THE EFFECT OF JURISDICTIONAL BOUNDARIES AND INFORMATION SUBSIDY ACTIVITIES ON THE PERFORMANCE OF SECTION 208 (PL 92-500) PLANNING By Alfred C. Hamilton A THESIS Submitted to Michigan State University in partial fulfillment of the requirements fer the degree of MASTER OF SCIENCE Department of Agricultural Economics 1978 ment can dually a tion is multitud good). and info tlon in ‘ effects ( Process c examl'ned. Hide 5qu fr°m poli ABSTRACT THE EFFECT OF JURISDICTIONAL BOUNDARIES AND INFORMATION SUBSIDY ACTIVITIES ON THE PERFORMANCE OF SECTION 208 (PL 92-500) PLANNING By Alfred C. Hamilton This thesis addresses the question of how a large unit of govern- ment can get many smaller subordinate units of governments to indivi- ' dually adopt programs which are in the national interest. This situa- tion is similar to the problem in economics of organizing a bid among a multitude of potential benefactors for a high exclusion cost good (public good). The problems of jurisdiction, transaction costs, free riders, and information costs are all relevant to the situation. The Federal Clean Water Act Amendments of l972 provide a situa- tion in which a natural experiment is carried out. In particular the effects of the jurisdictional boundaries of the water quality planning process on the probability of local control of nonpoint pollution is examined. A negative relationship is hypothesized, tested via a nation- wide survey and validated. This thesis draws on relevant literature from political science and economics and concludes with policy and research suggestions. great]; Dr. Eci politit as a t6 economi major a but not Hainrit assista Agricul Researc Additio have be ACKNOWLEDGMENTS The author would like to thank the fOllowing people who have greatly assisted in the preparation of this thesis. Firstly, the members of my committee, Dr. A. Alan Schmid and Dr. Eckhart Dersch. Dr. Dersch assisted with his insights into the political aspects of water resources. Dr. Schmid has affected me both as a teacher and philosopher; his contribution to my understanding of economics has been invaluable. A special thanks goes to Dr. Larry Libby for his support as a major advisor and friend. He has allowed me sufficient room to struggle but not enough to drown. I would also like to acknowledge Ms. Diana Hainrib for her support over the last year and her invaluable editorial assistance. Finally the author would like to thank the Department of Agricultural Economics and the U.S. Department of Agriculture, Economic Research Service, Natural Resource Division for their financial support. Additional recognition is due to Dr. Lee Christensen of the Natural Resource Division for his supervision and assistance. While all of the previously mentioned people and organizations have been instrumental in the preparation of this manuscript, any re- