seven states not included in the apportionment analysis were: Alaska and Hawaii, because of their relatively recent entry into the Union, Colorado, Vermont, and Nebraska, because of inadequate data included in their legislative manuals, and Wyoming and Montana, because they lack metropolitan populations. 21The states included in the 26—state sample are: Alabama Louisiana Rhode Island Arizona Maryland South Carolina Arkansas Michigan South Dakota California Nevada Tennessee Connecticut New Jersey Utah Florida New Mexico Washington Georgia New York West Virginia Idaho Oregon Wisconsin Kentucky Pennsylvania 22 . The statistical programs used at various stages of the anal— YSis were the Least Squares, Least Squares Addition, and Least Squares Delection programs of the Agricultural Experiment Station and the Com~ puter Laboratory at Michigan State University, and the BMD and Statis— tical Analysis System program packages at the George Washington Univer- sity Computer Center. Appreciation is expressed to both for assistance with the analysis. In 1952, before 1' area were significantly1 PableIII-l shows the pm in countiee for 43 sta Motion for metropoli I”measures. As can b1 therePresentation in s was had over 50 perc States °n mfitronolitan CHAPTER III APPORTIONMENT SYSTEMS AND STATE AID TO EDUCATION In 1962, before the reapportionment decisions, metropolitan areas were significantly under—represented in state legislatures. Table III—1 shows the proportion of legislative seats from metropol— itan counties for 43 states together with the percentage of the states' population for metropolitan counties and the difference between these two measures. As can be seen, only ten of the 44 had 50 percent of the representation in state legislatures, even though 24 of these states had over 50 percent metropolitan population. The mean for all states on metropolitan legislative power was 34.9 percent with a standard deviation of 22.5 percent, and the mean for the metropolitan pOpulation was 59.1 percent with a standard deviation of 22.8 percent. Figure III—1* presents the relationship between population and repre— sentation more graphically. Note that all of the states were below the line of equality between metropolitan population and metropolitan representation. Considerable variation existed in urban as well as rural states in the degree Of metropolitan under—representation. For example, compare the distance of the comparatively rural states of Maine and New Mexico from the equality line, or the moderately w *All figures appear in the Appendix. 49 Idaho 29.1 90.1 86.1 68. 67. 46 14 78 6] (A) 50 TABLE III—l Pre—Reapportionment Metropolitan Underrepres entation 1960 1962 1962 Metropolitan Metropolitan Metropolitan Population Legislative Un derrepresen ta tion (Percent) Power Alabama 51.8 34.1 17.7 Arizona 71.4 40.3 31.1 Arkansas 29.1 18.1 11.0 California 90 . 2 65 . 6 24 . 6 Connecticut 86 . 4 36. 4 50 . 0 Delaware 68.9 19.2 49.7 Florida 67.1 27.6 39.5 Georgia 46.0 10.8 35.2 Idaho 14 . 0 4 . 5 9. 5 Illinois 78. 7 60.6 18.1 Indiana 61.2 49.0 12.2 Iowa 33.2 13.4 19.8 Kansas 39.1 12.7 26.4 Kentucky 34 . 8 16 . 2 18 . 6 Louisiana 53.7 42.6 11.1 Maine 27.8 27-2 0-6 Maryland 82 . 3 4O . 8 41 . 5 Massachusetts 97 . 4 94 . 8 2 . 6 Michigan 76. 2 64 - 5 ll . 7 Minnesota 49.9 33.6 16.3 Mississippi 15.6 5-9 9-7 Missouri 60.1 45.9 14.2 Inmupdfire In Jersey lea York Rm Carolina 501111 Dakota 3hio 0llléthom Wlmia We Island South camlina SW! Dakota Eelmessee 74.1 29.4 90.2 27.6 86.6 33.6 10.6 76.8 45.9 51 Table III—l (Cont.) 1960. 1962' 1962 Metropolitan Metropolitan . Population Legislative Metropolitan , (Percent) Power Underrepresentation Nevada 74.1 26.1 48.0 New Hampshire 29.4 16.6 12.8 New Jersey 90.2 53.5 36.7 New Mexico 27.6 8.4 9.2 New York 86.6 74.5 12.1 North Carolina 33.6 20.2 13.4 North Dakota 10.6 3.2 7.4 Ohio 76.8 59.7 17.1 Oklahoma 45.9 20.6 25.3 Oregon 58.7 50.0 8.0 Pennsylvania 78.8 69.9 8.9 Rhode Island 83.6 76.1 7.5 South Carolina 35.8 20.1 15.7 South Dakota 12.7 8.2 4.5 Tennessee 47.6 27.3 20.3 Texas 69.0 38.0 31.0 Utah 74.0 40.7 34.0 Virginia 53.4 20.5 22.9 Washington 63.1 54.5 8.8 West Virginia 30.9 25.0 5.9 Wisconsin 48.4 47.3 1.1 WW Wisconsin . J Conne appears t° be ins“ flutes. lith reapportionment. Manually reduced in all s “amnion of legislators ment of metropolitan popula Wily-Wee of the 44 states Mislature. Twelve have ove ilotted against 1967 metropoi than sizeable number of St “Presentation or equality. The relationship bet it»: 52 metropolitan states, Wisconsin and Georgia, or the highly urbanized states, Massachusetts and Connecticut. The amount in variation of inequality appears to be just as great for these three different types of states. With reapportionment, metropolitan under—representation was substantially reduced in all state legislatures. Table III-2 shows the preportion of legislators in 1967 for metropolitan counties, the percent of metropolitan population, and the difference between the two. Twanty—three of the 44 states now have a metropolitan majority in the legislature. Twelve have over 70 percent metropolitan representation, up from three in 1962. The mean for all 44 states is now 55.9 percent, with a standard deviation of 23.3 percent. When 1960 population is plotted against 1967 metropolitan power, as in Figure III-2, it is seen that a sizeable number of states are above the line of proportional representation or equality. The spread has clearly diminished. The relationship between metropolitan population and metropol— itan over— or under—representation is better shown in Figures III—3 and III—4. In both, the percent of population for metropolitan counties is related to the degree of urban under—representation or ratio of metropolitan population to metropolitan legislative power. Figure III—3 shows that in 1962 the degree of under—representation was greater in states with low metropolitan populations, although wide variation ex— isted also in states above 50 percent. Nine states with over 50 per— cent metrOpolitan population have less than 60 percent of the repre— sentation they merit on a population basis. Four states are deviant illinois Xmas laundry Mnisiana hryland tassadmsetts lichigan limesota Hississippi Missouri 14.0 78.7 33.3 39.1 34. 53. 27. 82 97 '76 4‘. 53 TABLE III-2 Post-Reapportionment Metropolitan Representation Alabana Arizona Arkansas California Connecticut Delaware Florida Georgia Idaho Illinois Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri 1960 Metropolitan Population (Percent) 51.8 71.4 29.1 90.2 86.4 68.9 67.1 46.0 14.0 78.7 33.2 39.1 34.8 53.7 27.8 82.3 97.4 76.2 49.9 15.6 60.1 1967 Metropolitan Legislative Power 57.4 70.0 35.5 89.4 84.0 67.6 68.8 39.6 14.3 75.2 29.2 45.8 22.8 47.1 27.6 63.8 98.3 84.3 51.7 6.4 56.7 1967 MetrOpolitan Representation —l.3 +1.7 —6.4 —3.5 +6.7 —12.0 -0.1 —18.5 +0.9 +8.1 Gregor 58 . 7 Mnsylvania 7 8 - 5 Rhode Island 83 .4 Sun Carolina 35. 501101 Dakota 1 2 - Mmessee 47 . E'exas 69 ‘u‘tah 74 Virginia 5: lashington 6 Rest Virginia 3 Kisconsin A W 54 Table III—2 (cont.) 1960 1967 1967 Metropolitan Metropolitan Metropolitan Population Legislative Representation (Percent) Power Nevada 74.1 70.0 -4.1 New Hampshire 29.4 45.3 +15.9 New Jersey 90.2 63.3 —26.9 New Mexico 27.6 11.7 -15.9 New York 86.6 84.1 —2.5 North Carolina 33.6 33.2 —0.4 North Dakota 10.6 12.2 +1.6 Ohio 76.8 73.7 —3.1 Oklahoma 45.9 43.6 —2.3 Oregon 58.7 59.2 +0.5 Pennsylvania 78.8 79.2 +0.4 Rhode Island 83.6 91.2 +7.6 South Carolina 35.8 31.3 ~4.5 South Dakota 12.7 11.6 —1.1 Tennessee 47.6 48.0 +1.4 Texas 69.0 87.4 +18.4 Utah 74.0 67.3 —6.7 Virginia 53.4 38.8 —14.6 WHShington 63.1 57.6 —5.5 West Virginia 30.9 30.2 —O.7 Wisconsin 48.4 49.1 +0.7 W mitten to begin with. W' metropolitan representatio Its. his can be attributed Monetize progressivism and mflict in Wisconsin politics M its metropolitan resident bite. This may in part be do Negation poses to the non—m ilsorelated to the yankee p Mien of Maine politics. W limpolitan populations sig States with similar metropol: MILO at .80 and .65, re §°I this pattern unless it i Empolitan industrial bast inherit rural interests in The most extensive 01962 are in states well Mever, the pattern is di ”Nation is seen to be t of its population was met: this group was represente 55 hithe direction of equality in metropolitan representation——Maine, Wisconsin, West Virginia and South Dakota. Massachusetts has almost fifll representation for metropolitan residents, but it is 98 percent metropolitan to begin with. Wisconsin is only half metropolitan, but its metropolitan representation is as equitable as that of Massachu— setts. This can be attributed in part to the reformist tradition of la Follette progressivism and in part to the lack of urban and rural conflict in Wisconsin politics. Maine is only 28 percent metropolitan, yet its metropolitan residents are fully represented in the legisla- ture. This may in part be due to the lack of threat the metropolitan delegation poses to the non—metropolitan majority, but it is probably also related to the yankee puritan tradition and the Republican domi— nation of Maine politics. West Virginia and South Dakota give their metropolitan populations significantly more representation than other states with similar metropolitan percentages, yet both are considerably under 1.0 at .80 and .65, reSpectively. No simple explanation exists for this pattern unless it is that neither state has a large enough metropolitan industrial base to constitute a distinct threat to the dominant rural interests in the state. The most extensive cases of metropolitan under—representation in 1962 are in states well known for their malapportionment. Again, however, the pattern is difficult to explain. Georgia's metropolitan population is seen to be the most under—represented: only 46 percent Of its population was metropolitan in 1962, yet only 23 percent of this group was represented in the legislature. Delaware had a Moi them were represent uliiomia represented a mjori Hperomt and 73 percent res : mtflopolitan populations, laser metropolitan states. In some, wide variatio plitan areas were under-repr Mitan areas proportional re sander tended to under—rep Minion to their percenta Figure III—4 shows Ration and metropolitan undel tionment. Although there is tional representation) . Stat 1-0 than in 1962, particulai ”Motion is above 50 pero this pattern are worth noti 15“ percent representation 931M. In contrast, Miss “Presentation of its 16 P rent from 30 percent to 42 SOPUlation. New Jersey m1 WPercem: metropolitan p 56 69 percent metropolitan population, yet only 28 percent of this popu- lation was represented. Connecticut had the fifth highest proportion of metropolitan residents in the country (86.4 percent), and only 42 percent of them were represented in the legislature. New Jersey and California represented a majority of the metropolitan residents, with 59 percent and 73 percent respectively, but compared to their 90 per— cent metropolitan populations, the inequity was as great as in many lesser metropolitan states. In some, wide variation existed in the degree to which metro— politan areas were under—represented. Only three states gave metro— politan areas proportional representation in the legislature. The remainder tended to under—represent metropolitan residents in direct Proportion to their percentage of the population. Figure III—4 shows the relationship between metropolitan popu— lation and metropolitan under—representation for 1967, after reappor— tionment. Although there is still variation (about 1.0, or propor— tional representation), states now cluster much more closely around 1.0 than in 1962, particularly those states in which the metropolitan Population is above 50 percent. Several significant deviations from this pattern are worth noting. New Hampshire moved from 56 percent to 154 percent representation of the metropolitan population in this period. In contrast, Mississippi moved from 38 percent to 41 percent representation of its 16 percent metropolitan population. New Mexico went from 30 percent to 42 percent of its 28 percent metropolitan population. New Jersey moved from 59 percent to 70 percent of its 90 percent metropolitan population; Kentucky from 47 percent to Although these figures representation decreased betw hiemunt of the change res ntmpolitan power can be po tile III-3 presents the fi crease in power, broken into retropolitan delegation rema flined a majority, and state Bjority status. is can be seen, the Winn 1.5 percent in Mai] {We is 78.4 percent, wit ibirteen states more than c‘ the legislature and 21 inc: Shwed less than a 20 para Table III-3 also points or isalso misleading as an 1' States which showed signii Etropolitan majority to ”Rt remained a minority ‘ 57 66 percent of its 35 percent metropolitan population. The greatest metropolitan under-representation continues to exist after reappor— tionment in states with small metropolitan populations. In states such as New Jersey and Maryland, however, metropolitan under— representation continued, despite predominant metropolitan popula— tions. Although these figures show that the degree of urban under— representation decreased between 1962 and 1967, they do not indicate the amount of the change resulting from reapportionment. Change in metropolitan power can be portrayed in Several different ways. Table III—3 presents the figure for 1962, 1967, and the percent in— crease in power, broken into the three groups-—states in which the metropolitan delegation remained a minority, states in which it re— mained a majority, and states in which it moved from a minority to a majority status. As can be seen, the amount of change varied considerably, rang- ing from 1.5 percent in Maine to 281 percent in North Dakota. The mean figure is 78.4 percent, with a standard deviation of 80.78 percent. Thirteen states more than doubled the metropolitan representation in the legislature and 21 increased it by over 50 percent. Only twelve showed less than a 20 percent increase in metropolitan strength. Table III—3 also points out the fact that the percent change figure is also misleading as an indicator of legislative power, since many states which showed significant percentage increases either have a metropolitan majority to begin with, or have a metropolitan minority that remained a minority. In order to test the hypothesis that a ll. 15. 11. 18. a.— 5‘ 20 Kansas Kentucky Louisiana Maine Mississippi New Hampshire New Mexico North Carolina North Dakota Oklahoma South Carolina Seuth Dakota Tennessee Virginia West Virginia Wisconsin 58 TABLE III—3 Metropolitan Legislative Power in 1962 and 1967, with the Percent Increase ’in Power, by Type of Change 1962 PWR 1967 PWR PCT CHG Minority Status Before and After Reapportionment 1. Arkansas 18.1 35.5 96.1 2. Georgia 10.8 39.6 266.7 3. Idaho 4.5 14.3 217.6 k 4. Iowa 13.4 29.2 117.9 5. Kansas 12.7 45.8 260.6 6. Kentucky 16.2 22.8 40.7 7. Louisiana 42.6 47.1 10.6 8. Maine 27.2 27.6 1.5 9. Mississippi 5.9 6.4 8.5 10. New Hampshire 16.6 45.3 172.9 11. New Mexico 8.4 11.7 39.3 ' 12. North Carolina 20.2 33.2 64.4 13. North Dakota 3.2 12.2 281.3 14. Oklahoma 20.6 43.6 111.7 15. South Carolina 20.1 31.1 55.7 16. South Dakota 8.2 11.6 41.5 17. Tennessee 27. 3 48.0 15.8 18. Virginia 20. 5 38. 8 36.1 19. West Virginia 25-0 30-2 20-8 20. Wisconsin 47.3 49.1 3.8 EM ' status Before and Mmmrthment 1. (Minnie 2. Illinois l. Massadmsetts 4. Michigan 5. New Jersey 5. New York 7. Ohio 8. Pennsylvania 9. Rhode Island 10. Washington linority Status to Majority isNResult of Reappgrtionme 1. Alabama 1. Arizona 3. Connecticut 4. Delaware 5- Florida 6. Indiana 7- Maryland 3. Minnesota 9. Missouri 10. Nevada 11. Oregon 12- Texas 13. Utah 59 Table III-3 (Cont.) 1962 PWR Majority Status Before and After Reapportionment 1. California 65.6 2. Illinois 60.6 3. Massachusetts 94.8 4. Michigan 64.5 5. New Jersey 53.5 6. New York 74.5 7. Ohio 59.7 8. Pennsylvania 69.9 i 9. Rhode Island 76.1 ' 10. Washington 54.3 Minority Status to MajOrity Status as a Result of Reapportionment 1. Alabama 34.1 2. Arizona 40.3 3. Connecticut 36.4 I 4. Delaware 19.2 5. Florida 27.6 6. Indiana 49.0 7. Maryland 40.8 8. Minnesota 33.6 9. Missouri 45.9 10. Nevada 26.1 11. Oregon 50.0 12. Texas 38.0 13. Utah 40.7 1967 PWR 89.4 75.2 98.3 84.3 63.3 84.1 73.7 79.2 91.2 57.6 57.4 70.0 84.0 67.6 68.8 63.5 63.8 51.7 56.7 70.0 59.2 87.4 67.3 PCT CHG 36.3 24.1 3.7 30.7 18.3 12.5 23.5 13.3 19.8 6.1 68.3 73.7 130.8 252.1 149.3 W :3: and the 1968 2 high necessary. The nee macabetween 1962 and 19 manpower divided by .510, resents the figures for chan legislative majority ranked f Using this measure, a Mange but no change in stat sale, North Dakota drops fr Ihilansas from 3 to 12. Texas, on the other llto 2, and California from role that the metropolitan f reapportionment. It will b ropolitan states are ranked 32, Washington 34, Wisconsi the demographic compos itior lent. Without such change Study, no changes in the hence the low ranking. T Siderable amount of Chang latures as a result of 1: Changes in state aid to 60 change in metropolitan status from a minority to a majority produced the greatest change in state aid, a measure that reflects both the amount of change and the 1968 position related to a legislative ma— jority is necessary. The measure adopted for this purpose is the difference between 1962 and 1967 metropolitan power, times 1968 metro- politan power divided by .510, a legislative majority. Table III—4 presents the figures for change in metropolitan power relative to a legislative majority ranked from most to least change. Using this measure, a number of states with a high percentage change but no change in status are dropped down on the list. For ex— ample, North Dakota drops from No. l to No. 38, Georgia from 2 to 16, and Kansas from 3 to 12. Texas, on the other hand, moved from 11 to 1, Connecticut from 10 to 2, and California from 24 to 6. These changes reflect the new role that the metropolitan delegation played in the states following reapportionment. It will be noted that a number of predominantly met— rOpolitan states are ranked quite low. Rhode Island 28, Massachusetts 32, Washington 34, Wisconsin 39. This is due to the lack of changes in the demographic composition of the legislature following reapportion— ment. Without such change, at least in the context of the present study, no changes in the distribution of state aid could be expected; hence the low ranking. Taken as a group, these measures show a con— siderable amount of change in the demographic structure of state legis- latures as a result of reapportionment. The analysis now turns to the Changes in state aid to education. 26. 27. 28. 29. 30. 31. 32. 33. New Mis Ore Nor Vi: Sot Mas 61 TABLE III—4 Changes in Metropolitan Legislative Power, Ranked High to Low 1. Iowa 877.80 2. Texas 844.74 . 3. Connecticut 785.40 i 4. Delaware 643.72 3 5. Nevada 601.43 E 6. Florida 556.20 I 7. California 416.50 { 8. Arizona 406.89 9. Utah 356.44 i 10. Michigan 326.70 . 11. Alabama 309.89 12. Kansas 297.90 13. Maryland 287.50 14. Rhode Island 270.29 15. New Hampshire 252.56 16. Georgia 224.64 17. Illinois 214.62 18. Ohio 203.00 19. Oklahoma 199.50 20. Tennessee 194.58 21. Minnesota 182.81 22. Indiana 181.25 23. Pennsylvania ‘ 144.15 24. New Jersey 121.52 25. Arkansas 120.06 26. New York 120.00 27. Missouri 119.88 28. Oregon 106.72 29. North Carolina 84.50 30. Virginia 78.28 31. South Carolina 68.32 32. Massachusetts 67.55 33. Louisiana 41.40 34. Washington 37.29 35. West Virginia 30.68 36. Kentucky 29.70 37. Idaho 27.44 38. North Dakota 20.70 39. Wisconsin 17.28 40. Maine 11.04 41. South Dakota 7.82 42. New Mexico 7.26 43. Mississippi 6.00 W raidionula is ostensibly signedto equalize the eve . tinting students. In actua rid the equalization princip through the legislature. Thi vious dlapter. state aid is impriated per pupil, or p- hastate is more related t {Mic education than to the tribution of that aid, howe ital factors and is the foct It should be noted a inthis study have no neces “Educational program, for Mconclusive evidence on i achievement. More money m; tional programs. Second, tion in the cost of educat needs of students. Final? the wide variation in the the states . 62 State Aid to Education State aid to local school districts in most states serves as a supplement to the amount raised by the school districts themselves. The aid formula is ostensibly based on some measure of need, and is designed to equalize the overall resources available to districts for educating students. In actual fact, significant compromises are made with the equalization principle in order to get the aid appropriation through the legislature. This process has been discussed in the pre— vious chapter. State aid is usually examined in terms of the amount appropriated per pupil, or per—pupil aid. The level of per—pupil aid in a state is more related to the historic role played by the state in public education than to the immediate political situation. The dis— tribution of that aid, however, is more directly influenced by polit— ical factors and is the focus of this study. It should be noted at the outset that the aid figures examined in this study have no necessary relationship to the overall quality of an educational program, for several reasons. First, there is presently no conclusive evidence on the relationship between resources and achievement. More money may not produce significantly better educa— tional programs. Second, the figures do not take into account varia— tion in the cost of education in different states, or the different needs of students. Finally, and most importantly, they do not reflect the wide variation in the proportion of total revenues contributed by the states. nun. In the present study Itmepoint in time is not as lanolitan and non-metropol‘ thedistribution and amount 0 In 1962, the per—pupi inthe study ranged from $78 into: all states was $181. htmpolitan per-pupil aid r hkxico, with a mean of $ heaid figures for each st imhigh to low. As can b- mpeI-pupil aid figures an andfor1968 it is .968. '1‘} the two aid figures can be rBtropolitan per—pupil aid Mt show differences in ait Vidual school districts. dI'Stricts within metropoli aid appropriations . Thes‘ The generally hig' years is due primarily to Students and the high cos 63 A high per—pupil aid figure may simply indicate a large role for the state rather than a commitment to quality education, although many students of school finance argue that the two are necessarily related. In the present study the total per-pupil aid a state provides at one point in time is not as significant as its distribution among metropolitan and non—metropolitan school districts, and change in both the distribution and amount of aid over time. In 1962, the per—pupil aid for the twenty—six states included in the study ranged from $78 in New Jersey to $330 in New Mexico. The mean for all states was $181.51, and the standard deviation was $63.33. Metropolitan per—pupil aid ranged from $70 in New Jersey to $303 in New Mexico, with a mean of $165.73 and a standard deviation of $65.49. The aid figures for each state are presented in Table III—5, ranked from high to low. As can be seen, the per—pupil aid and the metropoli— tan per—pupil aid figures are highly related. The r for 1962 is .971 and for 1968 it is .968. This unexpectedly high relationship between the two aid figures can be explained in several ways. First, the metropolitan per—pupil aid measure is an aggregate figure which does not show differences in aid between cities and suburbs, or across indi— Vidual school districts. The educational needs and local resources of districts within metropolitan counties vary considerably, as do state aid apprOpriations. These are masked by the measure used here. The generally high metropolitan per—pupil aid figures for both Years is due primarily to the special educational needs of central city Students and the high costs of operating central city school systems. State Per Pupil Aid l. lamina 2. Oregon 3. Keviork L HERE 5. iashington 6. Ibrth Carolina 1 Florida E. Utah 9. Pennsylvania 10. South Carolina 11. California '3- Kentucky 3 Arizona 14- Georgia 5-hnyhmd 16. Louisiana 3-5 Rest Virginia 17.5 Alabama Tennessee 20- Michigan 21. Rhode Island 12‘ Arkansas 23. Idaho 24» Connecticut 35. Wisconsin 25. New Jersey $330. 295 274, 266 259 220 209 20 20 w M N N M m m N H H H H H H H H H H m u: A w M H o m u u m m b w M H o . . . . . . . . . . . . . . . . . Q) (I) \l 0‘ U1 vh- U) N H . . . u o - - . - 1962 State Per Pupil Aid and Metropolitan Per State Per Pupil Aid New Mexico Oregon New York Nevada Washington North Carolina Florida Utah Pennsylvania South Carolina California Kentucky Arizona Georgia Maryland Louisiana West Virginia Alabama Tennessee Michigan Rhode Island Arkansas Idaho Connecticut Wisconsin New Jersey TABLE III-5 Pupil Aid, Ranked $330. 295. 274. 266. 259. 220. 209. 203. 202. 201. 186. 179. 176. 173. 166. 156. 149. 149. 142. 136. 122 119. 114. 110. 107. 78. 00 OO 00 00 OO 00 00 00 00 00 00 OO 00 OO 00 OO 00 00 00 00 .OO 00 00 OO 00 00 Metropolitan Per Pupil Aid 0’) \l 0‘ U1 4}- w M H a . . . K) 10.5 10.5 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. New Mexico Oregon New York Nevada Washington North Carolina South Carolina Utah Pennsylvania California Arizona Florida Maryland Alabama Georgia Michigan Kentucky West Virginia Tennessee Louisiana Rhode Island Connecticut Idaho Arkansas Wisconsin New Jersey $303.00 295.00 261.00 259.00 248.00 214.00 211.00 196.00 183.00 177.00 177.00 154.00 148.00 138.00 137.00 132.00 131.00 129.00 123.00 122.00 120.00 110.00 101.00 95.00 75.00 70.00 W idiot-ulna are attuned more hgolitiml coalitions, and - daresteristics across differ Isililar proportion of the t milhasis. This finding w' dapter. Given the strong I study is on metropolitan pow aid figures will be used onl “Metropolitan per-pupil - figures will constitute the The relative advant 1962 ranged from a low of .1 (Malina. The mean for all 93' .106. The values for a] fate that the metropolitan Arizona and South Carolina districts. The remaining four being located in the W198. six in the seventy We. and one in the fif‘ these patterns will be pr 65 Central city school systems rely more heavily on state aid for their programs than suburban school systems. A second explanation for the high correlation is that state aid formulae are attuned more to educational needs and resources than to political coalitions, and that metropolitan schools have similar characteristics across different states that results in their getting a similar preportion of the total state aid appropriation on a per— pupil basis. This finding will be given further attention in the final chapter. Given the strong relationship, and since the focus of the study is on metropolitan power and aid patterns, the state per—pupil aid figures will be used only intermittently in the remaining chapters. The metropolitan per—pupil aid and metropolitan relative advantage figures will constitute the two major dependent variables. The relative advantage of metropolitan school districts in 1962 ranged from a low of .49 in Florida to a high of .107 in South Carolina. The mean for all states was .808 with a standard deviation 0f .106. The values for all states are presented in Table III—6. NOte that the metropolitan districts in only three states (Oregon, Arizona and South Carolina) are at parity with the non—metropolitan districts. The remaining states trail off to Florida's .49 percent, four being located in the 90 percent range, seven in the 80 percent range, six in the seventy percent range, and four in the 60 percent range, and one in the fifty percent range. Possible explanations for these patterns will be presented in a later section. q-. 19. Ne 20. Ge 21. Ke 22. C: 23. Lt 24. W 25. N 26. I 66 TABLE III—6 1962 Metropolitan Relative Advantage, Ranked 1. South Carolina 1.07 2. Arizona 1.01 3. Oregon 1.00 4. North Carolina .96 5. Michigan .95 l 6. Rhode Island .93 g 7. Utah .90 V 8. Washington .89 9. New Mexico .89 10. Nevada .89 11. Idaho .87 12. Alabama .87 13. Connecticut .86 14. West Virginia .82 15. Tennessee .78 16. New York .77 17. Arkansas .76 18. Pennsylvania .71 19. New Jersey .71 20. Georgia .70 21. Kentucky .68 22. California .67 23. Louisiana .64 24. Wisconsin .60 25. Maryland .59 26. Florida .49 manoeuvrredinthepo a Nation in state legislat dammed in the amount - M969. This section will = In related. 1962 Power and State with power and metropoli - iHippenclix. Several patte hckof any systematic relat creased metropolitan power - iii expenditures to metropo Empolitan per—pupil exper 5393. has the second smalle: WY: 27.6 percent. The c ables is .048. The range in metro; States with considerable ml TSiqnificant metropolitan P°1itan per-pupil aid in l Mitan power in the legi! The relationship “lative advantage is son Zinct. As seen in Figure 67 Metropolitan Legislative Power and State Aid The previous two sections have established that significant increases occurred in the potential voting power of the metropolitan delegation in state legislatures between 1962 and 1967, and that changes also occurred in the amount and distribution of state aid between 1962 and 1969. This section will examine the extent to which these changes are related. 1962 Power and State Aid: The relationship between 1962 metro— politan power and metropolitan per—pupil aid is potted in Figure III—5 hlAppendix. Several patterns emerge. The most obvious pattern is the lack of any systematic relationship between the two variables. In— creased metropolitan power does not have any apparent influence on state aid expenditures to metropolitan districts. In fact, the state in which netropolitan per-pupil expenditures are the greatest, New Mexico, with $303, has the second smallest metropolitan population of any in the study, 27.6 percent. The overall relationship between the two vari— ables is .048. The range in metropolitan per—pupil aid is just as great in States with considerable metropolitan power as it is in states with insignificant metropolitan power. The conclusion must be that metro~ Politan per—pupil aid in 1962 was not systematically related to metro— POlitan power in the legislature. The relationship between metropolitan power and metropolitan relative advantage is somewhat different although not much more dis— tinct. As seen in Figure III—6, metropolitan power does not appear ‘N as states . Illomhtinn between the t Mpite this lack of ‘ Wit states, several in “to support the hypothesi W. urbanized states - M and to the right from Pennsylvania, California, Ne l‘ichigan could also be inclu ism through these states Mesized direction. A second group of s trialized, begins with Geor< through Kentucky, Arkansas , into Utah, Arizona, and Ore States would also be in the 5“! the two regression lim With low urbanization and iosupport public educatio therefore rely on state re 59m the case in the Soutl 5etter in such states bot? “Greater role played b t v 68 to advantage metropolitan districts in the distribution of state aid. States with metropolitan majorities show just as much variation in their aid patterns as states with metropolitan minorities. The over— all correlation between the two variables is —.O45. Despite this lack of relationship among the variables for all twenty—six states, several interesting sub—patterns can be seen which tend to support the hypothesis. The first is a group of fairly indus— trialized, urbanized states beginning with Maryland and extending up— ward and to the right from the figure through Louisiana, Wisconsin, Pennsylvania, California, New Jersey, New York, and Rhode Island. Michigan could also be included in this group. If a regression line is run through these states alone, the pattern is clearly in the hy— pothesized direction. A second group of states, generally less urbanized and indus~ trialized, begins with Georgia and extends upward and to the right through Kentucky, Arkansas, Tennessee, West Virginia, Alabama, and then into Utah, Arizona, and Oregon. A regression line drawn through these states would also be in the hypothesized direction. One explanation for the two regression lines would be differences in resources. States With low urbanization and industrialization have fewer local resources to support public education than urban, industrial states, and must therefore rely on state revenues more heavily. This has traditionally been the case in the South. Metropolitan districts are likely to fare better in such states both because of their greater need and because of the greater role played by the states in funding educational services. hm-Etmpolitan districts loot mmicipal overburden ' fixing service costs) . and th lively tax the wealth of the uliance on state aid. is metropolitan legi States, the relative advanta- file this interpretation of ltpresent, it offers a part rarefully with additional da 1967 Metropolitan L- Ed: The post-reapportionme ‘Jtive power and metropolita memst striking change in States with a metropolitan l liar. Another significant State aid. Whereas the ma: “"99 in 1962, in 1968 the In terms of patter iihles is no better than 1' xmible in Figure III—7: increased metropolitan p0 69 Although urban, industrial states generally have metropolitan areas with greater local resources to fund public education, and hence more justification for distributing a greater proportion of state aid to non-metropolitan districts, it should be noted again that the prob— lem of municipal overburden in the central city (declining tax base, rising service costs), and the inability of central Cities to effec— tively tax the wealth of the outlying suburban areas, increases their reliance on state aid. As metropolitan legislative power increases in both groups of states, the relative advantage of metropolitan districts increases. While this interpretation of the data is only tentative and preliminary at present, it offers a partial exPlanation that can be examined more carefully with additional data. 1967 Metropolitan Legislative Power and 1969 Metropolitan State Aid: The post-reapportionment relationship between metrOpolitan legis— lative power and metropolitan per—pupil aid is shown in Figure III—7. The most striking change in the figure is the increase in the number of states with a metropolitan legislative majority. This was noted ear— lier. Another significant change is in the increase in the amount of state aid. Whereas the majority of states were located in the $100—$200 range in 1962, in 1968 these states clustered in the $200—$300 range. In terms of patterns, the overall relationship between the var— iables is no better than in 1962 (r_= .063). If any pattern is dis— cernible in Figure III—7, it is one of decreasing metropolitan aid with increased metropolitan powar. Other than New York, which as a result lists. his is consistent v blastsection. The ar u: Mackof support among mo- hexpected to be greater. Paulina and New Mexico) pro nettqnlitan school distric- Iljority of non-urbanized, aetmpolitan districts than his is partly attributable cipalover-burden, but it . would be expected if the r The relationship b! relative advantage in 1968 correlation coefficient is sane support for the stud} below .5 on the vertical arately, it is seen that to 50 percent of legislai increases most dramaticai and malapportionment hav California and New York. until the metropolitan d Of State aid, metropoli' 70 of Governor Rockefeller's educational program has significantly in- creased its aid to local schools, the highly industrialized urbanized states are notably limited in their aid to metropolitan school dis— tricts. This is consistent with the resources argument entered in the last section. The argument is somewhat invalidated, however, by the lack of support among most of the southern states where aid could be expected to be greater. That three states (North Carolina, SOuth Carolina and New Mexico) provided higher levels of support to their metropolitan school districts fails to offset the fact that the greater majority of non-urbanized, industrialized states provided less aid to metropolitan districts than states with greater metropolitan wealth. This is partly attributable to lower educational costs and lower muni— cipal over—burden, but it also indicates a smaller state role than would be expected if the resource argument were valid. The relationship between metropolitan power and metropolitan relative advantage in 1968 is also low, as shown in Figure III—8. The COrrelation coefficient is .Ol6. Again, however, sub—patterns show some support for the study's hypothesis. If the patterns above and below .5 on the vertical axis, metropolitan power, are examined sep— arately, it is seen that metropolitan relative advantage decreases up to 50 percent of legislative majority, and then begins increasing. It increases most dramatically in states where the urban—rural conflict and malapportionment have historically been the greatest: Connecticut, California and New York. On its face this pattern would suggest that until the metropolitan delegation is able to control the distribution 0f state aid, metropolitan school districts are disadvantaged in the was: educational p .1 This explanation impl sively, which we have alreac Explanation is that compromi lithe or caucus that allows it relative to non-metropol ities lose relative to the his that unless it is ei iuemd has the favor of . it has control of the legis amot expect to be able to itan school districts in th Changes in Metropo] iii: When changes in metr‘ “Why is plotted agains Phil aid between 1962 and Pattern emerges. The stai Metropolitan legislati' iier, show the greatest i ilan, Arizona and Florida ihe least change in metrt Vida. and Connecticut) . 71 aid formula. This may be due to the political rivalry between metro- politan and non-metropolitan interests, or simply to the greater local resources the metropolitan school districts usually have available to support their educational programs. This explanation implies that the metropolitan bloc votes cohe- Sively, which we have already said it does not. The only alternative explanation is that compromises are made within the delegation in com- mittee or caucus that allows both city and suburban districts to bene— fit relative to non—metropolitan districts, even though the central cities lose relative to the suburbs. The pattern in this figure sug— gests that unless it is either a very small percentage of the legisla— ture and has the favor of the non—metropolitan delegation, or unless it has control of the legislature itself, the metropolitan delegation cannot expect to be able to gain an advantage position from metropol~ itan school districts in the state aid formula. Changes in Metropolitan Legislative Power and Changes in State Aid: When changes in metropolitan power relative to a legislative Kajority is plotted against the percent of change in metrOpolitan per— pupil aid between 1962 and 1969, as shown in Figure 111—9, a curious Pattern emerges. The states that experience the greatest increases in metropolitan legislative power, using the definition described ear— lier, show the greatest increase in metrOpolitan per—pupil aid (Mich— igan, Arizona and Florida). But they were also the states which showed the least change in metropolitan per-pupil aid (California, Utah, Ne— vada, and COnnecticut). The remainder of the states in the study 18. M 19. Ni 20. M 21. I 22. l 23. ( 24. ‘ 25. 26. 72 TABLE III-7 Percent Increase in Metropolitan Per Pupil Aid, 1962—1969, Ranked 1. Florida 1.883 2. New York 1.375 3. Arizona 1.367 4. Michigan 1.280 5. New Jersey 1.271 6. Wisconsin 1.107 7. Idaho 1.069 a 8. Tennessee 1.057 ‘ 9 South Carolina 1.024 10. Arkansas 0.968 11. Oregon 0.966 12. Rhode Island 0.925 13 . West Virginia 0. 845 i 14. Kentucky 0.817 15. Alabama 0.812 16. Georgia 0.781 17. Pennsylvania 0.765 18. North Carolina 0.659 19. New Mexico 0.637 20. Maryland 0.622 21. Louisiana 0.598 22. Utah 0.566 23. California 0.469 24. Connecticut 0.436 25. Washington 0.315 26. Nevada 0.178 W“— in metropolitan powe lumticut, all of which as power, did not increase the‘ the more to the demograph 91mm of the state than to system of public education t themst progressive in the state aid formula combined ‘ through a rapidly expanding and therefore legislative p iii. A similar situation In strong New England sense of teases in state aid would ”increase state aid to me hihutable to the local we: ifidReno. The local wealt] it relatively low service Elimihates much of the nee Why Michigan, Ariz aid to metropolitan school 5‘3 Speculated upon . Mi chi litEd to the power of the 73 cluster together on both variables. Again, the interpretation must be that significant increases in metropolitan per—pupil aid during this period required but did not necessarily follow from significant in— creases in metropolitan power. The fact that California, Nevada, and Connecticut, all of which experienced major increases in metropolitan power, did not increase their aid to metrOpolitan school districts may be due more to the demographic setting and the previous educational system of the state than to shifts in metropolitan power. California‘s system of public education has traditionally been considered one of the most progressive in the country. The progressive, "professional” state aid formula combined with the local resources made available through a rapidly expanding local economy could have reduced the need and therefore legislative pressure for increased metropolitan per—pupil aid. A similar situation may have existed for Connecticut, although a strong New England sense of localism and a resistance to major in— creases in state aid would probably also be a factor. Nevada's failure to increase state aid to metropolitan school districts is probably at— tributable to the local wealth of its two metropolitan areas, Las Vegas and Reno. The local wealth available to support public education, plus the relatively low service needs of Nevada's metropolitan population, eliminates much of the need for additional state aid. Why Michigan, Arizona and Florida substantially increased their aid to metropolitan school districts during this period can also only be speculated upon. Michigan's increase is probably most directly re— lated to the pOWer of the educational establishment, both teachers and gimme. During the :m: Mively bargain with local uniting rapid increases in gained an increased local sures for additional state A The federal money ma nary and Secondary Educai inthe size and the role of intum made it mre effect aid. Finally, the racial x between 1965 and 1967 focu slStems and produced press Arizona‘s demograp outage of the population difference between the tw‘ 10ml wealth that Nevada higher service needs. Or through reapportionment, have felt more pressure “is greater state aid to Florida‘s incree be as directly related ' Mmination of Florida‘s under malapport ionment 74 administrators, in actively lobbying for increased state aid in the legislature. During the same period, teachers won the right to col— lectively bargain with local school districts over salaries, and the resulting rapid increases in teachers' salaries throughout the state produced an increased local tax burden which in turn produced pres— sures for additional state aid. The federal money made available through Title 2 of the Ele- mentary and Secondary Education Act of 1965 permitted a rapid expansion in the size and the role of the Michigan Department of Education, which in turn made it more effective in lobbying for and administering state aid. Finally, the racial violence in Detroit and other Michigan cities between 1965 and 1967 focussed attention on the problems of urban school SYstems and produced pressures for additional state aid. Arizona's demography is similar to Nevada's, with a large per— centage of the population living in several metropolitan areas. The difference between the two states is that Arizona does not have the local wealth that Nevada has and it has a metropolitan population with higher service needs. Once it attained a majority in the legislature through reapportionment, the Arizona metropolitan delegation may well have felt more pressure to satisfy these service needs, one of which was greater state aid to metropolitan schools. Florida's increase in metropolitan per—pupil aid can probably be as directly related to reapportionment as any state in the country. Domination of Florida's legislative policy making by rural legislators under malapportionment has been examined at length by Havard and Beth. ' its priori' Ilshicts. Increases in met-rope level of state support for m not indicate whether such clf unopelitan districts. Ta} ment increase in the rel; hicts between 1962 and 196 were made in only six of th in York, California, Idahc istered gains of from zero inpower. The Florida inc: that reapportionment was a increase did not occur une Wuld have been expected 1 lease of non-metropolitan The state with th relative advantage was Ne last section because it <' itan power as the other : traclition of metropol ita gains of metropolitan di that even though the met 75 Reapportionment permitted the metropolitan delegation to redress some of its past grievances. The position of Florida in Figure III—9 indi— cates that one of its priorities was state aid to metropolitan school districts. Increases in metropolitan per—pupil aid show changes in the level of state support for metropolitan school districts, but they do not indicate whether such change is greater or less than for non— nmtropolitan districts. Table III—8 and Figure III-lo present the percent increase in the relative advantage of metropolitan school dis- tricts between 1962 and 1967. The figures show that significant gains were made in only six of the twenty—six states in the study: Florida, New York, California, Idaho, Maryland and Alabama. The remainder reg— istered gains of from zero to twenty percent regardless of the increase in power. The Florida increase of 91 percent supports the argument that reapportionment was a significant factor in the change, since the increase did not occur unequally uniformly throughout the state as could have been expected from the last figure, but rather at the ex— Pense of non—metropolitan districts. The state with the second greatest increase in metropolitan relative advantage was New York. New York was not discussed in the last section because it did not show as great an increase in metropol— itan power as the other states, but it is another state with a strong tradition of metropolitan out—state legislative conflict. The relative gains of metropolitan districts following reapportionment indicates that even though the metropolitan delegation was already in a majority 16. 1 17. 18. 20. ‘ 21. 1 22. 23. 24. 25.5 25.5 76 TABLE III-8 Percent Increase in Metropolitan Relative Advantage, 1962—1969, Ranked 1. Florida 85.7 2. New York 66.2 5 3. California 52.2 4. Idaho 46.0 5. Maryland 37.3 6. Alabama 27.6 7. Kentucky 16.2 8. Rhode Island 15.1 9. Wisconsin 13.3 10. Washington 12.4 11. Connecticut 11.6 12.5 Louisiana 10.9 12.5 Arizona 10.9 14.5 New Mexico 09.0 14.5 Nevada 09.0 16. New Jersey 08.5 17. West Virginia 07.3 18. South Carolina 06.5 20. Tennessee 05.1 21. Michigan 04.2 22. Arkansas 03.9 23. Georgia 01.4 24. Utah 01.1 25.5 Oregon 00.0 25.5 North Carolina 00.0 . of state ll mainly other factors, sil law. also contributed to neutral of the legislature- mall. While California SI of state aid to metropolit- change in metropolitan rel; severely malapportioned Se: state aid formula. Figure isted this control and a1 lute the already high leve Idaho's 50 perceni cannot be attributed to i: leaves limited resources areas as the most likely Maryland‘s increa 63.8 percent and Alabama b0th provided the majori I”Gunilla and increase the tricts. In five of the relative advantage , the “aPPOrtionment provide i, t 77 in 1962 in New York, with 74 percent of the seats in both houses, it took the additional 10 percent provided by reapportionment to shift the distribution of state aid more in favor of metropolitan districts. Certainly other factors, such as municipal over—burden and urban vio— lence, also contributed to this change, but the history of up—state control of the legislature makes reapportionment a likely major factor as well. While California showed relatively little change in the amount of state aid to metropolitan districts, it had the third highest change in metropolitan relative advantage. In 1962, California's severely malapportioned Senate allowed rural interests to control the state aid formula. Figure III—10 indicates that reapportionment elim— inated this control and allowed metropolitan legislators to redistri— bute the already high level of aid more toward metropolitan centers. Idaho's 50 percent increase in metropolitan relative advantage cannot be attributed to increased metropolitan legislative power, which leaves limited resources and a shift in population from rural to urban areas as the most likely explanation. Maryland's increase in metropolitan power from 40.8 percent to 63.8 percent and Alabama's increase from 34.1 percent to 57.4 percent both provided the majority necessary to take control of the state aid formula and increase the relative advantage of metropolitan school dis— tricts. In five of the six distinctive cases of gains in metropolitan relative advantage, the increases in metropolitan power resulting from reapportionment provided a clearly plausible explanation for the fimlnytoexanine the in htiunof state aid is in t itan delegation. The hypot mat of change in both mt relative advantage will be delegation moved from a mi] Mined either in a minor avalue of 1 to states in Which it remained a majori status, the relationship I state aid can be portraye< relationship for changes cance of this pattern is response of these three t have a minority metropol: fairly modest and unifor States which continued t ation in their increase: became a majority folio least amount of change. Change for states in c1 78 changes that occurred. While the overall relationship between the variables is small (r‘= .199), the deviant cases give support to the basic argument. Changes in Metropolitan Status and Changes in State Aid: A final way to examine the influence of reapportionment on the distri— bution of state aid is in terms of changes in status of the metropol— itan delegation. The hypothesis guiding the analysis is that the amount of change in both metropolitan per—pupil aid and metropolitan relative advantage will be greater in states where the metropolitan delegation moved from a minority to a majority position than where it remained either in a minority or a majority. By arbitrarily assigning a value of l to states in which it remained a minority, 2 to states in which it remained a majority, and 3 to states in which it changed status, the relationship between changes in status and changes in state aid can be portrayed graphically. Figure III—ll presents this relationship for changes in metropolitan per—pupil aid. The signifi— cance of this pattern is that it shows the distinct variation in the response of these three types of states. States which continued to have a minority metropolitan delegation after reapportionment show a fairly modest and uniform increase in metrOpolitan per—pupil aid. States which continued to have a metropolitan majority show more vari— ation in their increases. States in which the metropolitan delegation became a majority following reapportionment are seen to range from the least amount of change, in Nevada, to the most, in Florida. The mean change for states in category 1 was 94.9; for category 2, 91.4; and ' pupil aid 'I’nerelationship bet dancinmtropolitan a ... « hmistency. Category 1 - tively little change with . healso grouped together, ifCalifomia and New York. only Florida showing more . thanstates in the other c In sum, it would a vided a plausible explana tion of. state aid to loca asclear or as strong an :‘ 10H correlation between an bution of state aid sugge Hotel are necessary. One the composition of the m Which it is more and les 79 for category 3, 84.0. These findings do not support the hypothesis. States moving from a minority to a majority status show less of an increase in per pupil aid than states in the other two categories. The relationship between change in metropolitan status and change in metropolitan advantage in Figure III—12 shows a similar inconsistency. Category 1 states are grouped together, showing rela— tively little change with the exception of Idaho. Category 2 states are also grouped together, showing little change with the exception of California and New York. Group 3 states are spread out more, with only Florida showing more change in metropolitan relative advantage than states in the other categories. In sum, it would again appear that while reapportionment pro— vided a plausible explanation for changes in the amount and distribu— tion of state aid to local school districts in some states, it was not as clear or as strong an influence as had been expected. The generally low correlation between metropolitan power and the amount and distri~ bution of state aid suggest that additional refinements in the basic model are necessary. One such refinement is to examine more precisely the composition of the metropolitan delegation and the conditions under which it is more and less likely to influence the state aid formula. The metropolitan de representatives from a vari inediversity of demands a tuncies nuke simple typol- and coalitions within the . tions can be made. A sign mic-economic and the de suburbs. Other research officials of central citi- °flb0th the cause and the filt’ther distinctions may 1 safe to assume the basic of central city (particul lators. Distinguishing c1 0f their constituencies le‘lislative district bou trich boundaries often 6 Labeling districts as on CHAPTER IV METROPOLITAN DEMOGRAPHY AND STATE AID TO EDUCATION The metropolitan delegation of a state legislature includes representatives from a variety of different types of constituencies. The diversity of demands and expectations that flow from these consti— tuencies make simple typologies and characterizations as to cleavages and coalitions within the delegation hazardous. Still, basic distinc— tions can be made. A significant body of research has documented the socio-economic and the demographic differences between cities and Suburbs. Other research has shown that the citizenry and the elected officials of central cities and suburbs have clearly different views On both the cause and the solution of most urban problems. While further distinctions may be on shaky or empirical grounds, it seems safe to assume the basic difference in values and role orientations of central city (particularly larger central city) and suburban legis— lators. Distinguishing central city and suburban legislators in terms Of their constituencies is difficult because of the size and shape of legislative district boundaries. Particularly in state senates, dis— trict boundaries often encompass elements of both cities and suburbs. Labeling districts as one or the other inevitably involves subjective 80 ltobe‘ file'onelan, one '4 Mmthas meant that l inouqh proportion to thei what arbitrary but real for distinguishing between Memine the proportion o lamenting each type of . themetropolitan populatio of the metropolitan delega cities. The distinction . Want here as the distin Whities. The assumpti the two in the definition tion, and that these diffl lators. Where a sizeable tram larger cities, greai walition building than ‘ the delegation represent: m’Ie likely to be able t Jetropolitan delegation 81 . . l judgments that are likely to vary frOm one "expert" to the next. The alternative approach, to be used in this study, is to use a surrogate neasure based on the population characteristics of metropolitan areas. The "one man, one vote" criterion used as the basis for reap— portionment has meant that central cities and suburbs are represented in rough proportion to their percentage of the population. Using the somewhat arbitrary but realistic cut-off point of 100,000 as the basis for distinguishing between central city and suburbs, it is possible to determine the prOportion of the reapportioned metropolitan delegation representing each type of constituency. For example, if 35 percent of the metrOpolitan population lives in cities over 100,000, 35 percent of the metropolitan delegation would be assumed to represent central cities. The distinction between central city and suburb is not as im— pOrtant here as the distinction between larger and smaller metropolitan communities. The assumption is that clear differences exist between the two in the definition of problems and in approaches to their solu— tion, and that these differences are reflected in the views of legis— lators. Where a sizeable proportion of the metropolitan delegation is from larger cities, greater conflict can be expected in bargaining and coalition building than where it is an insignificant minority. Where the delegation represents demographically homogeneous districts, it is more likely to be able to act cohesively in dealing with the non— metropolitan delegation of the legislature. ta gupulati ‘32 mamas over taxes and mm life style values outage of non-whites in L fienetmpolitan delegatio indealing with the non-me The evidence avail study is limited. Since t Me: in any sub—category omnot be treated as conc the mtropolitan delegati politan legislative power lowing analysis will test tmlegislattors from largc iiite population in these if state aid. “1'99 City Representation and State Aid Table IV-l shows lation living in cities t than for all states is 3 20-57 percent. The rang EMania, South Carolina a i i 82 A second source of conflict within the metropolitan delegation is race. The special needs of larger cities are usually aggravated by a large non—white population. Basic differences between central cities and suburbs over taxes and services are reinforced by the threat to suburban life style.values that non—whites pose. The greater the per- centage of non—whites in large metropolitan cities, the more likely the metropolitan delegation is to be divided and therefore ineffective in dealing with the non—metropolitan delegation. The evidence available to test these hypotheses in the present study is limited. Since the total number of states is only 26, the number in any sub—category is likely to be so small that the findings cannot be treated as conclusive. Yet the question of division within the metropolitan delegation must be examined if the concept of metro— politan legislative power is to have any theoretical utility. The fol— lowing analysis will test whether or not the prOportion of metropoli— tan legislators frOm large metropolitan cities and the percent of non— white population in these cities influence the amount and distribution 0f state aid. Large City Representation and State Aid Table IV—l shows the percent of each state's metropolitan pOpu— lation living in cities over 100,000, ranked from high to low. The mean for all states is 34.31 percent, with a standard deviation of 20.57 percent. The range is considerable, frOm zero percent in Idaho, Nevada, South Carolina and West Virginia, to 70 percent and 76 percent l6. ] 17. i 18. 19. 20. 21. 22. 23. 24. 25. 26. 83 TABLE IV—l Percent of MetrOpolitan Population Living in Large Cities in 1960, Ranked t L i 1. New Mexico 76.79 2. Arizona 70.20 3. New York 64.50 4. Louisiana 54.00 5. Tennessee 53.59 6. Washington 49.20 7. Wisconsin 45.39 8. Georgia 41.50 9. California 40.09 10. Alabama 40.00 11. Michigan 38.00 12. Kentucky 36.89 13. Maryland 36.79 14. Oregon 35.89 15. Rhode Island 28.89 16. Florida 28.59 17. Utah 28.39 18. North Carolina 28.20 19. Pennsylvania 26.50 20. Connecticut 26.39 21. New Jersey 21.29 22. Arkansas 20.70 23. West Virginia 0.00 24. South Carolina 0.00 25. Nevada 0.00 26. Idaho 0.00 W iilrizona and New MeXiCO' ofoeuopolitan residents in of metropolitan legislators truly homogeneous and presul the sample are from the fou As already noted, t tion will be inferred from tionment was based. Using tion living in cities over line the size of the domin hthe metropolitan delega 0f metropolitan power, thi 05 the dominant metrOpoli' Hill hereafter be referre Since the present Voting Power of the metrc maslire need only reflect Ch anges Occur in the eve 0 || . f defections" from the 84 in Arizona and New Mexico. Note that only five states have a majority of metropolitan residents in large cities, and by inference a majority of metropolitan legislators from larger cities. In contrast, the only truly homogeneous and presumably cohesive metropolitan delegations in the sample are from the four states with no large cities. As already noted, the composition of the metropolitan delega— tion will be inferred from the 1960 census figures upon which reappor- tionment was based. Using the proportion of the metropolitan popula— tion living in cities over 100,000 as a base, it is possible to deter— mine the size of the dominant coalition (either large city or suburban) in the metropolitan delegation. As a proportion of the overall level of metropolitan power, this figure becomes the potential voting power of the dominant metropolitan coalition in the legislature. The measure will hereafter be referred to as Power 2. Since the present analysis is concerned only with the potential voting power of the metropolitan delegation, refinements in the power measure need only reflect potential "erosion" of that power, not what changes occur in the overall alignment in the legislature as a result Of "defections" from the dominant coalition. Figure IV—l plots the relationship between this measure and 1969 metropolitan per—pupil aid. As can be seen, although there is still considerable dispersion about the regression line, it is much less than for the simple power measure. The correlation has increased from .063 to .367. Deviant states are Nevada, Rhode ISland, Connec— tiCUt, and New Jersey, all with relatively high levels of power but low levels of aid; and New sore extent deviations from tens of the historical ecc states discussed in the 135 plained in terms of the con which will be discussed in The refined power 1 retropolitan relative advaJ states fall outside the ex being the most conspicuous creased slightly over the association between the tw The second refinen racial homogeneity of the v‘nites living in large met Idaho, Nevada, South Carol than The mefln for ali a L standard deviation of l for all States, n ”‘6 measure herea Simllified b metropolitar “Ohm. 10 n to e Percent $5sure ( 85 low levels of aid; and New York, showing the opposite pattern. To some extent deviations from the expected patterns can be explained in terms of the historical economic or other conditions in individual states discussed in the last chapter. To some extent they can be ex- plained in terms of the composition of the large city delegation, which will be discussed in a moment. The refined power measure is much less effective in explaining metropolitan relative advantage. As shown in Figure IV—2, a number of states fall outside the expected pattern, New Mexico, Idaho and Nevada being the most conspicuous examples. While the correlation has in— creased slightly over the basic power measure from .016 to .062, the association between the two variables is still negligible. The second refinement of the power measure is based on the racial homogeneity of the central city delegation. The percent of non— whites living in large metropolitan cities in 1960 ranged from zero in Idaho, Nevada, South Carolina and West Virginia, to 35.89 percent in Alabama. The mean for all states in the sample was 14.32 percent with a standard deviation of 12.62 percent. Table Iv—2 shows the figures for all states. The measure hereafter referred to as Power 3 makes the over— simplified but necessary assumption that the interests of blacks and whites in public education are essentially incompatible, and that the POwer of the metropolitan delegation is therefore diminished in direct Proportion to the percent of blacks in large cities. The resulting measure (100 minus percent of non-whites in large cities times the 17. 18. 19. 20. 86 TABLE IV—2 Percent of Non-Whites in Large Metropolitan Cities in 1960, Ranked 1. Alabama 35.89 2. Maryland 35.00 3. Louisiana 33.92 4. Georgia 33.64 5. Tennessee 31.73 6. North Carolina 30.32 7. Florida 23.50 8. Arkansas 23.50 9. New Jersey 20.03 10. Kentucky 18.00 11. Michigan 12.36 12. Connecticut 11.75 13. California 10.46 14. Pennsylvania 9.99 15. New York 9.98 16. Rhode Island 5.80 17. Oregon 5.60 18. Wisconsin 5.40 19. Washington 5.39 20. Arizona 5.10 21. New Mexico 2.90 22. Utah 2.10 23. West Virginia 0.00 24. South Carolina 0.00 25. Nevada 0.00 26. Idaho 0.00 W mléflgefis: the u WhenPovet 3 is v unrelationship again in 1 figure Ill-3, it is seen th lostates with high levels delegations domina' ted by l milieu Mexico are the be (Nation of whether the v isof metropolitan power. next chapter. When the Power 3 tive advantage in Figure Although the correlation leasure (r: .160), it i: creasing metropolitan re seen to be present for 1: extending through New Y extremes of Idaho and N Virginia, the Carolinas the relationship. To some extent Vida's position, for e 87 percent of metropolitan population living in large cities, plus the refined power measure just examined) is somewhat circuitous, but es— sentially it reflects the "white power" of large cities combined with the power of the dominant coalition in the metrOpolitan delegation. When Power 3 is correlated with metropolitan per—pupil aid, the relationship again increases from .367 to .489. When plotted in Figure IV-3, it is seen that the Power 3 measure gives special weight to states with high levels of metropolitan power and metropolitan delegations dominated by large, predominantly white cities. Arizona and New Mexico are the best examples of these states. This raises the question of whether the variable is as much a measure of wealth as it is of metropolitan power. This question will be dealt with in the next chapter. When the Power 3 measure is correlated with metropolitan rela- tive advantage in Figure IV—4, the results are again disappointing. Although the correlation is again slightly higher than for the Power 2 measure (£_= .160), it is still insignificant. While a pattern of in— creasing metropolitan relative advantage with increases in power does Seem to be present for the group of states beginning with Georgia and extending through New York, the numerous deviations ranging from the extremes of Idaho and New Mexico to Rhode Island, Pennsylvania, West Virginia, the Carolinas and Alabama, reduce the overall strength of the relationship. To some extent the deviant states can be accounted for. Ne- vada's position, for example, is largely an artifact of the definition ,z, mu an: gamma are praenty Ye mg larger and smaller me My also present. BO - uplitan delegation and the advantage in state ai defining larger cities at The second group 0 lucky. West Virginia, the higher metropolitan relat' uterus of the Power 3 m plausible. The first is ivlarge cities is high, ' interests are probably no they are in northern stat ical culture and to the J has allowed traditional . ical process. The confl legislature may thereto :easure would indicate. The second expl P°litan relative advant siveness of many south 88 used. Since it has no cities over 100,000 and therefore no large—city non—white pOpulation, neither of the factors that would reduce its power score are present, yet legislative conflict is probably present among larger and smaller metropolitan cities and racial friction is probably also present. Both would erode the effectiveness of the met- ropolitan delegation and could result in a reduced metropolitan rela— tive advantage in state aid. None of this is caught by arbitrarily defining larger cities at 100,000. The second group of deviant states consists of Arkansas, Ken— tucky, West Virginia, the Carolinas, and Alabama, all of which have higher metropolitan relative advantage scores than would be expected in terms of the Power 3 measure. Several explanations are at least plausible. The first is that even though the percentage of non—whites in large cities is high, which produces the lower Power 3 score, their interests are probably not as well represented in the legislature as they are in northern states. This is due both to the southern polit— ical culture and to the rapid growth of many large southern cities that has allowed traditional civic elites to retain control over the polit— ical process. The conflict between large and small cities in the legislature may therefore be less in these states than the Power 3 measure would indicate. The second explanation for the disproportionately high metro— politan relative advantage in these states is the increasing respon— Siveness of many southern states to national values in policy—making Standards. Whether out of a desire to attract northern industry, the hides, the highmetm- ‘ ml uy be a response to The third possible nlative advantage is desi. fiecentral cities. South inlany other parts of the costs of suburban schools the generally high percemi cation. While these explax It least a partial ration. the three metropolitan me Alternative A rtionment and State Aid Patterns The analysis thu: three measures of metrop cation. It remains to t Dore effective as a preq fires that have been de‘ As noted in Cha based essentially on tion in the size of le in showing the degree lect theoretical relat 89 need to qualify for federal funds, or simply a desire to redress past inequities, the high metrOpolitan relative advantage of southern states may be a response to these types of nationalizing trends. The third possible explanation is that the high metropolitan relative advantage is designed to benefit the white suburbs more than the central cities. Southern suburbs are growing faster than those in many other parts of the country, and the high capital construction costs of suburban schools are borne mainly by the states because of the generally high percentage of funds they contribute to public edu- cation. While these exPlanations are only speculative, they provide at least a partial rationale for the low correlation between any of the three metropolitan measures and metropolitan relative advantage. Alternative Apportionment Measures and State Aid Patterns The analysis thus far has examined the relationship between three measures of metropolitan legislative power and state aid to edu— cation. It remains to test whether the metropolitan power measure is more effective as a predictor of state aid patterns than the other mea— sures that have been developed. As noted in Chapter II, other apportionment measures have been based essentially on the principle of equity, or the degree of varia— tion in the size of legislative districts. Such measures, while useful in showing the degree of under—representation in a state, have no di— rect theoretical relationship to state policy choices. Put differently, hues‘rspsfie g:- e. of tuning whether malappo Wtage urban interes - Theneasureused ‘ criticisms by focusing dir fore and after reapportio theoretical relationship Mot have, the validity tingent upon its superior pressed to explain why th lated to state aid than - Still, if the po it should be a better prr relatively low explanato: With state aid than the In order to test Entropolitan power and i «efficient of variatio related with metropolit Vantage bOth before and aPPOI‘tiomnent measures P°1itan per—pupil aid are presented in Table 90 a high degree of inequality in legislative district size probably indi— cates an under—representation of urban areas, but unless urban under- representation is specified directly in the measure, there is no way of knowing whether malapportionment is producing state policies that disadvantage urban interests. The measure used in the present study is designed to meet these criticisms by focusing directly on metropolitan legislative power be— fore and after reapportionment. Since the power measure establishes a theoretical relationship with state aid policy that the other measures do not have, the validity or persuasiveness of the analysis is not con- tingent upon its superior explanatory power. Indeed, one would be hard pressed to explain why the other measures should be more strongly re— lated to state aid than the power measure. Still, if the power measure is theoretically more defensible, it should be a better predictor of state aid patterns. Despite its relatively low explanatory power, it should be more strongly associated with state aid than the other apportionment measures. In order to test this hypothesis in a preliminary fashion, metropolitan power and two other apportionment measures——the inverted COefficient of variation and the Dauer—Kelsey measures—-have been cor— related with metropolitan per—pupil aid and metropolitan relative ad- vantage both before and after reapportionment. Changes in the three aPPOrtionment measures have also been correlated with changes in metro— POlitan per-pupil aid and metropolitan relative advantage. The results are presented in Tables IV—3, IV~4, and IV—5. Correlations B - PWR PWR m2 .495 no .372 . 799 ICV .498 . 117 ”K .605 . 313 Wt .063 . 368 HT .017 .062 ‘Note: Full descripti following tabl 91 TABLE IV-3* Correlations Between 1962 Apportionment Variables and 1962 State Aid Variables PWR ICV DK MPPA MRA ewe 'Icv .276 DK .352 .563 nrpn .048 .030 .010 MRA .045 .160 .231 .433 TABLE IV—4* Correlations Between 1967 Apportionment Variables and 1969 State Aid Variables W PWR PWR2 pwa3 ICV DK MPPA MRA PWR PWR2 495 PWR3 .372 .799 ICV .498 .117 .227 BK .605 .313 .295 .854 MPPA .063 .368 .489 .305 .332 MRA .017 .062 .160 .160 .192 .543 *NOte: Full descriptions of the variables abbreviated in these and following tables can be found on pages 42—44. . . s» Correlation “"4" ' ‘ ' 1962-19 92 TABLE IV-5 " a Correlation Between Changes in Apportionment, 1962—1967 and Changes in State Aid, 1962-1969 CHGPWR CHGICV CHGDK CHGPWR CHGICV . 682 CHGDK . 34 3 . 71 8 CHGMPPA .076 .027 .166 CHGMRA . 236 . 450 . 214 variables minted to either of the Imeof the three measures politan per-pupil aid or . power measure was slightl aidandthe other two nea itan relative advantage. In Table Iv-4. i tables is significantly Again, the power measure strongly associated with other. The relationship Significantly stronger a tionship between metropo increases from .048 to . .030 to .305; and Dauer- refinements in the basil metropolitan per—pupil lation with metropolita Teasures. When changes i1 aid, in Table IV—S. it among the apportionmen Coefficient of Variatj 93 As shown in Table IV—3, in 1962 the relationship between the apportionment variables was not strong. The power measure was less related to either of the two measures than they were to each other. None of the three measures was strongly correlated with either metro— politan per—pupil aid or metropolitan relative advantage, although the power measure was slightly more associated with metropolitan per—pupil aid and the other two measures slightly more associated with metropol- itan relative advantage. In Table IV—4, it is seen that the relationship among the var— iables is significantly higher after reapportionment than before. Again, the power measure, now in three different forms, is not as strongly associated with the other measures as they are with each Other. The relationship between metropolitan power and state aid is significantly stronger after reapportionment than before. The rela— tiOnship between metropolitan per-pupil aid and the basic power measure increases from .048 to .063; inverted coefficient of variation from .030 to .305; and Dauer—Kelsey from .010 to .332. As noted earlier, refinements in the basic power measure increase its relationship with metropolitan per—pupil aid even more. Power 3 has the highest corre— lation with metropolitan per—pupil aid of any of the five apportionment measures. When changes in apportionment are related to changes in state aid, in Table IV—5, it is seen that again, the highest correlation among the apportionment variables is between change in the Inverted Coefficient of Variation and change in Dauer—Kelsey (.718), although w p, Metropolitan strong interrelationships Int miables, only chang is significantly related level or less, having a c. Relative Advantage. As n ing is diffith because i been the two variables . the variation of the siz increases in the distrib tricts than either incre creases in the minimum p majority of the legislat two should be more relat: leasure isn't more strox the change measures is ‘ not included in them. City and suburban coali conflict, within the ct in these two measures . Provides as much expla aPportionment variable 94 the correlation between change in the Inverted Coefficient of Variation and change in MetrOpolitan Power is also high (.684). Despite these strong interrelationships in the amount of change among the apportion— ment variables, only change in the Inverted Coefficient of Variation is significantly related to any of the state aid variables at the .05 level or less, having a correlation of .450 with change in Metropolitan Relative Advantage. As noted earlier, the interpretation of this find— ing is difficult because of the lack of a theoretical relationship be— tween the two variables. All one can conclude is that a reduction in the variation of the size of legislative districts was more related to increases in the distribution of state aid to metropolitan school dis— tricts than either increases in metropolitan legislative power or in— creases in the minimum percent of a state‘s population able to elect a majority of the legislature. Intuition would suggest that the latter two should be more related. One reason that the Metropolitan Power measure isn't more strongly related to the 1962 power measure or to the change measures is that the refinements in the 1968 measures are not included in them. Changes in the relative strength of the central City and suburban coalitions, and in the degree of racially—based conflict, within the central city delegation, could not be included in these two measures. Even without the refinements, however, it Provides as much explanatory power in most instances as the other two appOrtionment variables. mm far the anally sigh measures of associa apportionment systems and ndevelopinq interesting [items in the data, the ntory or predictive power function of the regressi- intone that measures th variables are able to ac iable. In the present 5 ation in the two measur Detropolitan power measr aseries of control vari individual impact of ea: and of these chapters , the cumulative explanat Chapter VI. Table Iv—6 pre: aliportionment variable tistics of most use in a ables are P, F, and r‘ C‘ilnputed F ratio or p' Chance. The lower t 95 The Predictive Power of the Apportionment Variables Thus far the analysis has relied on descriptive statistics and simple measures of association to examine the relationship between apportionment systems and state aid. While such statistics are useful in developing interesting interpretations and explanations of the patterns in the data, they do not permit any assessment of the explan— atory or predictive power of the model being tested. This is the function of the regression, a similar statistic to the correlation, but one that measures the extent to which one or more independent variables are able to account for variation in a given dependent var— iable. In the present study we are interested in the amount of vari~ ation in the two measures of state aid that can be explained by the metropolitan power measure, the two other apportionment measures, and a series of control variables to be examined in the next chapter. The individual impact of each of these variables will be summarized at the end of these chapters, and the interaction effect of the variables and the cumulative explanatory power of the model will be discussed in Chapter VI. Table IV—6 presents the regression statistics for the different aPPOrtionment variables and Metropolitan per—Pupil Aid.3 The two sta- tistics of most use in determining the explanatory power of the vari— ables are P, F, and r2. The first represents the probability that the computed F ratio or proportion of explained variance could occur by chance. The lower the probability, the greater the significance of Change 1962-67 CHGPWR CHGS‘I‘ATUS CHGICV CHGDK 96 TABLE IV-6 Regression Statistics for Apportionment Variables and Metropolitan Per Pupil Aid Variable b F P > F r2 ngg METPWR 15.23 .056 .809 .002 ICV .090 .022 .877 .000 DK .003 .002 .962 .000 1281 METPWR 30.94 .096 .757 .004 PWR2 .227 3.74 .062 .135 PWR3 .153 7.15 .013 .230 ICV .005 2.46 .127 .093 DK 1.17 2.97 .094 .110 Change 1962-67 CHGPWR —.001 .140 .713 .006 CHGSTATUS —5.48 .004 .950 .000 CHGICV .000 .018 .889 .000 CHGDK .002 .680 .577 .027 .‘-_=_. neF ratio; r2 represents pendent variable "explains When one examines Table IV-6, it is seen the variables for Metropolitar tenths of one percent in 1 both cases the Metropolit variation. Note also tha dropped substantially, tl the .01 level. The figures for 1 iidare generally simila 10“ r2 values. When the regress tags are examined in Ta} exPlamtory power of th apportionment variables 5-3 percent of the vari 97 . 2 . . . the F ratio; r represents the proportion of the variance in the de- pendent variable "explainedf by a given apportionment variable. When one examines these statistics for 1962 and 1967-69 in Table IV-6, it is seen that the explanatory power of the apportionment variables for Metropolitan per-Pupil Aid moved from a high of two- tenths of one percent in 1962 to twenty-three percent in 1967-69. In both cases the Metropolitan Power measures accounted for the most variation. Note also that the significance levels of the F ratios dropped substantially, the Power 3 measure in 1967—69 nearly reaching the .01 level. The figures for change in apportionment and change in state aid are generally similar to those for 1962, with high P values and low r2 values. When the regression statistics for Metropolitan Relative Advan- tage are examined in Table IV—7, there is much less improvement in the explanatory power of the variables from 1962 to 1967—69. None of the apportionment variables in either year is able to explain more than 5.3 percent of the variance, and the P values are all quite high. The unusually high explanatory power of the Inverted Coeffi— cient of Variation measure in accounting for change in Metropolitan Relative Advantage, as already noted, remains anomalous. In sum, the concept of metropolitan legislative power has been found to be a valid and useful basis for examining the influence of aPPOrtionment systems on state policy. The Power 3 measure is clearly SuPerior in accounting for variance in the level of MetrOpolitan lWill ICV DK Change 1962-69 / 98 TABLE IV—7 Regression Statistics for Apportionment Variables and Metropolitan Relative Advantage b F P < F 129; nmrpwn -3.19 .049 .820 ICV .106 .629 .559 BK —.019 1.35 .256 1967-69 METPWR 1.14 .006 .934 PWR2 .005 .093 .760 PWR3 .004 .239 .634 Icv .000 .965 .663 DK .095 .916 .650 .037 1 Change 1962—69 CHGPWR .003 1.42 .244 .055 CHGSTATUS 62.98 1.58 .218 .061 CHGICV .003 6.09 .020 .202 CHGDK .002 1.56 .293 .046 _____________________________________________‘_.______________‘____~__— variables are not better.. unexpectedly high expiana‘ dent of Variation variah relative Advantage. Some 4" littems in the Metmpoli vented in the final chapi 99 Aid, and although it is not able to account for much of the variance in Metropolitan Relative Advantage, the other apportionment variables are not better. The major anomaly of the analysis is the unexpectedly high explanatory power of the change in Inverted Coeffi- cient of Variation variable in accounting for changes in Metropolitan Relative Advantage. Some tentative explanations for these unusual patterns in the Metropolitan Relative Advantage variable will be pre— sented in the final chapter. lThe procedure °f basis for classifying dis with the heads of the Den lnllichigan indicated the all the states in the Stl for 1962, much less class and would have been more inal analysis of changes tiOn following reapporti 2 . The measure 18 that white central city tors, when the suburbs a delegation, a violation warding coalitions. St: MEI of the dominant f2 0f the dominant element . . wire. This measure pro were tried. 3 The four regre ‘ . (l) The unstax lhlch is the slope of i represents the amount l with a given change in variables being held 0 (2) F, or the Pendent variable expla (3) P > F or m, significai chance; 0f freedo occur by CHAPTER IV Notes l . . . . The procedure of subjective evaluation was conSidered as a basis for classifying districts in the present study, but discussions with the heads of the Democratic and Republican state central committees in Michigan indicated that it would be too cumbersome and imprecise for all the states in the study. Further, even obtaining the district maps for 1962, much less classifying them, proved to be a problem in Michigan and would have been more difficult in other states, making a longitud— inal analysis of changes in the composition of the metropolitan delega— tion following reapportionment next to impossible. 2 , . . . . The measure is obViously unsatisfactory, in that it assumes that white central city legislators will align with suburban legisla- tors, when the suburbs are the dominant coalition in the metropolitan delegation, a violation of the assumptions in the first refinement re- garding coalitions. Still, a measure was necessary that related the power of the dominant faction of the large city delegation to the power Of the dominant element of the metropolitan delegation in the legisla— ture. This measure proved to be the most satisfactory of a number that were tried. 3 . . . . The four regression statistics included in the table are: (l) The unstandardized regression coefficient or beta (b), which is the slope of the regression or least squares equation. Beta represents the amount of change in the dependent variable associated with a given change in one independent variable, the other independent variables being held constant; (2) F, or the ratio of the proportion of variance in the de- pendent variable explained to the proportion not explained; (3) p > F or the significance level of F for 26 and 24 degrees Of freedom, significance being the probability that a given value would occur by chance; (4) r2 is the proportion of the variance in the dependent var— iable "explained" by the independent variable. More detailed explana— tions of the mathematical derivation of these statistics and their applications can be found in Dennis J. Palumbo, Statistics in Political and Behavioral Science, esp. Chapters 7 and 8 (New York: Appleton, 1969), and N. R. Draper and H. Smith, Applied Regression Analysis (New York: John Wiley and Son, 1967). 100 _ - *.;.- - ALTERNAT IVE The discussion ar tionment and state aid i! of alternative equally P tions for state aid. Th Power 3, accounted for C per-pupil aid, and that for more than 4 percent vantage, indicates that the overall explanatory thepersuasiveness of t lower relative to other The present Chi Selies of intervening v analyzed in the last t thrte categories: ecc CHAPTER V ALTERNATIVE EXPLANATIONS OF STATE AID PATTERNS The discussion and analysis of the relationship between appor— tionment and state aid in the last two chapters has ignored a number of alternative equally plausible and theoretically grounded explana— tions for state aid. The fact that the "best" apportionment variable, Power 3, accounted for only 23 percent of the variation in metropolitan per—pupil aid, and that none of the apportionment variables accounted for more than 4 percent of the variance in metropolitan relative ad- vantage, indicates that these other factors may significantly increase the overall explanatory and predictive power of the model. Further, the persuasiveness of the Power 3 measure depends upon its explanatory Power relative to other sets of variables, not independent of them. The present chapter will examine the relationship between a series of intervening or control variables and the state aid measures analyzed in the last two chapters. These variables can be grouped under three categories: economic, social—demographic, and political. Economic Characteristics of States and Patterns of State Aid The sizeable body of research on state expenditure patterns discussed in Chapter II suggests that state fiscal policy is more related to the level of economic development and the wealth of the 101 datepmvides to local I‘ muesithastodraw sources is per-capita p 1 ship between per-capita : [upfl aid in 1968.1 As l strong (5 = .049) . New the highat income to al pupil aid. The remaind states. A tentative pat inthe lower left of th non—southern states (wi in} metropolitan per—pu esis is not supported b When disposable advantage in Figure V—I low overall relationsh: southern states are no the conclusion must be it is a negative one. sion local wealth, ap} stimulant to state ai< A more practi Wealth on state aid i 102 state than to its political characteristics. As applied to state aid, the argument in its simplest form would be that the amount of aid a state provides to local school districts is directly related to the resources it has to draw upon. The most direct measure of such re— sources is per—capita personal income. Figure V—l shows the relation— ship between per-capita personal income in 1966 and metropolitan per- pupil aid in 1968.1 As can be seen, the overall relationship is not strong (r_= .049). New York is the only one of the six states with the highest income to also provide a high level of metropolitan per— pupil aid. The remainder provide as little as or less than poorer states. A tentative pattern is revealed when the eight southern states in the lower left of the figure are not considered. The pattern among non-southern states (with the exception of New YOrk) is one of decreas- ing metropolitan per—pupil aid with increases in income. The hypoth— esis is not supported by the data. When disposable income is plotted against metropolitan relative advantage in Figure V—2, the pattern becomes even more dispersed. The .082) is still unclear even when the low overall relationship (3 southern states are not included. In terms of the present study, then, the conclusion must be that if income has any influence on state aid, it is a negative one. Put differently, personal wealth, and by exten~ sion local wealth, appear to act as a substitute rather than as a Stimulant to state aid expenditure. A more practical and direct way of examining the influence of Wealth on state aid is in terms of the educational tax burden, or the ashigh(_r_= .454) as " Still, there is consid- wdiscernible pattern . the greatest educationa Mexico and Arizona-~are growing urban centers . sonal rather than co .- aid. The relationsh advantage is presented (y .373) than for me higher than for any of cases in the diagram, . Politan advantage than Again, the states diff Explanations can be p: bases which reduce the and Alabama both have central cities. Idah ficult to understand. neither the political metropolitan populati 103 percent of personal income paid in local and state educational taxes. Figure V—3 shows the relationship between tax burden and metropolitan per—pupil aid. The correlations between the two variables is almost as high (£_= .454) as when the metropolitan power measure is used. Still, there is considerable dispersion about the regression line and no discernible pattern among the most deviant cases. The states with the greatest educational tax burden—-Louisiana, Oregon, Utah, New Mexico and Arizona——are all relatively non—industrial states with growing urban centers, which accounts in part for the reliance on per— sonal rather than corporate income and the high metropolitan per—pupil aid. The relationship between tax burden and metropolitan relative advantage is presented in Figure V—4. The correlation is lower (£_= .373) than for metropolitan per—pupil aid, but it is significantly higher than for any of the apportionment variables. The most deviant cases in the diagram, Alabama, Idaho, and New York, show more metro— politan advantage than would be indicated by the states‘ tax burdens. Again, the states differ greatly from one another, and only tentative explanations can be provided. All have substantial industrial tax bases which reduce the individual educational tax burden. New York and Alabama both have large, high service need populations in their central cities. Idaho's high metropolitan relative advantage is dif— ficult to understand, since the one metropolitan area in the state has neither the political power nor the concentrated high service need metropolitan population that might explain such an advantaged position, misinthahy‘pfithfe mthevariables is M lineappears to be sligh- increaaes in personal in aid. Again, the southe . linear pattern by having with only average increa however, there is consi- nd only moderate suppo - When changes in utmpolitan relative a ship is negligible (5 = metropolitan relative a percent increase in per increase in income. T1". The position of metropt State aid does not see! of increases in wealth come. Although the 1 deviant cases, in gen« terns than the other Semingly obvious con 104 When changes in personal income and changes in metropolitan per—pupil aid are examined in Figure V—S, it is seen that the overall pattern is in the hypotheSized direction and that the relationship be— tween the variables is moderately strong (r_= .259). The regression line appears to be slightly curvilinear, with states having moderate increases in personal income showing the greatest increases in state aid. Again, the southern states are distinctive, creating the curvi— linear pattern by having the greatest percentage increase in income with only average increases in aid. Even without the southern states, however, there is considerable dispersion about the regression line and only moderate support for the hypothesized relationship. When changes in personal income are plotted against changes in netropolitan relative advantage, in Figure V~6, the overall relation— ship is negligible (r_=-=O7l). States showing marked increases in metropolitan relative advantage range from Idaho, with only a 48.2 Percent increase in personal income, to Florida, with a 74.7 percent increase in income. The southern states are again educationally inert. The position of metrOpolitan school districts in the distribution of state aid does not seem to have been improved significantly as a result Of increases in wealth, at least as measured in terms of personal in— come. Although the tax burden variable does not account for individual deviant cases, in general it is a better predictor of state aid pat— terns than the other variables included thus far in the model. The seemingly obvious conclusion that a heightened tax burden produces high ltnoalthflgfi. but intakes the differen = plots. they suggest that ability of the political onthe citizenry, are no levels in accounting f0 becomes, "What are the . explanation is the char politan environment its Sacial-Demograhic Fac and State Aid Patterns A second alte the argument that educa rather than economic 1: educational expenditur lated to the characte: Wpulation. In the p: test this hypothesis : pOpulation; the size living in large citie P°Pu1ation that is In Public schools. The 105 metropolitan state aid has several important implications. Since per— sonal income is not related to state aid, it would appear that it is not wealth pg£_§§, but the commitment of that wealth to education, that makes the difference in expenditures. While the data are incom- plete, they suggest that political factors, measured here as the ability of the political system to impose high educational tax burdens on the citizenry, are more important than economic factors or income levels in accounting for metropolitan state aid. The question thus becomes, "What are the conditions that foster such commitments?" One explanation is the characteristics and educational needs of the metro— politan environment itself. Social-Demographic Factors and State Aid Patterns A second alternative set of explanatory variables is based on the argument that educational expenditures are a function of needs rather than economic resources or political preferences. Metropolitan educational expenditures are seen to be largely pre—determined and re— lated to the characteristics and educational needs of the metropolitan POPUIation. In the present analysis, four variables will be used to test this hypothesis: size and proportion of the states' metropolitan population; the size and proportion of the metropolitan population living in large cities (over 100,000); the proportion of large—city population that is non-white; and the percentage of students attending PUblic schools. These variables reflect various types of demands on population (in thousands heseen, although the si states in the sample var h14,537,000 in New Yo cinder 2,000,000; the deviation, 3,913, 700 . apparent in the data. Of California, New Jers Ietropolitan population Figure V-B show 1966 state population ‘1 tive advantage. The co not only low, but also pothesized. States wit from New York, with 86. With only 14 percent. inmetropolitan relati California and 1.28 in tions range from .79 i the metropolitan popu] tion of state aid to I 106 metrOpolitan public schools that might affect educational expenditures. They will be examined in turn. Figure V—7 plots the relationship between 1966 metropolitan population (in thousands) and 1969 metropolitan per—pupil aid. As can be seen, although the size of the metropolitan population of the states in the sample varied considerably, ranging from 93,000 in Idaho to 14,537,000 in New York, all but seven have metropolitan populations of under 2,000,000; the overall mean is 2,933,000 and the standard deviation, 3,913,700. No clear overall patterns or sub—patterns are apparent in the data. The low correlation (£_= .029) and the position of California, New Jersey, New Mexico and South Carolina show the metropolitan population to be a poor predictor of state aid. Figure V—8 shows the relationship between the percent of the 1966 state pOpulation in metropolitan areas and the metropolitan rela— tive advantage. The correlation (E =-=052) shows the pattern to be not only low, but also to be in the opposite direction from that hy— Pothesized. States with high metropolitan relative advantage range from New York, with 86.5 percent metropolitan population, to Idaho, with only 14 percent. States with a high metropolitan population range in metropolitan relative advantage from .77 in New Jersey to 1.02 in California and 1.28 in New York. States with low metropolitan pOpula- tions range from .79 in Arkansas to 1.27 in Idaho. The percentage of the metropolitan population does not appear to influence the distribu— tion of state aid to metropolitan school districts. neq Wound on 1°°a1 1‘ state aid. An increase: tion should therefore 9 figure v-9 presents the ntmpolitan 9091113151“ correlation hemeen the metropolitan polllllatior among the states is ml the larger number of 31 it is apparent that th« Significant bearing on any variable included Just as the ed 107 Large metropolitan cities generally have populations with dis- proportionately high need for public services. These needs place a heavy demand on local resources, which in turn increases reliance on state aid. An increase in the size of large-city metropolitan pOpula— tion should therefore produce an increase in the amount of state aid. Figure V—9 presents the relationship between the size of the large—city metropolitan population and metropolitan per—pupil aid. Although the correlation between the variables is higher, at .436, than between metropolitan population and metropolitan per—pupil aid, the pattern among the states is not Clear when presented graphically because of the larger number of states with small large-city populations. Still, it is apparent that the size of the large—city population has a more significant bearing on the level of metropolitan per—pupil aid than any variable included in the model except for the Power 3 measure. Just as the educational needs of large—city populations influ— ence the level of aid at one point in time, so also should they influ— ence changes in the level of aid over time. The racial violence of the mid—19605 and the extensive documentation of large—city educational problems by the Coleman report and other studies, both produced an im— petus for additional aid revenues. Figure V—lO presents the relation- ship between the size of the large—city metropolitan population and the change in metropolitan per—pupil aid between 1962 and 1969. The weak relationship (£_= .180) and dispersed pattern shows that the educational needs of large cities did not produce significant increases in aid. When large—city metropolitan population is related to changes in metropolitan relative to .526. Figure V-ll Pl‘ self mnderstates the Pat the correlation is not 0 suggests that shifts in school districts during educational needs of la: To the extent t] tional needs, the metro increase as the percent increases. Figure v-lz is M supported by the homer, masks several The three stat ”Nation in large ci Politan relative advan from the states with r 108 in metrOpolitan relative advantage, however, the correlation increases to .526. Figure V—ll plots this relationship. While the figure it- self understates the pattern and points up inconsistencies, and while the correlation is not objectively high, the releationship clearly suggests that shifts in the distribution of aid toward metropolitan school districts during the period were the result of the demonstrated educational needs of large cities. To the extent that large—city populations have special educa— tional needs, the metropolitan relative advantage in state aid should increase as the percentage of metropolitan residents in large cities increases. Figure V—12 presents this relationship. The hypothesis is not supported by the data. The low overall relationship (£_= .096), however, masks several interesting sub—patterns. The three states with the highest percentage of metropolitan pOpulation in large cities are the only sub—set of states with a metro— politan relative advantage greater than 1.0. When Idaho is removed from the states with no metropolitan populations in the large cities, the metropolitan relative advantage of that group drops considerably, from .99 to .89. With the exception of Wisconsin, the states with Significant large-city populations but low metropolitan relative ad— vantage are all southern states, with high concentrations of non—whites in large cities. Louisiana's large metropolitan cities are 33.9 per— cent non—white; Georgia's are 33.69 percent; and Tennessee's are 31.7 Percent, the third, fourth, and fifth highest percent of the 26 states in the study. While rural poverty and educational needs may also be a factor in the disadvantaged position of metropolitan school districts, “than states I mm has the highest percentd cities of any State in i A third demogra] “ the proportion of non-W] J urban populations requi. extent to which their 8 determined by the amoun state and federal govei types of problems shoul Figures v-13 through v- metIOPOIitan populatio: itan relative advantag tween 1962 and 1969. 109 race is probably an important determinant as well. The use of race as the explanation for a low metropolitan relative advantage in southern states, however, is clearly contradicted by Alabama, which has the highest percentage of non—whites in its large metropolitan cities of any state in this study. A third demographic factor, briefly touched upon above, is the proportion of non—whites in large metropolitan cities. Non—white urban populations require a high level of specialized services. The extent to which their special educational needs are met is largely determined by the amount of inter-governmental aid provided by the state and federal governments. The responsiveness of states to these types of problems should be reflected in the distribution of state aid. Figures V—l3 through V—lS examine the influence of large, non—white metropolitan populations on 1969 metropolitan per-pupil aid, metropol- itan relative advantage, and changes in metropolitan per—pupil aid be— tween 1962 and 1969. In Figure V—l3, it is seen that the states with the largest percentage of non—whites in large cities are all in the South, with the exception of Maryland, which is a border state. These states are also among the states with the lowest metropolitan per—pupil aid, Florida and North Carolina being the exceptions. New Jersey is dis— tinctive in that it has the largest proportion of non—whites in its metropolitan population of any northern industrial state, yet it has the third lowest metropolitan per—pupil aid of any of the twenty—six States in the study, $159- some does n°t account tieperomt in cities Ofl large metropolitan citifi Iighthe expected. The ‘ ‘ smogests that the varial 4, old. When the influei vantage is examined in home states with even cities, only Alabama pr it does to non-metrOpol Wt 31110119 the states w: Finally, when cities is related to t 110 The great majority of the states have under 12 percent non- whites in their large—city metropolitan populations. This measure of course does not account for the variation in percent across cities or the percent in cities of less than 100,000, but it does show that most large metropolitan cities have a smaller percentage of non-whites than might be expected. The variation in per—pupil aid within these states suggests that the variable has little bearing on the level of state aid. When the influence of non-whites on metropolitan relative ad— vantage is examined in Figure V—l4, the pattern is again inconclusive. Among states with even 12 percent non—whites in large metropolitan cities, only Alabama provides as much metropolitan per—pupil aid as it does to non—metropolitan school districts. New York again stands out among the states with less than 12 percent. Finally, when the non—white metropolitan population in large cities is related to the percent change in metropolitan per—pupil aid in Figure V—lS, the overall relationship is generaally weak (.062). The six States with the highest proportion of non—whites in large metropolitan cities are fairly similar in their amount of change. Florida shows the greatest increase, followed by New York, Arizona and, surPriSingly, New Jersey, considering its still low metropolitan per- pupil aid. When the six states with the largest percentage of non— Whites are factored out, the pattern is generally in the hypothesized direction. Still, the dispersion is so great and the cases so limited, that generalizations as to the influence of the special educational needs of urban non‘whit‘ must be considered Only If the level an‘ the socio-economic char itan populations. then in aid patterns. The I changes in state aid is seen in Figure V-l6, ti esized: i__e_., the stat are those with the lea: WOnajor exceptions tI set the negative relat Increases in p with relative advan hYiothesized gains in “‘1 111 needs of urban non-whites on any of the three measures of state aid must be considered only tentative and preliminary. If the level and distribution of state aid are a function of the socio-economic characteristics and educational needs of metropol- itan populations, then changes in the population should produce changes in aid patterns. The relationship between changes in population and changes in state aid is presented in Figures V-l6 and V—l7. As can be seen in Figure V—16, the overall pattern is opposite from that hypoth— esized: i.e., the states showing the greatest increase in population are those with the least increase in metropolitan per—pupil aid. The two major exceptions to this pattern, Arizona and Florida, do not off— set the negative relationship between the variables (£'= —.277). Increases in population also had a negligible effect on metro— politan relative advantage (£'= .114). As shown in Figure V—17, the hypothesized gains in relative advantage of Maryland, California, and New Jersey were more than offset by states such as New York, Idaho, Alabama, Nevada, and Arizona, and by the general lack of change in either population or metropolitan relative advantage by most of the remaining states. In sum, changes in population were not significantly associated with either the level of aid to metrOpolitan school dis— tricts, or with improvements in their position relative to non— metropolitan districts. The final social—demographic measure to be examined in relation to state aid patterns is the percent of students in public schools. AS can be seen in Table v—l, the variation across states is considerable, 112 TABLE V-l i Percent of Students Attending Public Schools 1 in 1967, Ranked 1. North Carolina 98.29 2. Utah 97.79 3. South Carolina 97.50 4. Georgia 97.39 5. Arkansas 97.29 6. Alabama 96.50 7. Tennessee 96.29 8. Nevada 96.20 9. West Virginia 95.79 10. Idaho 94.70 11. Florida 93.00 12. Oregon 92.79 13. Washington 92.59 14. Arizona 91.70 15. California 91.20 16. New Mexico 90.29 17. Kentucky 87.50 18. Michigan 85.00 19. Maryland 84.50 20. Louisiana 84.29 21. Connecticut 83.20 22. New Jersey 80.59 23. New York 78.29 24. Pennsylvania 77.39 25. Wisconsin 76.50 26. Rhode Island 73.89 W Mord deviation of 7. had on public school hose with the least d- u Northeast. The proper-t hprimarily a function do send their children concentrated in the me of the metropolitan po- will be the need for - in turn should increas aid. Figure V—18 sh dents in public schoo overall relationship i patterns are apparent hand corner of the fi show an increasing 1e in the percent of pug highest metropolitan a low percent of stu also have a higher 1 on their public schc 0f the large and 5p ‘7"771 113 ranging from 73.89 percent in Rhode Island to 98.2 percent in North Carolina in 1969. The mean for all states is 89.64 percent, with a standard deviation of 7.64 percent. The states with the greatest demand on public school facilities are generally in the South, while those with the least demand are mainly in the industrial states of the Northeast. The proportion of students attending non—public schools is primarily a function of the immigrant and foreign stock population, who send their children to parochial schools. These groups are largely concentrated in the metropolitan centers. The greater the proportion of the metropolitan population attending public schools, the greater will be the need for and demand upon public school facilities. This in turn should increase both the amount and the distribution of state aid. Figure V—18 shows the relationship between the percent of stu— dents in public Schools and metropolitan per—pupil aid. Although the overall relationship is not strong (r_= .101), several clear sub— patterns are apparent. When the southern states in the upper left~ hand corner of the figure are not considered, the remaining states show an increasing level of metropolitan per—pupil aid, with increases in the percent of pupils. New York, again the exception, has the highest metropolitan per—pupil aid of any state in the study, despite a low percent of students in public schools. Michigan and Pennsylvania also have a higher level of metropolitan per—pupil aid than the burden On their public school facilities would suggest, both perhaps because Of the large and specialized educational needs of their central city m the relati publicschoolsandthei figure v-19, the same .. correlation is stronger inthe hypothesized dir the southern states ar the hypothesis that me- of students attending a Political Variables an The third set parison with the metr- Certainly metropolita of the state's politi< to education. Three ' Expenditure levels in electoral turn-out, a revenues contributed iables and state aid section . Party Compet has been one of the American state poli "WI 114 populations. The remainder of the states give support to the basic hypothesis. When the relationship between the percent of students in public schools and the metropolitan relative advantage is examined in Figure V—l9, the same pattern is even more pronounced. The overall correlation is stronger (£_= .230), and the regression line is clearly in the hypothesized direction when the deviant cases of New York and the southern states are removed. The data, in sum, offer support for the hypothesis that metropolitan state aid is related to the percent of students attending public schools. Political Variables and State Aid The third set of explanatory variables to be examined in com— parison with the metropolitan power measure are political in nature. Certainly metropolitan legislative power is not the only characteristic of the state's political system that is likely to influence state aid to education. Three variables that have been shown to influence state eXpenditure levels in different functional areas are party competition, electoral turn—out, and ”local reliance, or the proportion of total revenues contributed by the state. The relationship between these var— iables and state aid to education will be examined in the following section. Party Competition: As noted in Chapter II, party competition has been one of the major explanatory concepts used in the study of American state politics. A number of different measures of party The average pe held by Democr The average pe- of representat' The percent of the House in w To the extent that sta expenditure, long—term of state aid to metrop Interpreting t State aid measures wil iables in the study 5‘ As shown in Table V-2 South Carolina at .96! range is greater, par pretation of the Inde above are considered modified one—party De Petitive; and .1000 1 Since none of the st: We would expect to s .6999 range than in greater than the .9C 115 competition have been developed. One of the most frequently used is the Ranney Index, which is based on the average of the following figures over a twenty—year period: 1. The average percent of the popular vote won by Demo— cratic gubernatorial candidates. 2. The average percent of seats in the state senate held by Democrats. 3. The average percent of seats in the state house of representatives held by Democrats. 4. The percent of all terms for Governor, Senator, and the House in which Democrats had control. To the extent that state aid to education is a "political" type of expenditure, long—term party competition should be related to the level of state aid to metropolitan school districts. Interpreting the relationship between the Ranney Index and state aid measures will be different from the other independent var— iables in the study since the Index is not cumulative or uni—dimensional. As shown in Table V—2, the values for states in the sample range from South Carolina at .9659 to Idaho at .3723.3 For all fifty states, the range is greater, particularly at the lower end of the scale. Inter— pretation of the Index is as follows: States with Scores of .9000 and above are considered one—part Democratic; .7000 through .8999 are mOdified one-party Democratic; .3000 through .6999 are two—party com— petitive; and .1000 through .2999 are modified one~part Republican. Since none of the states in the sample fall below .3000 on the scale, we would expect to see greater levels of aid in states in the .3000 to .6999 range than in the .7000 to .8999 range, which in turn should be greater than the .9000+ range. This in fact the case. The mean for 116 TABLE V—2 Ranney Index of Party Competition, Ranked 1. South Carolina .96590 2. Georgia .96290 3. Louisiana .96120 4. Alabama .95290 5. Arkansas .90950 6. Tennessee .86910 7. Florida .86880 8. North Carolina .86050 9. Kentucky .75370 l0. Maryland .74160 ll. New Mexico .71120 12. West Virginia .69980 13. Arizona .66040 14. Rhode Island .61310 15. Washington .57940 16. Nevada .57150 17. California .54140 18. Connecticut .53030 19. New Jersey .48610 20. Oregon .45850 21. Pennsylvania .44260 22. Michigan .41990 23. Utah .41350 24. New York .38490 25. Wisconsin .37980 26. Idaho .37230 W m it is 5294-69i ' 3150,50. As can he 53% M11 aid within each <1 clear pattern of highel the .5000 through 5999 When the relat: relative advantage is ‘ tern emerges. Two-par advantage (94.9) than (86.1) or the one-part cometitive states are category. The relations} Nita“ Per-pupil aid 1nFigures v-22 and v- 117 the two-party states is $312.10; for the modified one—party Democratic states it is $294.60; and for the one—party Democratic states it is $260.60. As can be seen in Figure V—20, however, the ranges of per— pupil aid within each of these categories is considerable, and no clear pattern of higher aid in the most competitive states (those in the .5000 through .5999 range) is apparent. When the relationship between the Ranney Index and metropolitan relative advantage is examined in Figure V—21, a somewhat similar pat- tern emerges. Two—party states have a higher metropolitan relative advantage (94.9) than either the modified one—party Democratic states (86.1) or the one—party Democratic states (86.4). But again, the most competitive states are not distinctive from others in the two-party category. The relationship between the Ranney Index and changes in metro— politan per—pupil aid and metropolitan relative advantage are presented in Figures V—22 and V—23. As can be seen, no distinctive pattern is present. The mean change in metropolitan per—pupil aid for both the modified one—party Democratic states (83.6 percent) and the one—party states (98.1 percent) is greater than for the two—party states (67.4 percent). The mean change in metropolitan relative advantage for modified one—party Democratic states is greater (252.6 percent) largely because of Florida than for either the one—party Democratic states (101.0 percent) or the competitive two—party states (176.1 percent). In sum, it can be said that the levels of state aid are posi— tively related to long—term party competition using the Ranney Index, .., .:.:‘«§ to A second, more is the relative strengi adopting the state aid culated by taking the crate in the 1967 legi legislative seats and for the twenty—six sta Unlike the Rar cmulative, or uni-dim the greater the level lations in determinim Figure V—24 p: tition and 1969 metro; notStrong (r = .295) 118 but that the distribution and patterns of change in aid have no ap— parent relationship to party competition. A second, more direct and useful measure of party competition is the relative strength of the two parties in the legislature actually adopting the state aid formulas being studied. This measure was cal— culated by taking the difference between the Republicans and the Demo- crats in the 1967 legislature as a proportion of the total number of legislative seats and subtracting this figure from one. The values for the twenty—six states in the sample are presented in Table V—3. Unlike the Ranney measure, the party competition measure has cumulative, or uni—dimensional values; that is, the higher the score, the greater the level of competition. This permits the use of corre— lations in determining the degree of association. Figure V—24 presents the relationship between 1967 part compe— tition and 1969 metropolitan per—pupil aid. The correlation, although not strong (£_= .295) is in the hypothesized direction. Although sev— eral patterns can be seen in the group of states at the bottom of the figure and the one beginning with New Jersey and extending upward to the right, the states in the two groups are not the distinctive groups seen in the Ranney Index. Utah, normally a competitive two—party state, in 1967 was a strongly one—party state, while Florida, Tennessee and Kentucky, all modified one—party Democratic states in the Ranney Index, were among the most competitive in 1967. The combination of high competition and low metropolitan relative advantage in these three states, plus Wisconsin, Nevada and Pennsylvania, points up the 119 TABLE V-3 1967 Party Competition, Ranked 1. California .97 2. Michigan .96 3 Pennsylvania .94 4. New York .93 5. Arizona .92 6. Nevada .91 7. Washington .87 8. Tennessee .87 9. Wisconsin .84 10. Kentucky .83 11. Idaho .80 12. New Mexico .76 13. Florida .74 14. Oregon .73 15. Connecticut .64 16. Rhode Island .62 17. West Virginia .61 18. New Jersey .51 19. Alabama .49 20. Maryland .36 21. North Carolina .35 22. Utah .32 23. South Carolina .27 24. Georgia .26 25. Louisiana .03 26. Arkansas .03 W mmwmi tive advantage in Fig‘“ strong (E: .321). Thi Among the n03.00mpetit is a clear pattern of Among a second set of Jersey and extending t creasing competition a third group, the most thpolitan relative cousin's .68 to New Y Interpreting that short-term chanc 120 role of other variables besides party competition in explaining state aid patterns. When party competition is plotted against metropolitan rela« tive advantage in Figure V—25, the relationship is_again moderately strong (5': .321). Three patterns are apparent in these figures. Among the non—competitive states at the bottom of the figure, there is a clear pattern of increasing metropolitan relative advantage. Among a second set of more competitive states, beginning with New Jersey and extending through Oregon, the pattern is also one of in— creasing competition and metropolitan relative advantage. Among the third group, the most competitive states, however, the variation in metropolitan relative advantage is considerable, ranging from Wis— consin's .68 to New York's 1.28. Interpreting these patterns is difficult. One explanation is that short—term changes in the general level of state party competition creates temporary alignments favorable to metropolitan school dis— tricts. A second explanation, already noted, is that party competition Operates on the state aid formula only in conjunction with other vari— ables and must therefore be considered a necessary but insufficient Pre-condition for a high metropolitan relative advantage. In sum, while there is some indication that the long—term level Of aid is influenced by party competition, and that 1967 party compe— tition had an impact on the amount and the distribution of aid in 1969, neither relationship is strong enough to warrant any firm generaliza— tions. Any support for the hypothesis that party competition ML... W“ Ike legislatures more? lire receptive to high level of participation turnout has been founc‘ penditures in other f1. should also be related the level of metropol.‘ 0f participation is t' Percentage of the pop 1962 and 1970.4 Fi 911 the Milbrath Index pa The most dist in the participation The former are all be .40. This differenct 121 stimulated state aid to metropolitan school districts must therefore be considered tentative and incomplete. Electoral Turnout: Just as party competition in theory should make legislatures more responsive to community educational needs and more receptive to high educational expenditures, so also does a high level of participation make them more constituent—oriented. Electoral turnout has been found to be positively associated with state aid ex— penditures in other functional areas. It follows that participation should also be related to the levels of metropolitan per—pupil aid and the level of metropolitan relative advantage. A frequently used level of participation is the Milbrath Index, which is based on the average percentage of the population voting for Senator and Governor between 1962 and 1970.4 Figures V—26 through v—29 plot the association between the Milbrath Index participation and the four measures of state aid. The most distinctive features of these tables is the difference in the participation patterns of the southern and non—southern states. The former are all below .30 on the scale, and the latter are all above .40. This difference has no clear relationship to any of the four state aid measures being examined. The spread in the aid figures among high levels of participation is just as great as for states with low levels of participation. Except for a slight tendency among non—southern states to have lower levels of support and less change than among the states with the higher levels of participation, no clear sub—patterns emerged from the data. The correlations between the Milbrath Index and the aid measures canoes in “Mont“ mt be that long-term mt, the distributici cation. > As with party ‘ J term and long-term pat“ tionship between the 9 aeasures of state aid- hilbraith Index (g = ' Metropolitan Relative participation, short-1 on state aid to educa' Proportion 0f \ political measure to fines contributed by situational finance 1' levels of aid and to 122 are all low: .Metropolitan per—Pupil Aid (.080), Metropolitan Relative Advantage (.190), changes in Metropolitan per—Pupil Aid (—.215), and changes in Metropolitan Relative Advantage (-.016). The conclusion must be that long—term levels of participation do not influence the amount, the distribution, or the amount of change in state aid to edu— cation. As with party competition, participation involves both short— term and long-term patterns. Figures V—3O and V—3l present the rela— tionship between the gubernatorial vote nearest to 1968 and the two measures of state aid. The relationship is no stronger than for the Milbraith Index (£_= —.152 for Metropolitan per—Pupil Aid and .074 for Metropolitan Relative Advantage). The conclusion must be again that participation, short—term or long—term, has little apparent influence On state aid to education. PrOportion of Revenues Contributed by the State: The final Political measure to be examined is the percentage of education rev— enues contributed by the state.5 States having a greater stake in educational finance in policy making can be expected to provide higher levels of aid and to be more responsive to specialized educational needs and changing conditions than states with a relatively small role relative to local school districts. Figure V—32 plots the relationship between the state percent of total public school revenues for 1969 and Metropolitan per—Pupil Aid. The overall relationship is relatively strong (£'= .339). The majority of states follow along a clearly defined regression line in the hy— pothesized direction. still, the range of Metropolitan per-Pupil Aid is considerable. This own both seem to be. both providing considei ing their prOPortional pupil aid relative to ‘ nantly southern states Carolina. The relationsl". tures and the Metropol lithough not as strong in the hypothesized d advantage rather than Georgia being the onl ltan aid relative to '71 123 at different levels of state involvement, particularly among the states providing over 60 percent of the total educational revenues, is considerable. This variation is difficult to explain without a more detailed analysis of local expenditure patterns. New York and Oregon both seem to be the states being most deviant from the pattern, both providing considerably more aid than would be expected consider— ing their preportional contribution. States with low levels of per— pupil aid relative to their proportional contribution are predomi— nantly southern states——Georgia, Alabama, Kentucky, Arkansas, North Carolina. The relationship between the state percent of total expendi— tures and the Metropolitan Relative Advantage is seen in Figure V—33. Although not as strong (£ = .128), the relationship is still clear and in the hypothesized direction. Deviant states in this case tend to advantage rather than disadvantage metropolitan school districts, Georgia being the only state significantly disadvantaged in metropol— itan aid relative to its proportional contribution. When the state contribution is plotted against changes in Met— r0Politan per—Pupil Aid and Metropolitan Relative Advantage, no clear Pattern is viSible. The relationship between the first of these sets Of variables shown in Figure V—34 is weak (£_= —.029). AlthOugh Florida showed the greatest increase in aid, the remainder of the states with a high involvement in educational financing showed relatively little change. The greatest change occurred in states contributing between 40 and 50 percent of the total educational revenues and also among the three states contributing the least. When Florida‘s high degree of change (eXPlained earl and Nevada, Connecticu plained respectively i established progressi‘ ysis, the pattern is < in the state contribu‘ When state pe in Metropolitan Relat relationship in relat the distribution of s in the financing of < (.027) and generalize is clear. In combin state aid, it sugges financing are more g Ones in msponse to The failure of many raise new revenues that eduCation is n 124 change (explained earlier in terms of its tradition of malapportionment) and Nevada, Connecticut and California's low amount of change (ex— plained respectively in terms of local wealth, localism and already established progressive state aid policies) are removed from the anal— ysis, the pattern is clearly toward decreasing changes with increases in the state contribution. When state percent of the total aid is plotted against changes in Metropolitan Relative Advantage in Figure V—35, it is seen that the relationship in relatively few states showing significant changes in the distribution of state aid did so in direct proportion to their role in the financing of education. While the overall correlation is low (.027) and generalizations from five or six states risky, the pattern is clear. In combination with the pattern for changes in the amount of state aid, it suggests that states with a strong stake in educational financing are more prepared to redistribute revenues than to raise new Ones in response to the growing educational needs of metropolitan areas. The failure of many highly urbanized and industrialized states to either raise new revenues or to redistribute existing revenues also suggests that education is not high on the priority list of state legislatures. The Explanatory Power Of the Control Variables The regression statistics for the variables discussed in Chapter V are presented in Tables V—4 through v—7. As can be seen in Table V—4, several of the variables explain a significant portion of the variance of 1969 Metropolitan per—Pupil Aid. The tax burden Regressior l‘. Economic Variables PERSINC BURDEN Social/Demographic Va “K lETPOP LGCTYPOP NOWHITE PCTPUB Political Variables WNEY PTYCOMP MILBRATH GOV STPCT \ 125 TABLE V—4 Regression Statistics for Control Variables and 1969 Metropolitan Per Pupil Aid b F P>F r2 Economic Variables PERSINC .013 .059 .806 .002 BURDEN 68.99 6.24 .019 .206 Social/Demographic Variables METPOP .008 2.23 1.45 .085 LGCTYPOP .024 5.65 .024 .190 NONWHITE —2.40 1.82 .187 .071 PCTPUB 1.52 .250 .627 .010 Political Variables RANNEY —.002 .050 .819 .002 PTYCOMP 116.03 2.28 .140 .087 MILBRATH —.O50 .155 .699 .006 GOV —.001 .568 .536 .023 STPCT .309 3.13 .086 .115 Economic Variables PERSNC BURDEN 126 TABLE V—5 Regression Statistics for Control Variables and 1969 Metropolitan Relative Advantage b F P>F r2 Economic Variables PERSINC .003 .164 .691 .007 BURDEN 6.70 2.62 .115 .098 SOCial/Demographic Variables METPOP —.O37 .006 .795 .003 LGCTYPOP .075 .226 .643 .009 NONWHITE -.397 2.59 .116 .097 PCTPUB .484 1.34 .257 .053 Political Variables RANNEY —.001 1.48 .233 .058 PTYCOMP 17.77 2.75 .106 .103 MILBRATH .017 .901 .646 .036 GOV .000 .133 .719 .006 STPCT .016 .397 .541 .016 MNWEITE CGINC CGPOP 127 TABLE V-6 Regression Statistics for Control Variables and Percent Increase in Metropolitan Per Pupil Aid, 1962—1969 b F P > F r2 SOcial/Demographic LGCTYPOP .032 .800 .616 .032 NONWHITE 1.86 .094 .759 .003 CGINC .823 1.73 .197 .067 ; CGPOP —.677 1.13 .298 .044 Political Variables RANNEY —.002 .005 .942 .000 MILBRATH —.445 1.162 .292 .046 W SWial/Demogramic [MYPOP WHITE CGINC CGPOP 128 TABLE V-7 Regression Statistics for Control Variables and Percent Increase in Metropolitan Relative Advantage, 1962-1969 b F P > F r2 Social/Demographic LGCTYPOP .056 9.20 .006 .277 NONWHITE CGINC .131 .122 .729 .005 CGPOP .212 .318 .584 .013 Political Variables RANNEY —.008 .164 .691 .007 MILBRATH —.020 .007 .934 .000 M is at least PhD-113:3]-1y nificance level of th sive. Among the 5°Ci itan city population of the variance in th significant at .024. ‘ of educational revenu 10 percent in the val When the impa Relative Advantage is htition is the only plain even 10 percen‘ the Percent non-whit 129 accounts for 20.6 percent of the variance, although it should be noted that the limited range of the variable from 3.2 percent to 6.5 percent is at least partially responsible for the high value.6 The high sig— nificance level of the variable, however (.019), is clear and impres— sive. Among the secial—demographic variables, the 1960 large metropol— itan city population is clearly the most useful, explaining 19 percent of the variance in the Metropolitan per-Pupil Aid and being highly significant at .024. Among the political variables the state percent of educational revenues is the only one able to account for more than 10 percent in the variance of Metropolitan per—Pupil Aid. When the impact of the control variables on 1969 Metropolitan Relative Advantage is examined in Table V-5, it is seen that party com— Petition is the only one of the eleven control variables able to ex- Plain even 10 percent of the variance. The tax burden variable and the percent non—white in large metropolitan cities also explain approx- imately 10 percent. Again, no single group of variables explains sig- nificantly more of the variance than the others. When the four control variables discussed in the chapter are regressed against changes in Metropolitan per—Pupil Aid between 1962 and 1969, as shown in Table V—6, it is seen that they have uniformly Poor explanatory power. The Milbrath Participation Index is the most Powerful variable and it explains only 4—1/2 percent of the variance. Increases in the relative advantage of metropolitan school dis— tricts between 1961 and 1969, on the other hand, are largely explained bY the population in metropolitan areas living in large cities. This variable alone accounts for almost 30 percent of the variance in changes napteruhenitiscon ofnriation. it accoi The two depem in enlaining are cha: 1962 and 1969, and 19 apportionment variabl Measure than any of t ings challenge the c< Mt that show appor1 Whey choices. Before any f cance 0f the present ex“Milton! Power of tion rather than jus 130 of metrOpolitan relative advantage, and, as will be seen in the next chapter when it is combined with changes in the inverted coefficient of variation, it accounts for almost half of the total variance. The two dependent variables that the model is most successful in explaining are changes in Metropolitan Relative Advantage between 1962 and 1969, and 1969 Metropolitan per-Pupil Aid. In both cases, an apportionment variable accounts for more of the variation in the aid measure than any of the other independent variables tested. Such find- ings challenge the conclusions of most of the literature on apportion— nent that show apportionment systems to have no influence on state policy choices. Before any final conclusions can be drawn as to the signifi— cance of the present findings, it is important to examine the overall explanatory power of the model, using all of the variables in combina- tion rather than just one at a time. This will be done in Chapter VI. l'rhe same has is therefore omitted ‘ lietmpolitan Relative 2These figure of the States. 1970 ' M 1971), p. 76. 3These value: Ranney figures throw worth and presented . Conunities (New Yor? 4These value f0“mi in his article tics in the American 1972) . 5 . This measur It is treated as a 3 Variable because the hi?“ range in th my hi where th 9h betc CHAPTER V Notes 1The same basic relationship was found for 1960 and 1962 and is therefore omitted here, as is the figure for personal income and Metropolitan Relative Advantage. 2These figures are found in Research Report 1970—R1, Rankings of the States, 1970 (Washington, D.C.: National Education Association, 1971), p. 76. 3 . . . These values were drawn from an updating of the original Ranney figures through 1968 by Hugh L. LeBlanc and D. Trudeau Allens- worth and presented in their book, The Politics of States and Urban Communities (New York: Harper and Row, 1970). 4 . . These values are taken from an updated verSion of the index found in his article in Herbert Jacob and Kenneth Vines (eds.), Poli— tics in the American States, Second Edition (Boston: Little, Brown, 1972). 5 . . This measure is taken from Ranking of the States, 1970, p. 45. It is treated as a political, rather than as an economic or demographic, variable because the legislature determines from year to year what pro— pOrtion of educational revenues will come from the state. It is treated as an independent rather than a dependent variable because while the state role is the result of a variety of other factors, in the present context it has potentially important implications for the level and distribution of per—pupil aid. 6 . . . . Also note the high beta coeffrcrent this produces. Since the betas in a multiple regression are calculated by taking a ratio of the cross—products, such as b12 ' 0°13) (1023) = b 1 — (1023) (b32) 12.3 a small range in the denominator will produce a high beta value. A similarly high beta value is seen with the party competition variable, Where the values range from .00 to .99. 131 Having examin aid measures and a se Chapters III through on these findings ané P°ssible variance in “111 be to combine ti Vious chilP’cers into CHAPTER VI THE CUMULATIVE EXPLANATORY POWER OF THE MODEL Having examined the bivariate relationships between the state aid measures and a series of apportionment and control variables in Chapters III through V, we are now in a position to draw selectively on these findings and develop a model that will explain the maximum possible variance in each of the state aid measures. The procedure will be to combine the variables discussed individually in the pre- vious chapters into four separate multiple regression models, one for each of the dependent variables being examined. Three tables will be presented for each model. The first pre— sents the R2 value for each of the variables in the model and the re— gression statistics for the variable with the highest R . The second table presents the results of a technique called the “backward elimi— nation procedure," a form of stepwise multiple regression which begins with all variables included in the regression equation. It then elim— inates. one at a time, the variable in the equation contributing least to the variance in the dependent variable, until all of the variables in the equation have a significance level of .10 or less. The final table will present the regression statistics for each of the variables in the model. 132 effects of multicolini ginning Patterns 0 in State Aid Table VI—l p1 Efficients for the va seen, most of the co: nriance in the depei Variables is largely other independent va are underlined and v f'llldings . 133 In addition to the three tables for each of the four dependent variables, a correlation matrix will be presented for the cross- sectional and longitudinal models in order to examine the possible effects of multicolinearity on the findings. Explaining Patterns of Change in State Aid Table VI-l presents the simple product—moment correlation co- efficients for the variables included in the change model. As can be seen, most of the correlations are relatively low, which means that variance in the dependent variable explained by any of the independent variables is largely unique and not the result of interaction with other independent variables. All correlations in the matrix above .40 are underlined and will be duly noted in the interpretation of the findings. Changes in Metropolitan per—Pupil Aid——Table VI—2——presents the coefficients of multiple determination for each of the variables in the model. The "best" variable, Change in per—capita personal income between 1958 and 1968 (CGINC), is seen to explain only 6.7 per- cent of the variance, and to have a relatively low level of statistical Significance (P = .198). Only five of the nine variables explain more than one percent of the variance, and the two that explain the most—— Change in income and the Milbrath Index——are highly correlated at —.662. When the results of the backward elimination procedure are ex— amined in Table VI-3, it is seen that the total variance explained by the model is 27.53 percent, and that every variable in the model is smmAHS mDmaum gnu dmmeamo ma. Mvflfi ”urfluvw uh 134 mHe. oHo.| HeH.u who. eHH. mam. Hso.| sso.- mam. ome. amsomo eam.- moa.u mmo. Ham. man. mmm. $00.: mmo.a smo. amazomo nmm.- mmm.u ooo. smn. mmm.u mea. mmw. omm.s meammaHz wmo. mmm. mwm.| Hma. one. mmm.n aha. momsom Nee.» 500.: Mam. asm.u HHN.I mam. meHmsuzoz moo.- one.. new. mom. mmm.u momwmo swo.| oeo. «mm. emo.n momsnoma mmm.u mam.- omo.u ozHomo moo. oao. zmomam HwH. UZHWMMQ >oHumo mmzwmo fimmzwmo mBflmeHE mDmBom MBHEBIZOZ momwmo momwaowfl UZHUEU membm UZHWMmm >UHOmU Uflm wnmvw SH meQMHHm> Mom xHHpmz GOHHMHwHHOU HIH> mamflB I. . JwODt Zn “NJOC—1<> HIV-ul- IOmIIILIIK 11¢ dill.- ..OL GladdxohJI—A .- DIE! .- Ildfllififl> OEIIEIfiI'IH Lou IOdGIIiIECJUQ Oflndfldll 9° NflfllflUMWMINU N I H> Idl‘fl 135 I 03:30qu 3.333 30.10 aillwll il!~d~m~m~0 no I “.215, I ) Emmmwwsdnnln 1 iii, zl.0 . 3203.”. u. A 093 mg; n. 5:53. 25: mason no :3 “.0 , @0103 .y Ammo: w>om< th «Du h—u LO mo_hm—h to m_w>4 man’s z a talus: J Lu 1. wWJl(-¢() Wldaomlfl ‘d< Adult LIE Iflfld °§°hvl I fill UHAIdLI> anlvconln hon 055000..“ Iouvlfldlwun VHMD. In \ n I H> lil(k I iii]! 2 136 0 mac mbauu uz—nu aoauu >uuuu mambo: Zuaxzo I»<¢na—t uh~zxzoz l— mm4n1_¢<> was .39..“ non scoauonnzao: cu aha-£0 0.31:3, 0:06:0non non 0.56000...— naalilzu uncixoqn lIIl|lIIllIlIIIIIlllllllllllllllllllllll . . M I H> udl<fl w1130mc¢ Juno: z— an132 _ I tistieal significance findings. The conch budgetary increases of interacting force tical explanation, 0 developing the model] the concluding chapt Changes in l- siderably more effea advantage between 1! largely due to two Variation (CGICV) I wlitan cities in 1 percent of the vari districts relative Table VI—6 ables significant 2 P°pulation, the pe cities, and the ed taut variables in and are influence In Table ‘ but two majOr var 137 eventually eliminated, none being significant at even the .10 level. Finally, the low R2 of the model, combined with its low level of sta— tistical significance, as shown in Table VI—4, reinforces the previous findings. The conclusion to be drawn from these data is that either budgetary increases in state aid are the result of a complex network of interacting forces that is not amenable to quantitative and statis— tical explanation, or that several key variables have been ignored in developing the model. Some comments along these lines will be made in the concluding chapter. Changes in Metr0politan Relative Advantage: The model is con- siderably more effective in explaining changes in metropolitan relative advantage between 1962 and 1969. As shown in Table VI—S, this is largely due to two variables: Change in the Inverted Coefficient of Variation (CGICV), and the size of the population living in large metro— politan cities in 1960 (LGCTYPOP). Together they explain almost 48 percent of the variance in the aid increases in metropolitan school districts relative to non—metropolitan districts. Table VI—6 shows that these two variables are the only vari— ables significant at the .10 level. Percent increases in the state Population, the percent of non—whites living in large metropolitan cities, and the educational tax burden are the next three most impor— tant variables in the model, although they contribute only marginally and are influenced to some extent by multicollinearity problems. In Table VI—7, the relatively high significance levels of all but two major variables is shown, as well as the high overall E: m. undo: was. 2: .353 one uzuxu>c¢e§ 523cm; 12:02: u...» >0 ozacn duo: 392:; o .53. 1, , u u an... u >33 oz. can 5.. on zucu _. 2.533 z my :20: 30...:qu neenmru 0 w 1 .53: Juan: z— vinegar; .I.mlu..m.....m.mn...n .- lflwlmmozaflll duh-f- ..Ol Ilddflbflckflll I.“ Ola-nu 6% IIHIIiLl> OGOQEIQIBEN a. HIVOI IOGIIIEIOI vd< 0'90 IOHIIIFIII flail LEI thfllp—lfinlfl C}iUIfi-JIID hen Inddld ‘ I N> Ian‘k 138 ; 530310- 02538... mafia o . - moonwwo- flax?” . «swat cmmmnmamén :33an 33.0 . team 0- \ 3 . . 33%??? 333202 . $333.0- oooomiTS 22.0 235.0- Nmm280.0 92.00 oemmobbmé 02330.0 33.0 Nemalx . . 1- «0333.0 . 902,504 37.7.2.0 03032.0 903.0 «2.3.0 02330.0- .25“: 3323.0- 02.7%.}... 33.0 mmcmwdfl? it»:.-!)J~a$\..3.afl. . 2032. 3232.0 emcoizéi gema ONE: .0 9533.3- use»: iii: mmamozmél 23333: 300.0 games“- mettle“; . 02.00 . . . 0. 3:: o 38;: a 5mm 0 are. mmomooemwmwei .,\5-tz, 0 new 0 mum 0pm _e_ A moan ouzuox «on e mmnnwayml\« .\ :ril.r \i. \ .illiill i‘i‘l Illi \ I 32.5 0.83.0 3.3027303 33.0 2310 «0333.303 a. i\ :Laimwat i $3.0 2.310 oifiiméfifi 23.0 :03... 2528.82» a nelxzzcz 33.0 ~oo¢m.0 n3n-0o$0~eo 23.0 $3.10 33:34.22. H >050 .. 2.3.0 00201 Neoaomeodfimg 33.0 Fig.0 5:23.333 H .6353 003.0 303.0 m-$o~1~mm3 33.0 2.13.0 3333.333 H - 1.0.2212 i. - 33.0 2.30.0 2333.302 52.0 313.0 03302.02: a $950 ememé 320.0 toozewxzfim 33.0 038; 20~$~m£-¢2 A $55.. iii--- 300.0 3570 3.523.303.» 8.1.10 «:00 .0 $253.33.: H 02.00 Km .0 ~38: 3002005303 336 02.3; 03033.313 a . a A 005. ":3; H. mm 455:: u. A 09: m3; H. 3 #25083 a0 0:000 |i\ 330an 6:83 3 4:3 $53200 0~¢0m~3$m¢o2 ofiioomammfimw 2 - SSE -- 3335.0 0370 32.0.0 2333.2:2 caeoopemagooo 0 203.35%. .>.0 mingle a A 09: m3; a 3:33 24m: 3503 no :3 “.0 3.33 l Juno: w>om< wxh mow h—u no wu_hmukdhm oz< . mbzw—u—mmwou zo_wmwmowx . medh mu: <~¢<> no m_w>4ozmtu mz<30mlm tat~xm ozvow amooz w4m<_m<> 0 L3? we: 2 4.0.00: :03 m5 JIIIIIIIIIII uz—uu acmou >u~oo uZ~mama manhum zwomam Ihhuu4 m-¢mmhwuo o Juno: ammo: z— mm4m<~¢<> u¢490mem z. «mmzaz duvet :OMDIOLHOM od< awn: how nouvuaalom sea-logbox can Luzon hhoocccanxm o>wvcusaau v I H> mun ocooconovcH wcwlcuct < I‘I\|‘I| . .u I nnflhflodauo . . fl IAMOOIEIEI 5 «warms. . i.e.,..m. E muaom_¢H> uuqaamln « OUCIIU Gd dl‘fi‘ IOdlIOhIIl OIIOEI>V¢ O>iOIHOK IIDddOIDLflOX C MD IUCIquHUIQU Ildnlqhi> atlvcooouau Lou Iciclldlhlooa Idndvdlt n I H> “J.(h 139 nom>»004 - .- - i x \ . $33.0 afizammflq «83301.0 300.0 .0201“ wflfiSwJZ 232 - u t 1 50:3 II l-$3<>..m 05 0 03 SW 3.. A 000.. - 0.x...mxzmou+ - $3: 0 . . . x 00350.. I 030.0 ~53... $0053.33? 030.0 3310 «wahmmm .ovmmm l 01 l . \ . - . 001000 u. A 001... 03.; H. mm 025: 0 A 0010 03,; u. 3. #0500000 l-\. -010 - 33021§302 mN . , wifefiwfificm . $0230.90? 300.002.2030 «N ELSE [S‘is u S..~o.m3 3003:.0 030.0 ~53... «3033.003; ¢Nm0>mmm0000mm .0 3:? .izowmmmmm‘mV- 1\ l. .>.u . 05:3,}. u. A 005 03; 0 012.3 2:: $0403 00 :00 “.0 I} a xemmmoom. l . mhzw~u_mmmou zu_mww1uma . w4m JO mmm>412< . . amooz w>Om< ml? «Du w—u no wu—hmuh4Hw Qz< . 0001001. \(1\\ . a ~.§~3.llali\l1l. ,. Ejfiw oidllll-l)1l|l(.[1 . a In? maddd sNQManaa a ‘1 . . la‘!‘ . . 0230 3130.306 :11 -. {I 004004? zazaodo‘IWIJII; 3-.-. u and? I . . 0000: 20 $0015); $203.; .. ,5 $m§z . auto: :0«IIOLM0¢ 0.935;“: 0.5093: :nuamonofiuo: c.“ 093:0 cw I... noun—€23, oaovuomvvcH Lou :OHoauaauouva 0.3.31.2 mo uuaofliumuoo nII||IllI|Ill|ll|Il|llllllllllllllllllllllll m I H> mdnU-QU Q’s-“fit ZHDIDQ E‘Ifldnl Whats-E! LOttgd 9n§a~n§h0° ‘. 806° Ul~ Cut 0 Sb. , 4 Q1 4. (I. I q .2 [at . Q ~ 0 Juan! WI‘R—UWI! 2h IND-‘32 4W8! Z~ nwdfliulih aiflaaHQ: :1 .‘II-iu a‘afll‘r“ -‘flOI MOI '919'flilflfll ‘LIlv-D: |\|‘|\| -Hflldll} OEOVEUEOE EON Iii-‘00:.- O I H} lil‘h \ \ .140 {IJIIIIJIr{I!;:I It!!! r 2: I; , :“‘ ~p-omc¢. com—.33.,“ 3328.0 33.0 331w? : ‘ 53:30.0 >38 «mNmmmao.o o_oo.o m¢¢oo.n m~>~ommo.o 1 50¢»mmmJ u : ‘, I) owm¢mmoo.3n 75: «MA A mean pude: «am y . lau34w>‘d,‘-4 \: ‘5 mumaow‘, 3005 o: . « 1 oaoo.o mommm.mw ”ammoh¢~.oma~o~ m¢oo.o e¢-c.o~ pfimmo~¢~.om-o~ g ‘1- ‘-‘ »u_au u A no mm~omm¢TI~Em 030.0 733.: q~nm~mmn.0¢3mm d >3in!“ i1 x xa m: < 4 > u mm 4<_»¢ u mm 4«_hzm:omm mo magnum 1. . «I mfloemmma.¢~o¢o- m~ Japoh owpumaxou Inf 2323.. 53an 1323338.. 3 «9:5 0975.00 . . omep¢ooc.o oooo.o noamm.a~ Ammoooep.¢~ooo~ «caps—ac.oe~mom m zoummwxowz .>.u . u 4 $323 a 2...: u £333 23: 35:3 no 13.. mo muxaom dwaoz wy0m< wzh «on. but no wu~hmub no m~m>u(z< mono >9~QU Acawu >u—Qu z D m 9:: 202 a0¢>hga {Ill woomm uem so 0 aoauu >u—uu uz—mxwt 2wa¢3m wb—zxzcz «BAP—um... {If . ENG mm¢¢mno h uz—Qu 59.50 >U~Qu Uz~m¢wn zwoSn 9:13:02 ADAP—UUJ 92:; «can; >U~uu oz—wxwm 2mo~5a z><¢o4~x mk—ZZZDZ tattoo... emf—‘an-o ,o 2:2. 30.35 >u~uu Uz~m¢w¢ mambut zmoxa thime—t up 13:02 aoa>h m~o¢~m¢mulq o {flit , . Jun—at .58! z~ mwau<—¢<> . m¢<30mlx z" ¢wa832 I!!! omlvcl>v< OER-do: :cuanonouao: :w Gun-£0 c.3235» vuovconon Lou 9530095 :ouuuualwsm vn-onnn o I H> Hausa 141 mum.oo.&.w.. - If Illllll. IEIIINWMMMch.OI WWWMWMWMHQ.II., y- l‘ufl0mew . 1 .. :I.. :0mWoo.Q «omoqumuo mankum I! $23070. 0333 .o- 35.0 2310- «3038.0- 553:: ill - 0002.03.01 0330 m.0 33.0 0003.0- . .. 0.32310- 0280 00323.0- 0300 1.0 300.0 3%..0- 20330.0- 02:33 . - 03mm??? :omtmmmEo 3.3.0 .. 303.0... y. $333.31 59:5 $0203.? 2210 f. 33.0 . $23.0. 5330...? 2.11202 3,123.0 333%.0 33.0 303.0- .mnémm:.o- .0000 1.23.3.0 2.03 .0 «30.0 03.36 mommyooé :me . If: «No a 030.0 303$ « 2::umlflwwm0 w. 20:53 33; 0 20052.0? 2:: . 1/ Em I \4...:. . . 0 «gm Em 2.. A 093 enmuoz mom k $3; 0. .1 . 0353 . mo¢o.o ~¢ i-.. 1 .:W 1 1 «00.0 . 030.0 320.0 wommomfl.03 3.3.0 3.0.0.0 3333.03 0 gawk . 1-- . $3.0 022.0 03203.0: 2:.0 “$3.0 $323.0: 0 5, .. . 1520:: $00.0 $20.0 oggoiwA? 001.5 022.0 30330.8: 0 0280 a 1 $3.0 0003B: ¢~3mooo.mmm 2.1.0 03010 00333.33 H 023%: 1.3.0 NZZS m-~o~A.iof $36 3&6 :233103 H . 5950 33.0 27.70 0 3900.32. $30.0 22.10 3332.2: 0 - wtxxzoz 33.0 03%; 30033.28 33.0 $33.0 832.36%? 0 $0.60 - 33.0 132.0 3200:3503 38.0 30:; 2331.328 0 I >800 M. 00200: 333 $00.0 $01.0 «22.210022 0 Q0010”: . A no: m3: H. mm #25:: u. A 005. w3<> u mm 04:200me “.0 . , @853 floggifiog: 2 0mm» 85:10.0. E . 003002.302 mmwoompméofifi 3 .. GEE..- a 330.30 0 . .. . ~0¢~3m 0 .53 0 oimflé 30:30:32 Cméifiowflmo o 2053501 .>.... I 5:53. a u. A 00E 03; 0 $2000 25: mm .223. no 23 “.0 0350M . y Amooz w>om< wzk mam »— uo munhm_k no m~m>40¢a:_ wudoamlm z3t~xm DZDOm onoz w4m<~m<> o .300. 05 2.30% mammdumf. ozuou aoaou >uwwu uzummma mzapua Zuozam Ihhu04 0N0¢Nm¢mo01 o Juno: w1 Hove: :oaunonuom oumvcu>v< o>wqu0¢ :auaaononvo: a“ owcmgo :fi moannfinu> acmucoaovcH . :« Lou mowvmwuuum :OHHIOLMOK vca L030; shovucaanxm U>Huuasadu n I H> mun00 mEOOWEm mDmBUm WBHEBZOZ MEUOABUm momMBUOQ ZWQMDm m mzm proz EOflmmmHmmm pad wuwpm Nemf mmm.. mmm. NMK mmf wmo. omo. Ham. flaw. ©0H.I mom. va.I mvm. ohm. @mv. H®O.I vom.l mbm. mmm.| hoo.l MHM. vmv. woo. who. vmo. me. mhm.l mmm. ovo. #:wfidoapaommmmmlpwom aw moHQMflHm> Mom NHHHME coaumamuuou mlH> mam¢fi 00H. mmw. wvo.l HvH. 1 Haw. Hmm.l CNN.I I mmn. wmm. mow. flmmz EUmBm >00 mEOUNBm mDmBUm MEHEBZOZ NBUOHBUm momNBUUQ Zmnmpm m MBA tax bu: itan c: are ex explai mil 23.9 1 for 0 Chang ing Lhes met] deg} The it; CO‘ 145 tax burden (.469), percent of the pOpulation living in large metropol— itan cities (.733), and 1967 party competition (.411). When the R2 for each of the independent variables in the model are examined, in Table VI—9, it is seen that five different variables explain more than ten percent of the variance in metropolitan per— pupil aid for 1969. The best single measure is Power 3, which explains 23.9 percent, and tax burden, which explains slightly less. Aid levels for one year are clearly more amenable to statistical explanation than changes in aid over time. When all the variables are included in the model, it explains 67.2 percent of the variance in metropolitan per—pupil aid. When the variables not significant at the .10 level are removed, the five remain— ing variables explain 62.4 percent. As shown in Table VI-8, two of these five, governor vote nearest l968 and percent non—white in large netropolitan cities, are correlated at —.628, indicating a significant degree of common variance between them. Three of the variables in the model (PWR 3, GOV, and STPCT) are "political" and two (LGCTYPOP and NONWHITE) are social-demographic. The pOsitive association of PWR 3, STPCT, and LGCTYPOP with metropol— itan per—pupil aid indicates that the educational needs of large cities, combined with a high stake in educational financing, and a metropolitan legislative delegation that is both powerful and demographically cohe— sive, are all essential for a high level of state assistance to metro— Politan school districts. 146 .l-III .I IIIIII II; .2 .I IIIIIIIIIII..33$... - 023.306 - .-mosque-IIII-III|:§IW.:.~- .III. . mdmommfé II I 2.5 . I. .. : .. I. .iIiIiIIIlII; I III; I I . . -wmomfmhaa— 24m: \I‘il'i IIIII I I mania-mem- , - m Sm Sm . EI mIm-QFIII-li emf-Elea-om. t. inner; o. - $3.3 ii [I III'II. .I : -. I.| I... I II I -II-I -I-I III I I IIIIII II I i- I Iii IIIIIIII|II| I-z: I .I - I -. I . I I- I 33.0 imam; “nonwofifiw-m-i. . 0085- imie I .Idmwmmpmdwo: A $2.. III-‘IIIII‘IIIII I3 I, I- , - m A coin w34<> m I !lIMM.AWwdemlli IiIII-u M 001: w24<> u mm 4om< wt» «Du rum to mu h r oz< . w ZN—u m o 20 mmmxmwa . mdmhh MM2<—¢<> no m~m>4<2< wxzowuoxa hzwxm>ognln wx u .hmwm. wzh w— dwoox u>om< wtk may... ommwaol-Iqéi - .. I . . 595a 8332... a . II. I I- . 8:33 2.25 $.10 a j I I 5.2m 00322.0 a Iii!i :l-I - I - - 5.8:“. Comcgoé a iiii:.i --I--I.IIIII.. , I . I - I. >8 30220.9 a III-I - IIIIII. ..... I II I, I I - EEC; 33886 a _ doc: . EI- - I! . I.. $8215 3.355; magma-Ia . E. 5552 IIIIIIIIIII , .. Hove: .313-53. 3< 3%... tom :uuaaoawnao: 5 I-.II I-III . 33:5; 33:32:: you sewage-noose 333:: we 3:33:80 II-III . I -, o I H> mdnsuon «posum~m.CI Immowwmoo.o III mqoo.o ssome.nI o_oom~oc.cI >oo o~omohnm.oI cmmmmnso.a asco.o «sumo.~I saoessoa.oI m»_:xzoz mmaoo~n¢.o IIII :i.wmwmmsmnnm. I I Iomoo.o I sneasnw- I I.:-I s~qnmnom.o samba moons—em.o maomsmmc.c a¢mo.o q~sn~.~ smma¢~Ms.o max; I-I- I III II mongoomcéfi 23: mw34<> a new a «an ohm _h_ A moms guano: «on a mmsi<> a wuxsom mmeno ohuooun wonmu~oo.n~msa mmoo.o oe—oc.m aesom~so.nhnm n aca>huun poo.” anon .m qo-amn~.~o¢ma m¢~o.o ~o~op.m wea~o-o.~s~mm a >oo oooo.o o-on.¢ ~nmmmoma.¢~onn noco.o ocopp.¢ posesaNc.ooms~ IwI m~_:xzoz oeno.o p~¢n.a omosaoon.meo~n o¢~0.o conco.~ sooaeohm.nn.n¢ a euasm ¢~noo n aomm¢~nn.mncom -oo.o n_rn»n- amon~oo..oaos~ . auxa mouse“ to mm 4<_h2m30wm magi) u mm J l u A 901s . .pno-oa.o~hw Fae—ouco.nmc-u m~ goaum u onomu.o~ . ona~co~o.o “goo o neono.o mtocuon~.noeoc w-¢mman.o~nnc~ m. zo_mmuguw¢ . mu¢DOm a o>oU . w34<> w <30 I: u A some JNQOI w>om< wxh «Du h—u mo wU—bw~n no m~m>4 m1» 4w>w4 wuz<1 JwDD! m>om< wxp z screw >oo wt_:zzoz aoa>euaa maxi :ma~¢¢~s.e m III I euasm >oa snaps; wt_xzzoz aoa>tuoa zmoaam mega am¢ooooo.o a IIIIIIIIIIIIIII-IIIIII- I . . I roman >oo axooyea magnum mh_:zzoz a2a>puoa Zuamzo mama sneeoo»o.o a IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII .- boawm >oo azou>na mambo; w»_:xzoz aoa>euos >tuonhua zmoasa mega «oshospo.o a I Jill]- duo: ii . I I .II I 38: 2. 3.325; _ ELL-Er . ua< «ease you :uvaaonouaoz IIIIIIIIIII .IIIIIIIIII f oanuahu> 9:01:23: you oksvouofi— comalnmadam “Elisa-n I. - on I H> msmFMQAHQm I.IIIIawewwmmd.oI anwsamss.— aaw~.o - nw¢em.oI mnsowmnn.oI ¢nmhm~oo.o ommv00-.mm Owns.o «mwam.c:IilII.II..I.Ivaoowm:.wwI II. .IIImmmmmWM . .IIlIIIIIIIIIIIIIdmmMaMdAIAWIIIIIIIImHmmndmN«anI I. II.Ifiwmw.QI.-IIIIIIIIII.Imwuw¢qw meomemOHsfl cannon . monmom-.o oosssph~.m smoo.o laumv.o ,.-wsneminw mI.II touxtuqa..III.I IIIIIIIIIIIIIIIIdenu«sadism :.II.I mwmaoOAo.a - losoqm, I -I II oawsm.s . mammWMMqu >ca -o~m¢m~.OI chemoeoo.o com—.0 mowmm.»I um< n . H IIIIII.I .IIIIIII IIIIIIIIIIIIIIIaqqqduaIdI- .- .., -.1.- . I. I I. I I-.III IIII...I . II,IIIIIJWIIIIVI I. :«a~amd~.~ num~.o smonm aI mmwm «as m mHanmmmm [Ii-‘Ilwflmmflm immofg $2.0 $3.; £2236 awed- . $53310 .I. 3340 liquid «3.043410 NooHNMao.am~I zqmz manna) m ohm m.aam chm, .s. A mama ensue: «or a music) a I magnum _ IIIIIIImmmmwmrI mh~oo.o II o~ommoo~.ano mfios.o memoo.o oaommoo..o~o a >so¢anUa o . Hwaoaao commoemo.a~o I- smbw.o sstaao mmaoNomwaNmm a axoo>sa IIIIIIIMmmmquIIIIImwmwthII oaoo~mpc.oa¢a mNos.o ~m~oo.o Nqomromm.0ao a zmnaam ”on“. so dalmanmo.omos emawio NmaMQAMI oe QQmo ooomfiII a masses oaw4.o aOgaMMm mmn~nom~.¢smm~ omna.o mmo¢s.~ oo_am~oo.msmsn a aoa>poua mn-.o goes» a ommam~oh.omm- ammo.o moos~.m moaaoaeo.-~mn _ >ou IIIIIIIHawaAWI mamoo.n ~o~¢mm~:.~aana mamo.o sooaH.¢ homecomo.ooam~ a I ws_:xzoz mama. o m .m H-moo¢oaommo~ memo.o rosm¢.o oases «Inmaqfl. .aI parth 0 «Home ~ seamsmsm.ho¢¢a mmoo.o “mono.~n anom~o0a.ooohp H maxa u.A moan was<>.u mm nou wr<=amI¢ m A comm w34¢> m w¢om< mxh «on ham to mu_hw_h mo m_m>4<2< wxaowuoxaIb2w2m>o¢ax— wx<30mix tat—x4: th >0 QZDOm ammo: mama—x1) o .hmwm. m:» m_ JmDOI w>om< m1» soaps >oa arou>sa campus Manxzoz moa>eoon >huoabou zmaapn maze seasonpo.o o Juno: II . I I.I I II Jane: 2. mm;m«_a¢> axaaoaIm 21 «maxaz Auto: flow-IOHUOK v«< Hausa Lou :launoa0Luox Hi how lua».w0law Had-Ichloz was union shoe Qua-«LI> vsovuonoqu seam opauaaslsu —_ I ~> nanm .. 24m: IIIIIIIIIIIIIIIIIIIII . III 33.; a new a. NEW .olfil .III. I. A no: cafe: «on. e - .33.; a 333. IIIIIIIIII II.- .I III III .I II I. II! 336 «some; $332.43 33.0 «35% 3.333;; a .6353 m A 3.2 m3: u. ImmIiiimI i A as: was; i am. 1:533 to 318m . Naomnmfigmso m~ 528 $5323 . 32%.;an €333.83 ..~ «9.5 a 333.“; amooosmmdz 20:359. mugDDM door £62 2: so“. mm D mu m bomnxn wd‘Domld tax—z—x wt» >n ozaou awaoz waa‘_¢(> u .hmwm. wzh m— anOI w>om< mrk uumou zo_mmwxomx . wam< wuz<~¢<> no m_w>a8 $538.0 a . I 325.. 8358.6 a I. :8: I Iguana-.2. Sam—I5; $25.?“ I. E 5252 IlIIlII Hove! scannoumoz ounusn>u< o>qasao¢ covauomouoo: :u uouauasa> aconconovuu Lou :owuucalhoaon oaewausz no nunoauamuoou mu I H> mJfi00 QIUU>FQ mathUl wh~IXZJZ QD\)FUJI_ )FUOI—huk FWXFUK ZNOIDE MEXA m00~090¢to CH IIIIIIIIII JWQDI IIIIIIII I. .. I. I I I smock. 2. 33:3; II Missal... IEIIIuMmlbeII III III I O'Iu-h'xrmv‘ °>fiu..fi°¢ :‘ud‘fluaohuoz I III II IIIII OHAI1LI> 9EOQCOQOD L0H OLIVOUOLL Eddulliliufl VLIIv—Oln I II . n- I H} N11." I II\ 152 IHII . .II. II;:I III .IIIu . I %w.~L III II I IImmoom¢¢¢. I .I IIIIIIIIIIIIIIIllnmmmmmwmad ~¢m0m~o~.o na¢0.o won a wl_xxzoz _ , nonms¢~¢.o sowmo_~mon.III lands 0 «ones.~ .I- osmammoo.o analog _ IIIIIIIIIIIIIIIIII«omnoomwumIIII vvmmmfloo.o amoo.o coown.m mmmwmsoo.o aca>toan II II I IIIIIIIIIMmmmwmmwhwIIIIIIIIIIIIIbfimmIIIIIII mm25<> a new:; IItm mew MWMI I. _».IA more ouauo: moi n-IIII .I.mw:n<> m . moaaom IIIIIIIIIIIIIIIIIIIIIIIIIIIIII . o mon.o anaemo¢m..sp aloo.o oomom.« snoomo¢m.nas _ wpaxxzoz “MMM.M. nmmo~.~ mansmm00.¢-~ «oso.o noo¢~.o cosmonmo.~oo‘ a IIIIIIIIIIImmMHwMIIIIIII «moons ooo¢~.o Noosoosm.mmoa sano.o ~men..m ammoeomm.nso . ioa>hoon u A some m24<> a mm nansa w mm a<_»zm:owm to magnum .I~¢.o Fooo.o 4w>w4 wuzom< mzh z— ww4m<—¢<> m1» axou>ha mashed wh~zxzoz aos>huoa OmQOFONQno n atOu>hm mamhua wh_rxzoz nonbhucq moonm~m¢uo maahoa mh_:xzoz teaspoon zwozam ~¢o~en+¢.o m axou»~a magnum mb_xzzoz.a0a>poo4 emzeoa zmozsm mm¢¢m¢m¢.o o I >oo azou>pa mambo; mo_z:zoz aoa>suua hmxwoa zwouan ooooapm¢.o p nuaem >00 axoo>ba weapon me_:xzoz aoa>puoa auxsoa zwoaam mom~a0a+.o m eoapm soc azoo>ea maaeoa melzxzcz aca>~uo4 >pooahoa possum zwoasm mpaeooo¢.o a l hogan >oo arousal maaeoa ws_:xzsz aca>hoan >puoneoa emtnoa zuoxsm mass moancoo¢.o on . I. Juno: ouuuso>v< o>fiouaox :uuaao&onuoz ounuanu> esovconon Lou ohsuooocn :owuqswawam vnlxxoan «— I I I H> mafll OAOLSOI l1 OH£IdLI> vilfillnuficu XL°OIGINBIN O>u9lu3l50 IOLUQI Ollvfll>v< O>uaIAOl claim In! Vt! LIIOE #010: ‘0 L9H Ionflldvlflm COdlIOL V- l H> Hdn535.“ $26 SSTo 3333.5 33.0 ~33; £33343. a wfizxzoz i334?! 2.33 333333 “Sadr 32m! aofiodéfi a 2.35; onmoé 03:3 m3~2~m§om Zooé «221m 3333;; a “5:53 u A no: m3; u .3. .222: a 53?. was; a mu. 1:533 is 355m L $333.53 m~ 1:: 35528 o+onno$.~n~ 5:33.23 3 .35; a 333.3 2.3336 32.0 26:; c~oo$~<£o~ $323.1: 3 22333.. £5 333$. u. A 2:: ~31, u. 5.33 23: $533 no :3 “a 3:53 JmOOt w>Dm< wzh.mow w—m no wu—hw~h mo m—m>4ozdt— wxdoamlx t::—x<: m1» >m CZDOu Amoox mid<~a<> ca .bme. m1» m~ qwoox m>Omi mzh vodka >ow uzoo>ha mambo; w>_xxzoz aoe>wuoa >huonhum pwxhoa 2mD¢Dm maze macaooo¢uc aw Juno: nmooz z. mwam<—x<> meDOwJu zu zwotnz none! ical-Oh...- ofloouu>u< 2:01—01 nuamuonohao! nu indoor-Ix» uuovccnovcn v— I H> Han

2 x . I H“ ( IIIIPIIIIIIII . I . e moo-r . . v 06.0..“ . ...III+.........:....¢...3.13..a:z......Iun:...u.n:.|.:..n....wI...C....v........u..I+....:..24.0.33.5.. EII IIIlficnohloMII I. IiIampFIIIcIec . om cc... 3... £00.60 000.0h ocooom cannon cc.” on.— . It. . beast oar—"annoauoa wood 1:1 soap-Hanan uanfidoaonaoz oood ll HIHHH mu—DGHE . I.. II I I I— III - 190 III-[III IIIIII III . 03.3 . 80.2. cacao». I IIIbbEI 98.8 .585... O I I. II. :.... I. ...e-...o....eounue..-.+....eo...+..o.e......n..e....e....e. oco.o II . IIIII. .16- .i:!. I II ION» ..° -IIIIOVQ.PI-I.+.II.I . ., .. .. . . . .. . . . . . .. . . . I II.-.:I.H.bvflo: .II. . .........................w..........................n».n............................,................ II I a. i-.. boonom II I...) .-I.-bo.o...o~.l I I .. . 009.8 Doonon I000 .0... , 6.8.8. 23.8 oooamxII I...mom..a.vI.. I- -993. .- Lot—on o>queHnHMoa umeHHOQOHqu mead one coma—maqaom :suwaooouuoz 093 N I HHH szUHL _ 191 :owpcvconunaonhovea odouo amoé coo}. «c.06- . Noono.IIIIIIiI ”mu-D nDDoD III Ihoo.o I. . III-Imwcld ..I, .-I. I moo-v Dona o21..............l...............1.5.3....1.........I.........J..v..........-..»..I.-.-M.I.... . _ . I - I III IpnuhmI . I . II I .- I . -I - -.- . I I- . 4 66...; - - e _ 1&0 Z x . ..I II -IIIIIII «Ea m x . I I I III III I II - ..III III..- . . _ me. x . . _I I- II.I-...-III ..I. II: wait. I.- I ..... I III.” i. .- 25.2 I _ . in}: u _ - . . I -. - I.I .- - - . - .I I I ”I I I s I ..I—l II I II I I I II I III I .IIIIII I III I-. . I II I I I I Il . ..I I I ILENIIIIIuEIwIlIIIIIIIII l3 x Jae x = r x . I. IIIIII ..I. .- I.-.:..-I.I.. - mauw .xI- . III- . sooomm . _ .I - I . I III-III. II I no IIIm I. . - -.I-.I I. IIIIIIIIIIIII-ZAVI xI. I .u , o IonPum..I . E;- x .I._- I. anymo- xI (a. x . I - zzflrx. - . - . I 30.6.3 . 22:). x , IInIIIJII IIII.-5¢.x II I . m3. _ ma. ,, . . I III I I- oer , . , . - - ”I. I - I - ..I-I II.-III- I..- . 02. x o . 9 x1 01 I X >I g- x. I i l | I l | 00.: J:- 'l' clot I I | 000.30 441 x C) x . , . . -.:I.III-IIII-I.I IIiI . _ mau 2:». S<6H x >1 x . .. ..IIEI - II II.. - ..I _ o III I L I I- II III.I I I I I I I u n mmci x _ . 30...... . I I- I H n _ III. I I -- -- - - - I II - -I . 2.1.22......._...:....I......................................:.............I.................u.n.I.-... - . - -. I I .-I . .mBE .I. . -I-- .- Ices-AIII-II. IIIIGEBII . . -. II..IIImn.0an.III.I,. II III-I a“... c 000.0 500.0 zoa.o mood .330 «OOH .:o«uuvnonohno.—Lov:= :uuwaoaoe—uoz . m I HHH ”:5lo 192 sewoau:omohnunnou:= . 20.0 , mm .c . I. III - I .aoIomoII I II.II..I oo-ofloII I... 30.0 290 0:5 .306 30.0 I....o:.::.:....6...;-o.o‘oo.I....¢I....o...:...oo:..r-...........m...¢....mI.M~o.......:.I..I.I.-.I...n.o «new. , :03.» o - . - - . . o 0005 c.0343 . . I .II. I . I II. I I V0,: zCIAI boglonm 3 a.“ u . _ . 5:4; I III: I II - . I - . . zay x . «3L: dox. ..- ., . .. 35.0.. I . - I , I. II I ..I ME, 2 I 0,... . 2:; 3L III-III- - I I.-.-II,,.I.I.I- a .e. I... - II III. I - . - I. «3: I . _, III -II I . . _ m: x a>x III-II .....II..I--I II IIIIIIIII I 930.3 .. Wao x2?“ . I1 I II. II I: .II . I I II III+ «How-5m I I. we 0:. X _ . I I C III M . . .- 122-m . 1 ix. - n O .. I - I. III.-. -I ,. I. ...I . . I: :5“. x x -. , I.-.I..-. I..!I..-. . 06.3 ..I KW... x _ I . I I. . I .II III-.II II.I.-I.. o I - m .. IAééx .IEDII . . . >w_xx . .....I.. . I I... IE5 x!- EI x. . nos 3 . . I «Ex 2 ..I I ..I - . - JJH - ...I..-I..II. . x - . HKI .- I: >7.“ .. II III. I. II. I .II III I III I .I- II .I I... .II. ...I IDD)oan w: ZZUU . . . I LI . . .50.: _I: I . . II . I .. II. -I III . . 9.5.1. II .I I - ..I I. III . 1.2.:...I........:....+......I.....I...n........u....I......uI.....-.I.. .II. III-.II: .II... .- ..IIII...wor-...o.IIII- - - moo ...o , . 2.0-5 «86 noon .cowuuouononnonuovcb :uuuaonouooz I.I v I HHH mgUHm 193 I. I . I .. - . . . . v.3. _ . oouéum 08.0.; I. Spupwm “6:?me I . .44.. I323... Y.2132.15.23.25. ..:....EH .‘IOIO§’III.IIII.‘IOOOI canto + - I ..II I _.Imo|..... III . I II .I. ....... .I I I 69 Ixi.-. . -I...I.... I I-.- .II I II .:..I II..- . rm: .2 x . III .I . IIIII I .-I I I I. III... I . . . LEI. + I (to Ix. ...IIIIIII III «I .m.KI~I.: .Il . . - Iii. -.I-.. I.I.,.....I.I...I. I. I I . I. lIlI III . ...IiII ..I II II I .I. . III .III I... . I. I III II... II ...IDV» x III. I I. .I I I I. . . I .. omfl.o ; div W ¥d4 x . cud.< . NEH? : I I -.I”- I I E »I E: x . III- -.I. . . $22. .I I . I . 35.... I I . . Sn x zsz x . “wanna: . . . I I I..- I . III I. J I I .I I I .II .I .I I I IEIIEI. . I I III ...III .IIIIIIIIIII . IIIIII. I III. .I I I I I . .II. I .II....I ...I Ed x I. I III I I I . I I I I. . . 7.100 x . 3kg In. I. I III. -ngéuax 92 x I. -I I I I. . T}... I. . . d4 x . o O I H II I III II IIIIIIIIII I II I I .III III. IIIIIIII I III II .I I II IIIII IIIIII I II I I I I II 9 o . J 03.: . .I .-I I .II- .I II- - .53...“ III... I., LIT... . I ..II I I IWUOIW .. I I I . I I . I I II I . . I -MIII I- II.I..II-.:wc.m>. x -.-I I.. I I III. I - II I. aIz . II I III-.. I III. IIRAIBI . . I- I I III... I. II I ..I mwmnvIIIII IIII .IIIIIIIIIIII .IIIIIIIIIH. ...IIIIIIII. . I IIIIIIIIII I III II II I III III III. .II III I I . ...I. II. III . I II I I . I H I. I I I I one.) » II . .II. III. .II... II .-.IIIIIIIIIIIIIIII I I. III I I I .I I + ..35 II. . III Lutz «II II. I. I . uficu x . II ImIEIB » I . -III. .II -II ..E x}... I III.I-.I.I II- -I :II I I I .I 4. T .IplliiIIII . >2 x . . I IuIIIIIIIIIIIIII -I.IIl I .I....IIIIII; IIIHaIx aaooo~m cow-ppm woo-ham IIIII I IIICiKpr NIHIIII I I..III.IIIL w Lu. L c. . pecucmfl kbc.oo vat. Hana.— hom EGUMHOQOLOOZ 93 Islam o>wvunnwwoq :aufiaoaonaoz «03” m I HHH meUHm n-9-uo.+oooo¢.o-.>o.o.#...o+-...¢ou~.+....+ooo.+o...§....o-...>.o.oo..-.+oo.o¢....+p.o.+o...‘co..§-o. I.Phthbml..|lIIIIIIth ”Db IIIIIIIIIIII I I. .5 L I. o ..:.. II I. 194 ouuuum>v< u>waaaom _ emu; . coed 3.3.0. Mahmud... .. . I .I .I. ...II..II..I...II.. III. 3.. .IIIImE E 3w MES . .. mmm... III I ..\I . IIII .IIIIII.. II I. outotao-o§unuo§nacntunuu§ouuo.-voo‘-ovoonooobnoououooo.no-.¢:-oobuvun I...............,... omooo + . .I I III III I I III I . ¢ (n ( w I .I I I. I I III «.5an . . .. . . . . .1 . .. unmet (I . . , .IIIIE IIII..II.I.I .IIII. .I , . . IIII III! III .I ..:.)yx . . . . . .. . om~.o + . IIILhNI... III. I I I II....>WZ x I ¢>.3.x . . .. ,. . I. I . . . . I . ZzwIPx 5.5 x . OMIMUJ b . . I IIII IIIIII IIII...II IIIIII IIIIIIIIIIII III .IIIIIIIII. . II. II III II IIII I .IOI (MINI _. .IIIIIII III. I .I ¢IE x I . . IIIIIIIIIIIIIIIJLfli xIIIII. I. .IEID x . I I I . CE .III I I I. II I III: III “I. '06 (A x n A I I. I I I III II I II I I I I I I I III II I I IIIII IIII II III IIII III II I II I I I I 9 o J 3.3.0 o . I II II w,>>Ix II II I I I I . I. ...I I... .I IIIIImuw II II II . .-:I III III: ..III I .II I . I. ..wuoix. III... I I II III III I II II I I II IIIIIII II IIIIIII I I. I ...... IIII I.I III I O I II . I . :EEIHI I I II II II I to x II III III II I III II ...I I I. I II I I mhInID T I I ..III. ..IIPmImaIr ..III. II I III I II I I a. . O IIIII III I III IIIIII II I II I III I I III I I IIIIIII II IIII I II III IIIIII I III I I I I I t I I. I I IIIIIII I II I III II..III I I II . one-3 . I IIIII I I II III . .............. I . . . I .I III. . ...Ipmn... I. . IIIIII I.waI.oIII qu.v .mNH.H who.o I namfiL . msofio mw.m£L .. I . . IIIIIII.I III. II I I III Omavca>v< 0>H06Ho¢ :uvfiaonofivox Ucu Lozom 0>wpadnflw04 :muaaonouuwz noafl III , . I. eIHHH 35on 195 oooocmm ICE 0 IIIPPDUJII pro.obIm «.I....ct.a..+.o..o...o¢.-..~I..o:....+....¢.o...z.q....: Hi: I. ,I 80.93 dbéfl ...... I 63.-.} II .ISIOPRII ..I. III ..I..+.. C8,...IsNI 8,063 9633 I I I .II I .IISHIPIIIII II . . . . O I I I I III I IIII II II'I III III IIIIIII III II I I I I t 93.0 I . (LIL II IEIF. IxI I I III II ...I I II II II o NEH x . . . I . I II III I II I II III I. III IIIIIIII III. II I . I I II I . I II I II I II III IIDDNIIoID ¢ I II IIIIIIII .IIIIII II I IIII IIII II II III e ILuINou I I IIIII I . >v_ x . . . .. IIII IIIIII I II I II.II I I IiI In .I III . I III! ..III III I III II . con-o o 4) >> x * L m.u II. . I.. I I I . III IIIdSIm uII ...I! I I I IIIIIIII II III. III I . . I II I . d5 2 x . .I I . I .I Had Ix III II I ..I I I IIIDDwnoI. I III. I. . I IIIIIIII I III II: III .. II IIIIIII «w x I III. II I II I 73a. III I . . IIIII. szr x Idu x I . I II cons . I 92.: I I _mC . . I II II IIII IIIIIM IIII II I I I I III I I I u d I 0 I III II..II III I . I . M n I I I II III I I I IIII I IIIIIII I III III I I II I I O I «.de x 5d x . a IbbwupIu. I .Muo (I I I II III I. I III . III I. If OIHIII I IIIII I ”I II.-. I I III . I oi Ix I HZIx .I III HI. I I I It ET? I ..I I I I III II I. IEISIIK I II IIIIIIII I II I.I I I.I II I III 02.0 n #37: >02 x . I. TI . I. III I .II I III IIII . II I II I II I II .I II .II III I .II I II IIIIIIII IIIIII I I I II M I I I I I IIIIIIIIDDbupIII In I I I I I III I I IIIIIczzwa Ix. I I IIII.IIIIIII|. II , I I . III II I ...um.x.. II I III I szIxI I I I. IIIIEEIFIIIIIIIIII I 710v xI IIII . IIIIIIIIIIIIII II II. I . H oooa . III IIIHEwIleI I I. III .Co IIIIII . . IIIIII . I I III I I IIIlbb DPHII,» II II. ou‘ooo 8.0.03 08.03 _ III IA I III IIIIDQ u... DImmIIII IIIIIDD n... D vaI I III p uoIanM 3< 19:. non :uufioaouao: 33 was ..ozom 3333on 5333302 $3 II III. I I -.IIIIIIIMmIMI I _ . I I. .III IIIIIIIII IIIIIIIII IIIIIII. $ 0 abw¢..oo¢ouo.§oao-¢-uoo+ouvn§ooonosuuooo I... IIII III IIruLIIIwIIIIIIIIIIIII. «Oncoo§ouc.onnootcuooéuo ufiv I pr KHHmNII IIFKLI. Fm I.r I II I III CLO-COM LCCorCN -.O.-.06nolu§v-ou‘noo to: .00.» N. I HHH HMDUHE 196 "IIIIIIIIooo.~ IfooaJIIII. . com.“ 03.0 O+OOCIIOIIOIO§IIOII|AOIOviii. U .II-III.I.WIDIIOIIOAIII III.I IIvII III 20¢ . n, M» I. c u nPI-IpIoIoVI-I! - uHII-I-IIIMIHuIdI-Iq u ......M II... al..»I-I-Iu n§I- I .II-IVI-In - vIIIIIIII .I II I I II I I II 000 ooow + II III IIIIIIIIIII IIIIIIII II..IIIIIII. I. I I I I III. IIIIII I I I . 49 x x Ha < “792.2 1.3: >m2 v. . II.I.III II I III III.I I I IIIIIII I «A II I H. I I I I I II I IOI lflWle VN I zzou . .32 m. I. III... .I.. i x ..I I ...I x I... I I IIIII I IIImquIIII £3 x max :23 x . l — X IIIIIIII II II IIII I . 8333 . I . I : 42 x ZWP x I Q5 XImII .III II . III IIIIIIII. I III II . II II I I II I .II . . IIIII IIIIIII I I III I t... I III. I III. I ...ImuuHIIIHmA II. .I oooIIONN .+. £03 x III II I + www.mumw I III 0333. + . I 83.?qu . I II: III oQo.oo~¢II II + I III II. I I I II III III II. III.IIIIIII .III II I I I. m I ..I... . I III . II I- I I I I .II III IIIIIIIIII II... I. I. Ido I ”III. IIIIooouoIoRI I In I I I II I. .I I .I. . I . I III III I I I III. I . II IIIIIIIIIII IIIIIIIIIII I II II III III III.I.“ I m. I. I I I... I, 45k: I -.-IIIIIII I. I In 0 a m III IIIIoIq°I$o~IoI I III III III III-I 80.0on n IIIIIUIP....II.INI0I II I -I. ooaéowo . I III‘Oo IoIOoIIoIQI-I‘vuauo+norntooILIn§nnIyI|IIlIIII I08; I, I. 03.4 IIIInfiItI-ovItIIoorrIfinco- I I .I I IIIIIIII IIIIIIIIIOIQOJ I 30.0 Joutu IIIIIII I I ( I cam; codeI. II III.I... I SIIIOIIIII II III.I II II I II ocoHINooH $53. Hanan Lem :muwHOn—ouuuz fl.“ OMS—£0 «as. Autumn—402p Gav cowuwgaom unawHOQOLooZ Apfio ownqd owod {III I I .I .II IIIIIIIIIIIIIIII . m I > 35on z 000.27 + I 000.04I III I I III III‘II II . «.2 It: 2 x . . w<> x > x . I I (>22- 5 x :mES x . IIoIaIquIo o New ”Ix . Icuqud . . mauz x . . III. I 1am x . . >z x nzx wuo x . IIcIIIIfllmIIqI 03.93 . + 0.5): II..- . .IIIII I. IIIIIII III.I I IIIIII .II. I I III .I II I .. II . II III III.I: . Zzwr x . I I ...... :a x . . I . . JIqu I0 I. o (ugnrom . II I I Ha Im I a: x . I I I aJéIm I .III III II III IIII . IIquE. «I II III . 7 993.com I. :55 x . and“; . . II 1 . 9 . . l .NLMI x o I I I .. . Jéu x . N W . n w 94??? u + 3039‘ In M III‘GI - I . I336: .I in x I «383 I I. UI II . 52 x . . . H 03.0.3 I” II. I . I . III II I. III.I». owaI.IrIooII.IIII| I III I. I .I I II. I I I I -..IIII . III. I I. I IIIIIII I i 003nm: ”I I: I III. I I I ¢ ope/I: z 00 x II I.. I I I . n . . .H 2.6.00» n + 9.583 fizzy....o..I.:...:11.33....2: .I.. . . u .4 . I. . .I. .I.I. . I .. .I.. u o o i . . .I.I¢I.I.I.IuM.I.4uI.I‘4. I. .II.II.I. .IIIIIIII IIIIIIIII 80.2 93:: I 532 30.0 So.~ 30.3 08.3 9.5.2 Iowcé OS; I .I I II I 3342: 62 395 .3". I .- I :wuwHOQOLH—oz :fi owcuzo _can nooaluooa .Lozom o>wuwamwwoq :muImHOQOLuoz :« 093—5 0 I HHH HMDUHL _ 198 OMMDGN>U< o>flvmaom (I I _ . Doc-co ooo.rh youomm ;uo.gm gtC-La _ . 5: z x . III IIadH x Mm x « ¢> Z, w 14 ‘( x . all-IDIOIII o 1 0&5): m I I. ILIU unfit. I I . an: m deo x . .I I I . I IIImIb Z xIII ...IIIIIIIII . .>.z x o z x Vic x . IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII {a m coo.oo~ . I ohu.no~ IIJI III.I IIIIIII III I III O I. I I IIIIIIIIII I . . III II I 21.9.. x I. .I III.II ..II I IIIiIIIIIIIII H . I . 3 x . aaq.uam o I I . huddddw H ast I I I Ha x I II. . €44 x . . . II I IIIIWCm WI . a. .I. 920.com + Imrra x . cpo.pom IIIIIIIIIIIIIIIIIIII4IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIII . Ngad x J fiIIIIIIIIIIIIIIIIII IJfixfiuIflm II II I I II II I II I I I II I.I m. n I I 0 I II I II I I I I I I J ooJIdd: + ..ouo.me«I III 2.552 ” 5m x ..I OBIYbIm I III >67. x . 03.3.. H II II . 399:3 . 033.: m . 9:. f H I 2200 x I. . I I I I 93.03 . . 80.2; . . ..¢.m..o....¢......u......n.....;....vnfluuuu..“Md;..wuumtt....‘....¢M.Hw....uu..mn+..uflotfid.hufl..¢udNiIIIIIIIIIII!IIII I oooéml 000.3 8%.? .5qu 3.3.7 30.3 08.2 5.10m .39.: $0.3 $0733 639535. 03.933. 50130..qu :« owzmnu 6:6 noodluood .Lozom o>wumauwMog cavwaonoguuz :fi oucmsu o— I HHH mMDUHm 199 . va< . cg .H 8.3 IaHnIIIIIIoIoIo-IOIII IIIIIIII.In..I..NI.cIII-. ..I-III-I. III.-. I I poonw ooo.~ oc~.fi wow» I .. ou¢.o ......o........:.........z.....¢...2..............I;......- - ..+..-.-.-ol.I...o....w.......II..I......:....I. u o WIS II .I I I. I I II III I .II II IIIIII I II M LCM-II ad 40 (A I - I I o 9 zzwkx NIX « 5:; :Iuszxw «can {>3 ufiuz - onooa IbImIN.~ II IoIWquIHI H I IIIIIIIIIII-IIIII -I III I I I. I 5.5..” lbhboIFI .III. II I II I IIIII- . IIIIIIIII IN—EI Ono.~ I. >2 x I m IHM x E x II-IIIASIIMIIIWdI/Ixi..IIIII l «Umrw . . D o o 3 .I I III.- IIIIIIII. I. . IIII. m LW[I I II .I I E . ooo.m . Sn— x I . dz: 3: 95- Imamgaéz 3.67:. I . .. I III I III .I - III? I I umN.ImI o .aOou.-§conoéouuu¢InsIOouo.+-oo-Ooc-o+nnou$onscone-o‘ccooo'uouonun... ovooot+ono..~onu§uuouo~ano I II numb“ .I onwwh acoug I . I - chIC ocng . I I uw< Hanan no; :GOMHOQOLOOI c.“ own—5:0 «:3. 25.35 o>wouamwwoq cwawaoaouvoz EM omumzo ‘ mg I HHH mMDUHm ammonm>v< o>wumaam . 08$ 08.0 cooJI 39...“. 8:; Sim ou‘uoooOooOoIcnuonfi-o m~m.0I I, DOOaOI I. I III II. .II. .I III -III-I I I I I. I. u I- 55:3- .- I O I . . O IIII IIII I III IIIIIIIIJIII IIIIIII I I II’IIIII‘ IIIII III I I I II II “2.0 m. I I- III II IIIIIIIIfI-IIIIIIIII . fmfu I. II . . . III. I o QMNIO H 'IIII'I I II I 'I III | I I I I II I I I I O Cmfio. u IIR-a-Hx x x x x m (x 5. . IIIIIIMINHHH ” I I I .I-I-I-I III I -III-I II [If -I III I I - II - III- I-I-II . mm — ..4.. . - III I - I ..... - II I-- II II I .I . . nnUu. com; I” I I I mama-I. :SIIIII o no . o n o I IIIIIIIII. III II I II I I I I. II III 9 II I L C I n a H II. 53 a I.- -n -- szx I -IIII-,.-éu,.I-IIII III-I III-III.I II-IIII-XIIIxI x-I XII . . . D II I I I III . I II II I II I 0 w .I--III.I .III-IIIIII-I I - I 03. u I om~.~ mNOoN b InII II III IrIIIIIIII I ...II II - I. I. III III... I - II II I - -I. I www.mm - a: H 5: . . i I . I Ins-z Iii-II . : x.. i III-SIUIIIMIIIIII- IIIIII Ilmhm ..M.l-IIIII.w.IIII I III II III II- I . ant-rout oa+u O‘CIIIOIII E55 035 9&3 00m .N ..p.,.I»....¢.33.3... . 80.0 coo; poo-m I . you.“ anuaca>u< o>mvaaom I Nu I HHH mMDUHm .50 .M - :uafiHOQOLaoZ :w «um—3:0 «Eu usvwvm o>wouamfiwoq :aOMHOQOLDUZ :« oucmzu 0.0 201 coicmm 55.me . . I ...I..-III~I ESE-CE - - - . Eta-Em.- II-IIIII IIIEqu ---IIIII (HEW. III .- .I....I....I....I....I. .I.-.I....I...III.I....I ..I....II.. .I....I.I.....I... 9:03: o . .I I Iv LIUIUHWI .II. I 26.1 I >1 x 423 I . I (kc-..:.. anoim I . -.I ...II II. III-Mcuflz inIIII -II III. . IIZzw... «II-II -I5II.I-Iw§> x . ,I . lad x I . - I .543 xI (>45 x I ou.I,.¢mI'II.llllI|l|IIIIIII- III .I . . I (.2 ( I :33 OExII . xvi-..I I + (4:13 :25 x 99.1% II x - J10 « I. I I I I (cofum II-IIIIIIIIIII ...IIIlIIIIII I. I .I- I . - I 5» x I I I .III . II III... III. I - 22ou I . uni-.30 I Hm x I Luriiw I I IIII IIIII-II III. I II - I I >wZ x I . I.II I II IIIII.-I-II bog-qr I - I. I- I raooiu I . II - I. I wazszIIII-II-III. . I .I I. . . - - II II - I I I . - I 000.40 0 .,..I....I..I.uI-.I.I..I....I....I....I....I....I....I....I....I....I....I....23.} .o-no-uvocou§ouon&ono III'IIIIIEINPPWII I , - roc.cm¢ ocg.umm I I .II-(Inw-cmmIII -II -I - (Simm— (yciuuo CLO-0cm rCcCo¢ o<<.nrm Loofiuom 33 .393 hi :uvfiononooz $3 «..:.. « nozom $3 _ n .- >H $5on JOAOJ 202 In” rmmfp mhw. a 02m .H m>0.c I ommvcm>v< v>wuaaom I ¢Iunnu¢.-.>-u Cm FIJI. .32... mpcflu . I III. I I I II II. III III III I “no.: I I . I . 4D,. I I I I III I IIIIIIIInyVII .IIII I tsumfl - . I . III III II... .. I. III I I II IIIIIII IIII III II I .I OOUJN I III.IIIMIowHLIIxIIIIIIIII I I . I I III «5 y 93 I cccim . I I I! dIZIIIrImII I (JleixIIIIIIIIIIIII . I I . v32 x . III . I. III.I.IIIII III II I d5.w x (> .>> x . . 1m?) . . 03.3.. . I I. (.2 K II I . cEJm . WINS}. I I II III. III III .IIIII I. . . II . o<< Y . . Unmo.3. # II I III II I‘IIIIIIII I I I I I # CLO-0.x . I I II ..IIIII I .IIIIIIIPD x I I I I I IIIII I I . . H55 x I I tguIm Inca KI. . I I .III II I . 4.6/2 x .33; ..III.III»,IzIIx IIIIIII II II I30 x I I I . cacti I . I I I. I I I I I I I I (a x I I . . I .ZZDUIIx I I ; I iII I . IIIIIII III.I. .. I. I I cno.:o . Ha x . cuu.q( n I I I. I I I I I I II IIIIIII>wZ Ix II . 3.3.; I I I II I I I. . . bacir I UGO-#0 o IIIIIII ..IIII II I .I I II o Cho-«a HI I I II I III II I II X02 I.Z IxI I I I I I ..I I II I III I I . .JGG-¢mm IIIWIIII I I IIIIII I I I II II I II I III III III ¢ COO-.vo +0..u...uu§ounc¢ouoo§oI~In III III in RIP. .I. III I DON; «:4 mhcoc cmv.HI cucuc mmm.c .27: I owauflu>v< o>fiuwaox cauMHOQOLHO—A oooa flaw N Luzon mead N >H ":5me ......r...............I....«...r.....I.......I...u.....wp................ rpm .g cow.c JGMOd 203 . 33.. IOIOIOIAVIDWIIIIIIIIIII 600.09 I . ooo.omn 000.0nN ooo.oma 000.000 000.00» 000 60¢ coo .oon ooo .OON IllIIIIIIIIt'I III.I-wc-IodVI..II-I~.-IoI-I>Iu.1-Iq-IcI-I+uuouOcI-u~+oncu+ucoutuugnu-ovuéuu->-'-Oo-vt-o..;u...0oo.-¢-.-.t.u.u‘-uu II II IIIIIOOIOdNN I ¢ I I . I III IIIII II I I III. I III I. .I II III. III .IIIIIlIIIII MIIOOOJNIIIIIIII. I III I II I I I . II I. IIIIIIII’IIIIIII I I I I (AH x n I. I III III II ..III .I . . III II II II lIIIIIIII I wand xI (>2, x .IIIomOIIImmIIII .. II I . I I I -. II . I I IIIIIII-I-IIIIIMbBMnIIlII .u5.z x 7v. x «Ea x II . o III.II .Ooo.anIIIIIII .III.III IIIIIIIIII|IIIIIIIIIIIIIIII I. I I . I I I. . I I oooim . I I {.0 < . a1: x . ZZNP , u I . 0.... I I 000.60 I 92 .14 x flz ( o OOO.ww I I I o I IIIII III III III IIIIIIIIII.II I .. III I III I >mZ x I . III I I I . ..II.II- I I I him. «I 5.2.0 x 92, x . IIII|II1wc3Ix XI I. II I II I I I ..... I... I. I I I I I 1:5 . eooJo . . (a x . 80.5 1.82 x . ..II .I .III.IIII. . III II I I IJ¢OI a n I I I I I I I -II-IIII 08.3 I I- n .I I III I I I I I- I I 80.8 a I II I I I u w. I III IIIIII‘ III‘III IIIIIIIIIIIIIII‘ I I I I II II IIII - III I II I J o z t $031: In x. I II. IIIIII I III: I , I I I II I. I-.III.II III.I OOIQII—IiII II I. I. $sz I _ . IIIII. .II.IIII II II III. IIIIII I. I I I. .I I. I IIIOOO.0N~ oooowum . Iv 000.3: N I :IIIII IIIIIIIIIIII .. I . I I I I . I I I I I III. .II.III” OOIOflSVI—I II.III.IIIIIIIIIIIIIIIIIII I II III II I I . . I I I IIIooonE I g z IwI I I .I I. II I I I I. IIII. I IIII.-.I.IIIIIIIIIIIII I.IIIIIII'IIIIIIIIII II I0&00004.I..o+nIoIoIoIrIcIoIcIawmuocw-oI-I-+II0I-vo¢oocu§ooontuuno>~onot..-.Qnooot.-¢u‘na-§ooooroo- II IIIIII II I II II I . IiIIGIOnIVIJMDIMIIII III IIOOOMODV II I I 000 ..Omn 00°.OmNIIIIII I WOO-OD— ooo.oo¢ eoodom ooododII. IIIIIIII doodom 08.0% a: insane“. :unfloaonao: $3 “Ea m nozom 33 III .I. n l >H mMDOHh . I I 204 owuucw>v< oinoaaox 305;: II 02; II 05.0 30.0 2.0.0 on»; 08.. 000.0 000.0 0230 000.0 voauvtoc-obcocuboucn¥000m§ucccbocnu§0000000-0000.09000000300900000900000-005000090-uavooo-000000000-0 |III°OO-I IIIIIIIvIIIII I III I I III .I II I. I ..II I I I I I .III I IIIIIIIII IIIIIIIIII IQQQd HNIIIII|I II III.I... II.II. .. I III.II I I .. N(UIm x I <>I7>x I I H II III-IIIII08....3III.II I I- II... III I I I I I v.08...» I I I I . I . III III. I I ~32 x In.“ N OOOI‘D .I .III II I. .III.II IIIIIIIIIIIIIIIIIIIIIIIII IIIII .. I III I II . I. II II I .II III 0 OOO.#D I out fi ‘0 a. (I .J I4 x Z 5 P I< «D Z 00060 wflo Ix. II «4» ...I II II I I 93 x I I . .0000- o 0 9 ‘ a 0 § I I I I I I l I I CC 0. x I 000. _0 3.3 i X I u U— , ..l G I U x I .I 000.00 I I I £00. mm II 00‘ I I I I I I t I JOMOd IOOO._I:III.I 0 II. TZIImIIIIIIIIII!IIIIIiIIII IIIIIIIIIIIIIIIIIII II I IIIIII IIIII. III. III III III III IIIIIIIIIIOOOMI:I_ I IIIIII .I I I III I .I .IIIIIIIIIIIIEMG xI. II II I . I III III .III.I.IIIIIIIII I. I I I I ..I I III . II . . I #00005 Be... .0 .000. _S. 0 000 .02 . I I I I I I I I I I I I .0 III. III III.IIIIIIIIIIIIIIIIIIIII II III. II I II III. II I I I II I I Ioooém—I | I I IIIIIII I IIIIIIII I. ”III I I I .Ime. II IXIIIIIIIIIII I IIII I I. III II I .IIIIII III I In I II I. II III.I I.+0omIoItI0IoII-II-I0I'I0Il-00-.II-Io‘IHIquIanII-IoIIoIIoIM0-..Mtomuv.Ioo-o00000000-$.-o.v-o-ovovo.90-0......oo‘arontn-.000-0.0.0.05 I. II I II -ImmeIIIII III.I.IIIIIIDINI0HIIIIIIII. I I. 2.0.0 . . I 30.0 050.0 03.. I com; omofi I oomd _ 03.0.. IIII IIIIIIII IoIoImId IIIIIII IIIIIII .II.III [I owaocu>u< 0>M0m~om :nvaOQOLHOE oood can m hulom hem...” VI>H $5on I 205 ('1 7 I n a . I .I|| IIIMMQ F u I . x x . :lIllIla ada.a«¢~ + d¢u m + (no.0mm— a . (.E x . 0 III .. . . Zzwkx . « |||IIIII||l1I|lI|III|III|Il||I|IIIIIlIIIllIIIIIIIIIlIlIII .IIIIIIIJVfidIIIIdJIII . q.00m~ 000.00ma 9 . .Z x I KO Ix + 00» III|UII||I| IdInIUI d>—3 IXX I l||||llII|I|1|||||||IllIl|||lllllllllIIIIIIIIIIIIIIIII IJIII flu. :«~_ + ”WS_ 2 m a” .H CCC JP II I u||l|l|l.l|I,||lI||I Ildl 000.003 . :50 x . 000.00: IIIII|IIIIIJ|l|l|IIIl|lII|I|IIlllIllIIlllllldJmlflllll I . El . II . 92, x . . I o: + .Idddddflddlllllll . No x «a x . M I- I. o w Goo-OOAN + . 1.32 A. OOC Omfiw W n x 000.00: + >wz x I 0 ecc-005m ...+....0....0o.o.0..o.0.-..¢....+....+....~....+....0....o....+...o¢o...+....+....0....¢....¢....0.. ecc-00¢ cco.om¢ coo-COM CCOJomH Cue oooomho oooomNm ooo.mnm ooo.mNN occ.m~ Vu< awash Lem :uaMHOQOLaui ooou vua oEoocH Accouuom aUMQaU Lon ooou n I > mMDUHm 206 omau=a>v< 9.333" I.“ awn." cam." and." coo.d an.a da<.c _— 0....+....+....+....+....+....0....v...I..............o....0....+....+......zl....I....o....+....; .. a o . _ IIJIIIIIIIIIII. . . . x fiddx . . chdeIflfiAIII¢ I 4.100 .m. o can—3.4m. . _ . 13¢ x . M. I . Zzwk x . ~ . I. OOO0DomH . .06.2 x ¢J x 0 060.;me 000.003 . . III 0 ¢ 0 . IIJMIIX III . waIn u ooo.ooI—~ o 0 0 oooooafim n . I flWmE m flulfl n III II I ~ 0 dadqa¢qu o I NEE m 0 LI . 0 I 0 o 000.03.“ . an x . 000.00.; . . 72 x IIIIII IIImjj . aaa.a:o~ + + ccc.rq«o ooo.oa- 0 >wz x ¢ occ.oo>~ .....o..............¢....¢....¢....¢........................o.........+..............+............... p d m o mmm c rrc.c own; new; 03: I 000. 0 0:0 000.0 DECO-06>"; o>mvwao¢ :MDMdOQOLOoS oooa «Em oEoocH chomhom «amino hon oooa N I > mMDUHm I 207 II/I vun< I. I I“ noo.Omm . 000.03 000.03 9.0.03 3.0.03 000.03 vinonpt.0.00coo-#:0000vn0obo-006000 ERII. Marx . mN.~..m fIIII. . 000A . I .I 30% 63m x x . M..Im....|pI 0 7M .IImIPGA . 0.1er x . . . , «5.2 x mm x . . mMEP x VINE x 9 x . 0mm... . . 0mm... . «in. x I I I 1IzoIonIIIIII.IIIII III . ¢7IH3IIx I. I. IIIIII ..I I- II MINPUNII I I. . 45. x JuUIx I My.) x H m2; . d¢u.w xI 02. x «A, x . W a 72 >WI¢IX I I . ,u 03.“ . x 5:2 x . Him II . . . O I I 0 II «5 x I I» {Gum II I III mmImIKIII III II -I I II+ImNNHmIIII . 80.0 ..:...................I......I.I I. II Em 000.03 I I. 0.006%...“ 000.000 000.com 000.00.. 000.00m 9.0.03 I II 3% :03 no; 03300030: 003 0:... 000.53 53. 1003323 33 III nI> 350: . 208 I cumunm>u< 035233“ _ nhw... m2; mg 0 I 3.04.0. InHoIfioI 0mm!— OONJ omooa coo-o Ombuo ooooo IIIII #oooo+-n00+.»on‘cuoofiu...‘nnoo0.o-0.0u....u..ofilosuu.v on§no0-+.00I-6JIdI-40|-IqI-HJI 0 + 0 0 .0 . . . . o . U I. mNNcm .0 .HIM— x t mNNom . . H (I . . 0005 . .dqodm . . . 2.0 m . I I I. m2...“ . 2sz x I «w . I . , Iwmez x .801 x . . _ x x x . 0mm... o I12} mm . a: . Sm . . :28 x x . . II III «53 x H . :33 m3... . “:43 x x 23 x . m2... . 0. ~30 x 9% x H a H I . I“. p . I >uI2 x ..I m 03.0 I in x «6:2 x + 31m I Id atin I. «I. x . whim . MMO x .I 03.0 I I. I. owImI.m n-.o . I . 3T0 . I.-. I I . 4.32 x . 000.0 . . 30.0 I O .0....OOII‘OOI‘§IIOO.§O.|I‘ODOIOIOIO.OOCO§OOCI‘IDOO‘OOOI0050I§IDIOOOOO...I.C‘IOOO‘OUOOODIOO§IOOO§CIIO* BIN; m2: IImIsIono 30.0 Emuo .II... 02.“ cow: 03; 000.0 02.0 000.0 IIII . I. I III owao:a>v< 033.043“ :wuwaQuOLuo: «wood Una counsm xaa Hacowvmosvm hood . II 0 I > mane: 209 \ 3< IN IIIIapm: ., III III AIIIIIII IIIIIIIIS ..InIII ,II IISIIIIOIIII, IENIJ COOoN COO-H CfiNOH ccmao Cc¢oc . jjJIGIOINIUII IIIDI§IIOIIII.OI* I. .0 IIIOIHIIIIINIIIIIOIflflIIINIOIIIImIIIHj . C b IIIIII IIIII III.II III.II I I I o..+no.o;o'..>:.oe!.I....I .Iflro...5..I..«.-.-§....vc.dl. I (II IEFD x - oou utm , _ 4m x I . III. I ¢> 3 x 218 x . 932:3 . I 2; Z I I . I 000.90 M I I II III I III III III.II II. I I I>Ilmum.mcIm I . I I III I 91V I . m :2: x . a o . .. . I I 000:; A. I II I I .II II I II II III HIM K II III .I II I II 4 tug.¢> . Sm x I . I I II I o .I III III II I I I I . .I I .5: III II I I III .I. o . ZZWF.X o I In I I III. I l . I I II 000 z. H .9 ((7:05 0'0. 0 I «w x aoo.¢m 0" ¢ C I. O o x? 1) I I I I I i I I I I I I o I I r)! 0‘ ; m; I 0 §: I I ‘ . uI-JIolytI-MI-I-I IIIIIII I (Cmoc m I > MMDUHL 210 III II: III III II IIII IIII eczo . can.» . wt; I. . ImamIflN: I II ...Ik I . cg m 06c m wou.~ ; III I III III .I II IIIIII I. IIIIIIIIII III I II .vul.fi.>.anur..so+cvcu+oooIoII¢mmntbnoou+ outflut..fln&anontlolu.ny Vol-.IquIouo.r.I-o-+.uu.-.oo>1-JI I033 J III I I 03:3 . . IIIIII IIII IIIII III III III I I IIIIIIII IIIIIIII I. III I. III II II H I I . I I III II I III II II mmeIIIII . IIIIIIIIIII IIII 46x . oooém . III III. I. I I . I II I. I- II I I II II II . II I. III IIIIIIIII I I I I III. II . / I (kg-0m Omaniv 9 . IOOII . II II .II III II II I I III II I I I II 5 r .40 . >2 x mirx m l .I I . I . n 54x IDA; x . Idedo IHIIIIIIIIIIIIIIII I III III IIIIIII IIIIIIIIIIIIIIIIII I IIIIII I II I o o O CCLQOO I I I III I I I I: IIIII I I I Eh. I . _ n I III .II I .30»: x u . o a II 0! UGO-:5 o .IImJu XI I I II III III I II I II II I II IIIII II I Hm x I I 93.2. . _ I I III II III I I I nyx I I . 80.2 .I . I I II II I ..I III zzIwI,_IIIx . goouyfi o I I III .III IMSWIKI I III I+I wIqumIIII Im.mo,.I..I.wI.I...I... I I I III IIIIIIIIIIIIIIIIII £000.? I ecc-o ooofq . . poo... coo.» II I I IIIIIIoIi; I II Seam IooIcI N III-I I(.c I I III 0 I > MMDUHE 211 3; II I I IIIIIII 000.0mm 000.0m¢ ooudmm 300.ch 000033 000.000 000.com 000.090 oooéom 000.00N IIIIIIIII 0.000.090.0000.oc+Io.-H~o.-6a quvlnmmI‘d-Iflflmw;nvIflqjdgfiqujjg; 0.0 o ¢o_ x . 0.0. . «$3.2 x INNZIX HM, .I I I.I I I -I I IE0 09:: «>3. x «32.. . I. III . ddufi Wm x II III u I II .. IwaoIII I 3. x III III. IIIIIII. I I .II. III 000.009. + #52 x I (I; .I III I +IIdwc$rInH . 533 x «a 52 «A x I93 < . u I 12?? I 228 x . O I I. I I IIIII I O IIIIIIIIIIIIII II II DE ‘11 I I I IoobwbnomIIII. III I III I , .I I IWImqmuwmum III . I I .03me . II I I III II . o I II. II00.0.I00wI.I.III.III I . II . ..III.IIIIIIIEII 0 n _ 31.2 X 0 000.0000 ..I 00E x I II . £0.83 III C I I I dIII . . IIIIII 0 000.002 . . III.I...IIIIS m . . I I I. I. II . IIIIII IIIII W III I 0 l T0: I I 0 .I a s U I0100..ooo0 u I J mfilx n (60.0.; I 000.0003: . I I . I I I.. 30.0%.: I III00I0.000- . I .I I 30... 03 I00I0u00m3, I I I I I . I. cacaoow: I II III. III I I II>Z .II II JmoIIx II I III I . I I x I bopuglomif II II III-I.IIII III II III II. I 30.0 03 oooocoo¢oocuoonou0.ooo+o 0.0. ..v. .Qo.ou+. o0..c.¢..o.9o.u.¢.....IM..I;I....0:..‘oc..Iw..I..0uomI II II I IIII .II I III II cob-0mm ooo.oImI¢I III.II, I 000.com 00.0.03. 00.0.05“ vw< #0030 no; :mpwaoaonuoz ooo~ 3:0 Autumn—.55.“ ...I: cowuafiaom :mawfioaofiwoz ©00— n I > mMDUHm 212 oIIwapcm>w< 03.5.30: mpm.c _ooo.¢~ . . noo.wm _ o o I x n . E, 3 x 05? x . EIVIII W5 2 x >xIxI I III I III flow: ...I . 0 aqvw x . 000:} I. . Io. Iowan: I . tw « . I . 22w... x 92, x . mIIIIIIIIIIIIdIIIIIIIIIIIIIIIIIIIIIIIIIIIIdemImIIIIIIIIII I . I m 25ka .3 x I I.. 8. . m .II _ I O I I W . mwo x . M u . .u 80.3 o . 12:5 x I 80...; II . ¢4m x III I. II . . I52 x . III III; . . (I5 x I562 x III.IquInIIFII . :uzz x . . 4a x I u I I I I . . DEX -.IIIII IIIII 000 we I Hfi x I. 0”}..me ILIxIIIEF II I.II MIIIIIIEHBIMIII 02 xI II II III .I I II III I pnphnbIIIIIIIwIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII I II. oou.¢o w I > uzDUHL 213 I :I III I III. III I I. 1 _ III I IIomoaowwI 80.03 c oooémm 03:; $0.03 8053 08.93 000.com 3053 . O...-O‘DUIC’I0009'I.I§OIQ“O:I§OUO§‘33§434133I44I . «.50 x 52 x (>2 x «a: . III III ooo.oo~ I xwfii x 51H: x I Ha XIII #92 m I IquuqudN o NMO x ddUJL x x . I III I II .I 4.3% .x isz xI IIIIIIII223 JwII .IIII. i I I . 1mm; x ¢.w x 23 x . I . IIfimeI cEIXI < 5 x I I 80.83 II. In: I In.zII< In. ouoIJoINII, III 80.3% I. I .!I II IajI H. III.II; E? I II I I . I «m x III II I . I. III I oooBowm III. , . n 85.2% 25.8? I I 80.80? I I. .1 N a n d I o I I II n I I n I T. boowobwm I I 98.0on w . . I . J5 x . m. H . H . www.maog .II I I IIo 3.....IoIIoI II 30.83. I. I muodmwwm I oooéoww I I 93.8% 0 I III. I III.I I I I III .I II II. I I I.I IIIII 80.82. VII? I . . x .I II 0.00.”. ofitu .IoIb...ct...‘+.oII.'I§o.o.>.I.o.¢o..-Ou.IuIIofianonIIIIIII..III.I.HJII..¢I.I..III.I.I .IIIII I I .Io.o.oI.oImImI I I 80.03 03.03 08.03 000.03 000.03 08.03 000.03 000.com I II .I I II III I 214 3< I“ . I I000! 00...“ IIolojIII 90.9.0. I -0350 I II. 000.~ 0004 00~J 000.0 000.0 IIIIIIIIII IOI‘IIOQIONIIOIF‘OOIOOOIno.7..I...I.0‘00!I‘D-0"...n+0.OQ‘OOItfilo..§§jfi§dfljiak.1%qd| . 49 x x .8 x 062i «3.5 >mZIV.‘ .. . . III..- 0I0.I.I00I~ -III-I. IyIdmx 0m x (I .I. I I 00 .SN . maox «:14 ~20: x 2.10.0 .I. II- I . III. .I I .. I I.I :52 x 3. x i x . . mer/Ix «.wx 10.13 x . .Imam x zzmrx DE 03 .I 000.00I~I~ .II. I 07m x . «JII II 000.002 000II0IaIu a x I (0: 0 I. IIIIIII II IIIIII II I III I an. X 0 I I II I 000.003. I , I 50.00% @004me0. I. .I 00II..00IINI..IIIII! I 0 d I-I I I II 0 III I II 0 I. .0 0.00603 II I 000.0INm m I O I I B . I . .30 x . m . . . Im . 0.000.00an .II I I . I I I III I I {300.3031 I.. I. II I .I III II I. beefinofi. II I .I I 0.00.003 000.030 I n 000.3% III II ..IIIIIIIIIII II I I .II .I I I I- .-.III .I I I. I I I I I .II I I I. III.IIIIIII . I 000.080 I . :2 x O IIIIIII ...I...I.I.IIII...Iu.II.IIIqu......mIIIuI.I.I.u...IuI.Iu..I.I.I....I...I.I....I....I.I..III...I....I. I I : I. ...I. I. Icon.” ..III-I oIovIJ I. I Ioomflw: I, ecloho oo~.o I. I: ooo.~I.II.II.IIIIIIbowIJHIIIII I 09.0.... 00030 oo¢do III .I I III I... I I I III I: I I aconluooa .vw< Han—am Lo.— caoaaoaohuoz ca 09230 was Amt—.3395 :3 comawaanom :muwaononuoz hvflu own—3 com" o— I > mMDUHn— /. 2l5 000.0 outta-tycootbvocofinooI-tnI ou~un~>v< o>«vmaom, Ioooé III. 000... Don.m oco'm .v.w..o....¢ ........I.........>..t.;...LI... 000.~ 0.0 h..n.tmm.r>.mnlll|II||l|III|||| db_x I . III.I x N o 03.0% o I I I III .III IIIIIIHaIxIIIIIxII I .xI xIII +I000I.I00~ . I IIIII III IIII I IIII II I I I.I.IIII >¥ xI IIII II I I IwIII II In .I . 43¢ x xx . . ..I .IIIIéII IIII:I|I ..IRM.I II IIyIIIIEIlIuIIIIII . mAm x as x x x . 00.3.53 . . II I .. IIIIIIIII I . Ix IIIIIIIII .IIIII .II0I00I.I0I.02 . . O . I 08.33 I II 1.91% . 000.002 . . . , E III III II (a x I I . I I II III 000.003. . . I II I III: III. I0I0I,0I.I00I~mIIII a I. . I III I I I IIIIIII I I II I III III IIIII IIIII IIIII IIoII II .I ..dIII 000.003 . . 000.00? 0 u n d I. I I n a n I I 0 I I III I III I I 5 3 . I I I I I I II IIIIII II IIIII .I IIIIIIII voyII oogooowm + II I I I I II II I I IIIIIIIII IIIII IIIIIII II I+Iooo.oonI u 000.00% M III .I I III I I II I . 000.80on I I 000.002 +. I I I I I. II III I II III IIIIIIIIIIIII II I ..IIIIIIIIIIIIIoIcovpoth 000.00% M II . 000.008 cocooomy 0‘ .00.. 00+ ..o. ... IIIIIbbe.@IIIIIII 00040 I . 000.0 000_IN00_ Amnnmwsozv adv :ofiumazaom cmafiaoa0Luoz zufiu oman coo“ I ombunbwv a 0.....- 000.0 a...- .¢....o....>....o....0..m.omw.r+...r0....¢........ I I. ooo.¢ coo.~.IiII!IIIII:IIII owo IIIII ooo.m coo.m ooo.~ .um~0:m>v< 0>fiuMHo¢ GNPMHOQOLOOZ :« om:mnu ucm an I > mmDUHh 216 owner—mi; o>aum~om . EN; m2; I I305 II . 30.0 I A310 0mm.“ 00w: 03; 000. 033, 000.0 ‘IIIO.OIII§COOO§OOOO‘OIII.OI.OOIDI0+IoII§I0IIIOOIIIIvII+IOIO‘IInIOthIID§hOIIflIIIIIII#tI.IIdIOIaIlIl|-OIIQIHJ§CIIIIII§IIIIIQIO H I H II 000:7 . o 000:? I . <9 x mmz x “I 000... . I . 00049 0 I 000.3 I . 000.2 . I . , I . I I Ixm.< xI III III-IIIIIIIIIIIIIIIIII .I I II II III I I. 0.2 x . II. . . ISIID . 80.3 + :92 x 01E 0.me . 00 0 . «39.2 x N E x I O O o I >x - 000.3 . Imo x I 92 x x . 000.3 J :22 x u .m . I$ 000.0 I 093—523. o>wumaom amounaononuo: Goo" «05... I Anaconoav cowumasuom :muwHOQOLuo—L howu OWL—mu coo“ Nu I > mMDUHn— 217 UN< IlIII|IIIII|IlII|Il|lll||I|||II|||IIIIIIIIIIIIIIIIIIIIIIII ooc.omm 000.0m¢ 000.0mm ecc—6mm oco.0nH coo.oco coo.oom cocooo¢ ooo.rom 000.com c.5-ooot...o+.-..+.....v.. ct. .+.....v...u......b...c+....0....~.-..s..cut-...IvJJIIW-tIfi- ..§.II.II....J|..+.II.. fImnu .m x >mz x <> .>> x <9 0. . omhoo Iflgjx .. 00m.¢ v mmO I. paw...» . NE x x 393 x g; x . . _m Ix . D In N I Q 4. I O 0 ¢ c u *I N o a‘ x , 0m x . IIEI . .56. . ooo.NH IO- IEDE. x 228 x . 0.6.2 om~.MA . cowooa . o0 I0m~.n~ I‘ x I ¥m< x . 0.0.2 meUHm 218 omm.~ §--o§000-6aoo.t...u+oooovauon§uoc~6.uo mfim.~ <9 x ooN.H mmaofl awaacm>u< o>wum~om mhv.c one a coo-o mmm.c omhoo m~¢.c ono.c ..5....+....v....;....o......u..&...uflvnwfiu....‘....+....+ . ”How, x «>2, x >mz x . om» o u ‘ - . omp.¢ H «2.3 x . . E? x . Geno: ; 7 + . . jam x Imam N 93 x . Sm q . _m x . om~.m H final I v CmN.w l H x flm y” . . .53 x . [a x E: x . 50.: 03.3 H . (WT? . .rx x H W oom as H o rum.o~ W . 412 x . w . . n . . o om~.m~ ” 44w x «#3 x . £3; 1. IlIIlII|ll||fl|||llIIIII|||lIIII||IIlIIIIIIIl|||lIIIIIIIIIIllllllllllllllllllllll. . coo-hm o . . o ccoo>m l gjfiwfioz x . , . ‘ a“ . . 22m: . h 2 [I -«m w . . o Swim <13 x +n...+nnuu+-on0‘....Occ-oQooco+-osu+oocn+ 1’! own.“ mhm.fi LON.H mNH.H «No.0 amo.fl u..w....+....o....fl.... 000.0 +....ud...+....»m.nd‘.........+....+ mmm.o owmvcw>u< o>aaudoz caudfioao&uo . . 2 o cg van 009“ :w mowuflu :aawaononooz owhmq :% moawn3mcoz 9:00L0L v— I > MMDUHE cfié m.$.o ccc.g vw< a com.» 00¢." coo.” coo.c compo oooom coo-a OONoH oomoo 00¢.o ‘odI-Iuéuo..+-lnlqlutu.-Iu‘jbunuu 3.0 . he x x «E: E: x . o ocm.q . N_m> x I u . n E x H OmN-x o o Cmn.n . . *2 x (a x . u , .20 x . H 25.3 . :22 x zzwa + 50.3 9 93.2 . . (3.2 N l . . w 2 . . _ o _H_ X 0 ”\M l I I vs. 02.3 . . COWS m 0 11— X t l nom~.m~ . <4“. x E x Domo¢m + ; Owwo¢m . Zak . ........ ..t ..t +.. new.“ 00¢.H 060.“ 500.0 cowoo coo-N 000.“ CON-H QCmoo OCQ.O w. 33:33 £3 :9... .6; :mufloaonaoz 5 SEES “Em m Comm :fi WOMOHU :naHHOQOLOUZ DWLNA CM noawnzlcoz OCOOLOL m“ I > WMDUHh 220 . Uw< Hwnzm hum :muwaoaonaoz a.“ omcmco tam :owumdaom.c.w 0%.; I. A: I > MMDUHm 32 Iamde I IIIIIIIIII I 32$ I I (MCI: I IIIII I 0.: ..I Lcw .u I .I GOC.N bCOoH Orrou vg¢ug JVIDIIIOIflI‘IIIII 9 OIM I O O vfivIIlIt IOII‘, O ...I-I‘vICIInIdIDIIoIIIMIMIO - .IOI‘ I m I 0‘ III.II 05 I O .II-O aImInI IINIOIII O IOH .II a IILII O vIflWII cIIan§IIIIIll I‘ ovu Glofilu D I .II? slolnl. camohl + .. OLmawl IInI |.| . I I. II I II I I I ll . . anHIIIII IIIIIIII I III I II II .I 11111 II IICIF >3. mI III I I I II I I IwILmuflIIII III . >1 x «6sz . mo. :IIHMIwIIExI I I.II w_>.> x x (IE x .I II I & OLmob )z x I. III. :SIZ x 21er x £2 x .I I I I IIIIIdMIUMIx I IIIIII WWWZ . OOOoMA 6 . I VMOKI I ISIK .II . I I..II.I. II I , I I (:3 . < . . I III , I II II. nzIIx III.I -. I I «aIXI Irma x zIzwIUII I- I.I I I I III » IWQZ/x o . 02 x . Don-MN HI. IIIII III.III I .I II I II I I I .I I I I I III I II I ‘ angmN II. J4U X I I II.IIII III.I. IIIIIIII . . . 4...... x I. . 235m I .32 x I 19...". . . . I.IIIIIII.IIIIII-.I I..III:.II III.I.I III I I- III I I I I I I I . Im I a d o n . I III. I I I. I n . .8me I . if: m I III I I O III III 0 I . . u 0 II II I III I III! II III I I I IIIIIIII IIII I II III I I IOI III I..I II . . C O I II II III I elem..~m III I- i i I Emém O I I IIIII I III III.IIII.I I [ IIIII I III IIIIIIIIIIIIII I. I II I I I II I l l {.IIIIIIIIIIIII‘I! I I we: .00 . III II II I II I II I III I >wZI x I I. nIIngqucMIII III ' I I I 'II I I II I I III II II I III I I I I I. II I II I II I I O . . 03.3 I .I ,II III I 5%: 00"... 0’0. I III III] II III I I I I I I I I I II II C O oncotwmI-owIOIorI-¢ooochImdd‘ooImmooou-I IomnIIoquroIoquIuom - 0 II, I I I .. . I IIIIIII H II I mgc IIIIIII II.IIIII_IIII.I II III I IIII 000 N 0003... 0cm." omno 9.1.. I I 5 .I l 2 2 .r 33 3.333..» .3 .3333 . . 1c onuocu>v< 0.1.393“ 3, _|||[l co3o3om 3.3 I .>3 333333oco3vc 3 . 1 3.33 3 «.733..3w.I[3[3 [33. .3 Cmcl.m . r..v 000.0 OOCoh GCuom 3 000 .m LDC3-H [[JUIHoJIanJ3. .0. .3o3.3a,¢[.[-3..-o3 9.43.39. grim-H 9.1.. 3.,ovo3... . .ox. lule3..m.3.+ . .3-34H3...3. woin - u no.- 3. ..v o . o .v u u o .0 s a . .y n . . o. o u n 03.7 . . 3. 3 + 034.3 . . l[[a~d.d:[533df r[[33.:. . 3 .3 3333333 [3> 3.! 3 . w . 1.3 3. . 333 [Q>I?. x I. > (“Iac [ I 3 3 (a . , . 33 .. 3 63.sz x . can; . an; 5,: Ha x . com; 3 33 3 33. .33 3 3.3 Maufi . xx: . rimrdwm u fink. x Mad < . ..3 33 . ... . <3: «3.57:. . r3732 , 3 33 3 3.. -3 3 333333 3. 3-33-; - x 2.. xfia~3[3 t1 |3[3l-[ dmv I v 3: . . 3 3 13.31-. \ [-3[.3\\ .3 \ . - 3% [ . Sb. x , I . 3 d 26.3 33 3 334 x . 53.2. m. I V . I g ..3 3.3 3 333 [.33w3. 33 333 . 13333 3 333 3 3 3, 3 333‘ 3 [3 333l.1 3 . u 3 T. u 0 3 . -3 3 3.. 3.33.3.3 .3 3 3 -3 3 .3 1 [3.3 3 3 33333 - 33 3 33 - 3 3 1 . 33 u . . 3%.; + . . 3 3 53.2. 0 \ I ..[3113 33“ I [3 3 [\ oomomn 933.! I333 ..333. 33.I3 \[3.33 333.[‘.3.33 [ 3.l3[3![ .3[[ 3 33‘} . \ 3 3 l 33 .3 olwwmcmm 094033010 13.13.1333 3 . .33.. . 3‘ .33 33)! 1 3 1 33. I. .3333 w 3 3 33 3 I‘. >wz.w unnono 3 3 3 3 3 .333 , [.333 I... 3 33. I... i + oxmunm ..l3.3.[.3.3;3...13..........:....o.....¢....o....+...:....>..2+....+3..3.uo....3...3..u.. 3 3 . :3?“ :...:..3....... .l[|.i33 _ 3 . 92;. 93.03 can; co<.~ (.3. — 95.0 .. ocot. moonm3 . 9.30.! [ 13.333 3. 3 .- bung—5r“: o>fiummo¢ :muHHoaosuoz :..n 095—5 can :owvmdznom cw own—«cu II S a > mMDUHm I 222 .. 0: 3“. 000.0? 000.0? . 000.30 000.00... 000.033 [I 000.000 000.000 003.000 000.00m 000.00~ nnécvoovunoo&.-oo.vnu...u-o.0-.-o¢nuoc$ouuo+uuo0.00-.30....‘0000000-16-nuofio.ov.o-unn+|-. ...oo....uu-§3 . _m x . 000:; . . . 000:; I V 3C . mzs x .. 000.: . . (004.0 . «a x . . 9.2 . . .3 X 31 33[ 1[\‘ 0| 33 3 00mg; . . 000.2. . 0.32 x . 000.3 . 3 33 .30003.m3m3[[ . 32on x . I 000.00 . 3 35..»A v__ x o S . 3.333 3 3. 3 03 . . . m a 3c 0 l I BENZ. ,. 2 - . 3 . --. ...0w3m.00 a . _ x.m3§_.z x . . 33010. . . N._.m.< x . cocomo o . .3 0....0.Mo n 3 3-. 3-3- .3 [Iisk \ [I3 ..-3.333[!.-333 3 .3- 3 3 «[3[ . <4“. x mmo x . 000.00 M .93 x .. 0300.1. 3 3” . 3 . «is: 3 . >mz x 2K3? xx x . “ 000.2. . 3 NF .3 JHuHo . grow. x <0 x $3. 0. . . . 4.3.4.5... .. 3 .-- . 3 3 53 3 33 3 333. ..33333! 3 1. .-3\[333[[ 1.1.3.2....3.1.1.3....2......................... ......+......... 2.... . 000.03 000.300.. 3 -.3. 0003 033.3: 3 103003.000 . 000.0. H 000.000 000.000 000.00.. 000.000 - 3I[030m.003~1.3333 3 [[33 .I 223 cumucm>u< o>woudom m3; 024 “3.0 30.0 ..:.l.d[[0 03; 08: 000.0 000.0 03.0 000.0 #00oo‘uucoQouno...nootaooofi-0000000“oo-piurn-‘o-u-Qn00-600006-uoatcoo0*..-cfiaocotooou$0000OnIu|uo§oolrloO . . alx [[ 000...; . . . 00m...» H I. C . ms: . 000.: . . 000 E H .40. x - H . .>.z x n 000.: + . 000.0» 0 I . [[ni3x . - 000.~0 + , - .. 030413 [[228 x . 000....0 . as. x «ox 4000.010 . 192 x H l .3 mu 0 n a. 3 | I I. I. I II III. 000.3 . . . + 000.00 m. . :3 x . -- s - 3 o . a U. I 0 [I c 0 0 000.00 . . 000.00 r3 . . xm§3.z x . .., . [M340 x . - I . 3 NEI< x . 000.3 + «30.33.00 . Im<>>3x . . . mmo x 3m x . 00...; . 010.0. x . 08:3 . -3 . .3<>L3. x >mz x ..3 . an? x -33. 22”.:3 x . --3 30.; ¢ . 000400 . Exam x .3 xmKle <3. x a H . . 2&5 x . 3 .-.mu< o>womaoz :muwfionouooz 00¢“ new mean :a maoosom uwanal cw mocovSQm acoohum I [ . S 3 > $50: 224 - I34 uw< “a ooo.omm ooo.Om¢ coo-0mm coo-omw aauuawu oco.ooo 000.com ooo.oo¢ 000.com ooo.oo~ no...on‘o-g.0.nuo§outo+uo-¢-na.+uou00.nocénoun+ 0nov00-o§.o..9-oo.&~u-n...cca§00ou+oonu+-.dlj+.I-JuI-+I-IIII H . - n¢m.o .0 o m¢noo . 9.39 93 . n x x n .2 x m_ I . 3. 3 07.0 . I . 004.0 . 3E x 3 3 - - w. -. . MIC x J . . . wazIx mot-o o - . I. 3...m.o-¢oo . 27.8 x . . Add x . . 03.0 + >m7: . 03.0 H Im<3.x I- H - I . . I.- . -I.._m.-xI3-[ . I 333 I33 .. .-I[.I.I I m¢0oo 0 . 0 m¢onc . N03 . . I I In 3 I I . xw§32 x <>w>> .x .. I a CNN-O .0 5 ONNHC W . QE-x . - w[ H >x x . A mus-o » 3.+I.mi.-.o I ..I I : . . .II 0.30le II. I..3.III3. .I - . . 0:01... 0 <4". x ZZMF x [I II. [Ii II+ opm.o.-..[III . I [I [K x00 x .. 93.0 o . - I . I. I I - . I . Mid-0 . m3 9 <40 .2 3 x . u <0 - - .-II 0~0.~ . . 0N0.” O 0 ...:..0:....o.......o~o.:¢....o: ......0....§....o...-.o....0....o....§......I...o.... I...&3....-03....I. 000.0mm 000.09. 000.com ooo.o.m-N 3. I vegan-mm I.- cco.ooo . 000.00m ooonoo¢ 000.com 000.com . llll vfi< HHQSL Lon unawaonchumz 000“ vca acmafivonaoo hunam mo XGUCH >oczum ON I > mdDUHm 225 annuaa>v< o>fiuuaom I4 msm.~ mNH.H mpa.o mmm.o whooo a 03m; com; 93: 03.0 3.30 ..... I Io-om.o *....v-ac.o...o+....5....o...o+....v....¢..o.&....¢....§|....+.o-.o....+....t....¢...o.voo..¢...d¢:.3. I d m¢m.o + ‘ mqmoo I I . ,3: 9; x . 3“ . x2 x . .IEIDI . .I - owio + :22 N + o~¢.d . . 4m x .. O K 0 . LE0 434 x I now.o o IIII I I I. II-I I >Imo¢onlIIIIII . . 332200.. . I . Jmz x o 03an H 3:23.. . I .. I ”I . II . . E x I . .u. - . m¢o.o o J.n¢c.o .II [...I .I. . N_m¢ x o H c o E I..II I ul. . . w . ..meZ x § x . x o~h.o + h I o~h.o . ore-x .- I II N c, x H 32.0 I .3330 I . I562 x II 3.3 I 2.30 A. . (nix zzwkx . 05.0 . xmd x . - 3.5.0 . 3 I <4 .1396 . «.13 x mfiow x 4o ~ . 03; . .Iowo: ‘00no‘ouo~§no006-.-IOOOaOOOI-06..ao‘oon-Ooo-ooé-.hn§oooo+ocoo+noo0....0*..0063000‘10.o'Qo.-0Oooo-Oocoo.I mxwofl II - mNAwHI , II- I -mna.p.I-II- -IIII]m~@.bIII!I:I. IIIJIMFoqn II. I I .II ommJ cow; own; . .ooo.o 0330 oooé Il- ou~a=u>v< o>wunao¢ :auwHOQOLuOZ oco— can dowkuoaaco manna mo xoucH hoccmx an I > mmDUHm IlIII|I|lI|lI|llI|II|I|l|||l|||llllllllilllllllllllllll < oomua oo¢ou 6M coo-H hoc.c oo~.o W. cocoN cocoa com.” oom.o oo¢oo tin-DUOIOnoO-to-fiooo-bnsoofio .090 0.9..out...oOonoo+cvanvoanOnooo+utno+n9-§u ofionnc+ccofvntoo§¢otu IIlllIlIIIlIIIIlIIlIIlIlIIIIIIlIIIIIIIIIIIIIIIIIIIIIIIII!IIIIIIII|lII|II|IIIIIIIIllllllllllllllllllllllllI||llllllldllllllllllllllll m¢moo b n o~¢.o 33 i cmq.o I6 + m0¢oh D 2! 2: C) C) X 6 3.3.0 . Q4 0 9g 0 o . . . u . . 1; x . x . u 330 . . <3 Xmsrz . ..:..“ w . i . A £05! >¥x . O O n .I. n m¢uwx vccmm Illll «N I > mMDOHm 227 coo-m 000.0 awaoaw>von>fiuaHum mivo an I > masonm Q ..:......+.......................:...2....+Iu...-J...I..+......... 000.6 000.0 ooo.¢ COO-N 0.0 000.0 coo-h uuo.m coo-m Loo-H I ooo~vuoon gonna—=24; 355540“ :mafiaononvoz cw emu—5:0 ucm cowkuoafiou hug—mi mo xouCH zoccuz 228 F vw< 000.0mm 000.0m¢ coo-omm 000 OmN 000.000 ooo-oom 0c0\00¢ 000.com 000.co~ 000.00 o h 000.00 . «Ed. . . S. K . 000 cm o-Oo- uéu-o-Oonou¢ca-.+-onc6c-ou‘anvo5...-§ on a oofiooo 0‘ ‘ €00.09 oow.m¢ + I OLNunw III|I|||I|IJII|IIll|I|ll|l|Illllllllllllllllllllllll . mdo .w x . 03.3 . I 00 o r . . ..3-.5 x . . , IIIIIImKozu as x . IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII oo¢ooo .ou “I 0 I 60¢.oo . o m. . ..:2 x . m 0 III I r\ . e ‘ a a1 o .I. .. t a 03.3 I . {0.3 m. . 34.; x x . u I H - a 2200. . 03.90 H . I + :33 + (cmooo 93 x H NE< wz . . 30.3 ..I x >3 Ix . ”>2. «m x . . 5.5 x . 0no.00H + curvoo~ «co-oma 83 5:63 2;.ch 92.9; .- I . \I\ UH< Hfinzm hum cmowHOQOLuuZ moo“ vzm :OflawuanOU >vhmm Noam Ill 9N l > mmbUHm 229 aways—9:3. o>wauaom . thJ mNHJ whouo mmm.0 whouu 0mm.“ 00m: omoJ 000.0 om>.0 000.0 .I.... .... .......... ...........+........... ..:.... . .UIflddfilIfljfilI 000.00 . u 0.0.0.00 . IKE E . . . . . 02.90 . . 0.2.00 . . oo~.o¢ v . 00N.oo . . . . IEoIm x . . 03.2. . 0 . Em no u 3 , IEEC H 00.300 . . cufnw . D . . III o 03 on. . <34 x . cc. co m . . ..HZ x . .w o . o a... r. n t I. . . X . 0 000.90 4 u F + I I. . 000.3; o:..+.........+....¢.........+....+....+...:..I..................-.II.+.....:...I.....-...u.....¢...... mhmfi mg; 5.0.0 mNmé . m3)... 0mm; cow; I ,0ch 000.0 omio 9.0.0 oWaucm>u< o>m9ma0¢ enounaononvoz 000" can :oaawuoafiou .3725 500— MN I > QMDUHn— . I 230 vfl< ood$zmm . 02.03 505mm ouauoww 03402 08.0% 000.com 9.3.9} 80.02 98.03 *3 '92 u.ub..nuu+rnoow.¢:nro.\ot4oouo+..ol-.-0.H.o..m|-00u.o...so.n+ouoa}.muattoolu>u.41.“oHulolm.nI-J.iou.qolmlfidl¢lu¥9.5..«4dbonIH-Huuo I - . . ...i- - , 1. , £50..”VA . ¥ . .>Z .-v.. .I-)- -K . i - 1-1.2 ”11-..: 1.. \ “““ 55.53-11,-..21 . . - .4on 83m . :22: - 0....-.3, i- .. 00.060 .\|IJi +1... II - y I . O Famine . . f! , . . . . . ““““““ o o 03.7... .9 . w - .. 1 . .I 11,9. UWmloNn‘\l..}- .. - o . . . 21;- . t..- . . - .. iw- . . 11-.-}..- -1- 1---, . Ii-.. .. . z .I.].‘JI -..I. - w J 2-- I! -I--.. .. \. WEUHK Mobvl. 52.....‘3 Sodom . . l . ooofimm . .- , . . 1--..-. $1.15} c3500 000.com 0.5.03 1.... 03.6.3 3< 395 non. 533330: $2 3.... 53333:; we xounu 5951.: on .. > $32 . - 231 4 --3: wumoau>v< o>wawdo¢ .1 mNNoA m-.~ . mhwmo mNo.o mh0q0 . -II I! own-a CON.“ omo.H coo-o OmNoo noooc ’1 on.m.-.§.c.o.:o|HflInx-.n§ounloHl.lnouO-onus...-cvo'flr-Oooo..Iciun+nodlotlrumoéalqijbadpj+dluljti‘jo 30.0 o 9 30.0 I 5.4 ‘ .“l‘ ..... oomMM ..+... I . .w- r . . I. . .. .. . v n.0m.h II .. . . «:0 x . n m2 w Iowmvmw . Iii]. .2.on >x 6.; x NAM—aux Ono-H own»:u>v< 0>womaoz Canaan's—0.502 000" «E... :oaamawowuhmn— we RovsH :pmhnawz NNI> meUHm 232 I.. .III.IIII . I | l I I I 3 h 9;; 8.2 < IIoIIIoo Iqoqfi coma 00N0~ cow-O oo¢oo § .,o c o,» 0 o u a‘ n o n o. . n n .‘v o 3.. quI. .9 alpl-quIcIfl-UJI-JIqu-Jlfiquqadd‘IuI-InIuIIIuIdI-IAHJI-IWq-I-.idIIIIIIIII ooo.~ I .OI§.IIOOO.M IOrII‘ IrI IUW c. O O O o 0 0,6 t 000.0 d , II . ..... I . I II. I . I I I -I.. . I. II I . Iona-M .I. I . - I II II... III+.mI.m.~I .- II ..I I . «.ox . IonDHMN u I I 0.5.9“ . «J? . .I- . I. m S... x 22m: H - .monumeIIIr . I I. I-.- I I II I I I I I. . .. CmémI I I I I.- - I. -.I I , I I. I Imv_ xmsmz x n 80.3 m I . I .- . I.I- . 122-w .II _ . I.. -I.---I. I .- .- I . 3.....- I I . - >2 I} 2;. , I . . E x #3 . I ..I . wmo x «>2: >mz . . .-.ohowIlI---” I. I I I -1 - .- I -I I-- . @333 .-.I . pm x 228 x . . J9 x .3913 . . 00n.ho . I u I . . I .II III . -.I . I III I..-IIIIII. I. . I I . I II IIIII.II I I III-.I. ICvm .50 I . I . .I ..I... . I :III .II- . I - I I -I- -II . .I I , . 29.5 -..I .I. I , I 0+.aaoOo-on+nnrrrfio.ow+.o- III I I: .. com.” 3...” . coo.“ oooé ooo.~ I -.-Io-oWJ.» II I.- ooofiluoofi 45¢ Hung—am Lon :mOMHOAOLvox ca 09550 9:. :Ofiamnnoa le> HMDUHE 233 cool“. I coo; oMauu>u< o>wuaam 05... 08.... I I- 06 Ofimuh .r I I I III I I I IIIIIF - I IIIII I.- I I IIII .9 OIOIm .FI I III I . I .I ..II- II .II... .I. . II .II .II.III I . I..ZI Ix I- . .. . . - . I II 5.2 x I. .I . I I II I.. I III- - II. I I I I.- I. m_>> x 2 x...- I I. .. . .I .II I I 28.3 . Im<>>x In: I 8me . 130.... I I I . I .I I I IIIIII ,- «n. xI-II . .. >mz x . mmox . /.IIIII -II ..II . I I I III III- I I III I I. -I.-III. <>§ I I II I -III 303.00 ..I III I II - o 000.00 . zzoo -III.I.I I . II I- . III I.. II I I I I I .I I -II .E I.- . I I- II M, H/ <9 x :5: x I . III IooInImMoIIIIn .II.IIIIIIII I I I . I II II I I I- -II ...IoquNoIII 234 000o0m¢ Goo-0mm ooo.Cm~ Goa-awn coo-coo Ooouoom ooo.oo¢ 000.com oooooow ..0....#u...¢....t....0...o+.o..+oo..0.a..+..o.»...c+..uoyodu.+.u.o&...fiH.%4..m44d§q€¢dfi€fi41fld44ufldHHIIIIIIIIIIIIIII echoed 9 ES Lw 000.0mn I _ . - ...J...“ coo-ma <.._ x 000.3 00‘ 02.3. I. . .IIQIKANI mmz xx , 1 NE< x g x . . lad-OI? x I 3.12 x . mmo x . )1 N 7 coonm¢ .00 930A JOUJOAOO Oom.mm .0 XmEIZ x . n OI.>>x II I _ x. . 000::- ,.. Im<>> x E x m 3 IIII0-0Imaom 0005p . TEE? x . I-.I ¢ 000.0% 000.03 000.com 08.0.3 000 000.000 000.000 80.03 000.com 000403 I I 235 canon—55¢ Egan-mom -. m5: mi; 30.0 30.0 30.0 , 00m; 03$ 03; 000.0 03.0 000.0 ...22............:..........:... ..I....I:..+........:...:...{..II.IJ:2.2.32.3..2. 02.3 I . I 03.2 DEIIE I I dwx w 000;: 02.3 I M ommfiw 22m x . 3n + , . I I onoé-m , . II+ CARES <13. x >¥ . Don-muo‘ y :00? 02 x .- . -_ 9 w o. J . _ . w . . NEG. x . M¢uz mm?» x . J. . x .. A 0 m IIImmEPII 3me I E130 x x x H 2200 x022 x I I I. -I . )3 o o I. x I E x 93 x I .. -II 03.2 I wmo x E x . 0%.: .5 D X I E 0+ boo-09.00.050.06030u0-ooofioonofiuooofiooc-fin00.00.0-0 .32............+....+.......2:13.232?» I I Eula IImIE Ith-oflo mmmé . .Ithchu I .III.II.IIII I 0mm.H . cow-a Om”... (no.0 OmN-Ju 0000.0 undue—NZ: o>fivaaoz caumaoaonuoz 3.0— 65... mean anchmoz ovo> LOCL®>OU . _m I> NKDUHE 236 , Ifl II . III 02 - I» 08.03 000.09. 4d?” 000503... 04.0403 000.000 000.com .. - 000.000 000.0% 000.03 tn‘mIaooVIprIIoIoI‘o-IWIMGtu..V‘on. cotton-«V... 0n-ulm+b.p-JququI-bnfldIuIUI-Iud0HdIIJquIqI-Inlqufidlulqlvl1-I-Iqt.I-IoI-fiquI-IcuInIAI. III OOOuhH § o (usqNH . II.I - II I mmo x . O I I III II III 000.- .I I I. 000.- II II i I m;> “I. I .I I I II. I 00.0.3 I. 34 ,0 . .IIW 4m. III I . . .. , I 220qu . I bopnwm . ..I II... 0 II II I I.II0II0I0.I~mIIII. . P7. . . IE0 x I. I I ,u7_ X I 00mm: I 923 I. I. cIfiAmII . OI x xm< . . I 000.? . VIZ N_m< 0. I .I “I ..3 I II IIII .x I II I . II I III! I II I. I I III .III.I-IzzmkaIII. III .II I I. II. I. . ..II I I I I I ISbIIxI I I. ..II I. I .III.II .III.I . I 000.94. . . 803$ n 5.. x . 30d: I ..I I I I- <4»: I I III I. 1.59.0.3 III I . 5 x . .II II . . 53 m x nmss x < __< x . . . . -. ..I. . II . wa.zI0.A I II<@.x I H II I . I 0.00.”? I.. I. I. . . II I I. II I III I III .. I.. I I I.. . I I303» I . H mo¢ Hoozum OflHnsm Haven. «0 9:00.00“— ouuvm NMI> mMDUHm 237 oluuaa>v< Simonaox I Imwouu mwm.m whond :nUIH 060.0 omhoo 03.: I I I Eu}: I. W0 x .III I III. I 36.3 I . I 201% I H I Imzsix . I“ III oooIIIhINI III I I 7.2 x I I IIIrIIE§mIIIII . Q I I II I / 2200 x I III I I .II I I II IoIoIOImeII I II II III I . III.IIII , I. .IIIIIIIIIII'II IImIIIquINm IIIIIII . D _.m x . II . I... «6 x I . H fiz x H 03.; I. I 92 XI. ..III.I...RI “ 9.3: H I.. u onwwmfl .I III.I I I. I I.. I I I I II I .II III III I coed: m I D I I IIII I I 0 IIAIIII . ..I I .I I. 10:2 V... E x .I I I I .. I. I m. n _ <> 2, x xx? H n 3%.: III .. NEE x I an: - ”I II>zmI I .I I III I. .225. XIIIIIIII .II.III- I I II . I --.II IIIIIUI .I ..... II I I :53 I I I I I I. . I I .III ooo.I~Im Io . I ouo.mm . § I I Icoousw I... I. I I . 13.... x III III I IIIoII,IIJ.m III III. . . mfivaao¢ :mumaonofiwo: 003 can II . 00¢“ G..— moscv>o¢ Hoofivm Dada?“— Hmaos Mo 0:095“— ounpm mml> meUHn— 238 OQmIH 00¢.“ Ufl< ddo.. moo-o 0QM4O ooo.~ oamoa oomoo oo¢uo ..I‘IoI-IoInt..notIv-uIO-o.a‘. cbo.u.‘oonn..quInII~c § or...dedJouI-IdchWHjOHIJIvIIvI.JIqJ>-III-IHI>-Iuljrnlulflq 08.: I .II wmo x cocoww + . 93 x 80.: I .12 x .I. IIIZI.Z oobommIIII + x o oowom if . I u , : . 220.0 x III.III I II III.I . >Imz x U IoImm..~m I III. I I 02 xI I IIIIOIWIZNMI I I I III I III. I IIII I'd II” 82.9 x - . I m Iooo.~¢IIIIIII I III... . I.. I. ..III.IIIII ..IIIIIIIIII: III III. I I.. I I 505... w I (na II . ...I I II III. IIII..II II III S III. II I.. III IU=2Ix I. II v—maII x I I .I m. a X X . IIIII . .<>.>> . m 89.5 I NE”: I QILJII a II --.II I ”I I .I I I- $.sz 22m;K .. I. . .II.IIIIIIIII. .II I .II II . I I I I 1.3.3 x -.I I... III II I” I I 8pm.“ I . I 95.9.. H *M x .. IoooIman. II o (IEIx. I II I III I III III. II II IHIImunIHmIIIIII 0 II . I. I . <4 X c I m mMDUHm 239 . I .flII .. .ouaucgg 2333 . . one I I e920. 8c; . I .ocIoJ .II I II .II. .I II.IIIIIIIIIIIIIIIII 0030NN a I. I. II I I I I I I I II I II I IIIIIIIIIIIIII I I I II.II .IIIHIo.III.I IIII III.II! III-IIIIIIII.II.IIIIIIIIII III III... I II I I. IEo.~I~II IfIIIIIII IIIII I I. I I I. I. III. I II . . 93 w I II.II I .II . .I....IIIIIIIIIIIII I I I I I I I I - ¢ 0 (J O l N I“ II; <1 C) [X I I I I l I I l I I I I a? ‘( l I I I I > DJ 2 x; I 23.3 I- II III III II I I I u 92 x I I I I I II . Isnfiwi III {III ..I EJII III II I- I I. I. . m 85.? I III I III .I III . I III.IIIIIIIIIIIII . I . I I. I I I I I I I I ¢ Ooo.N¢ O - . II... I II .I I. wIII- . I I I : I I I I I III ...I.>.>> x .192 . I w J . . I an x x xm< . II o I I2. 3 I I I I I. . I.. . III”. II I II n; u .>sz N??? I 08.3 I I u ...I. II.II. .II.III... .. . I . I 22m: . . I IIIIIII II I I. I I. I I 14.—bx .. . 03.3 . . . . I I Mva II III. .I.. III IIIIHIbICIcIINMIIIIII 30.3 H <4... x I. II III. I I II I .. III III j/ I II II. III: <4 deOfi H ovoib .I III I <4< x :mis Y. I I I II I I - I a II I. .I . I I .I I I I . I. I I II I I I I . X t n I833 . I I . xmz 2 <0 . I I- . I no 0.3 II I III. II-.. .. I I I $.32. .xI. ..II-III 0.5 navauont-aocOI-oofio.action-‘.o00.noookunn..I-u. con->0... a... ...I n u o I III: II-IIbEwmI I 8.3 .8? I I Sow. +0. I. coo-N moofi ..:." moacv>oz ..Ioonum owanzm awuom. ho 9.595.“. ouaum mm I> $53... I II IIII 1 061 370 III.II. III MICHIGAN STATE UNIVERSITY LIBRARIES 3 . warYXEIrcIuil :3: .1... I:.. .. l..::.. 2.17. Tic. ..I. I ..:.I.....I:._ :f . f... ._... f . ..:. . . I L. .«va. ..:...T. ...I». 15 $7223“; . 7x . 1 .: ..:.; .I..»m‘vx. ; _ “In .r ”our...“ .;»I I _z. \ Liv‘Irigk' v;- I...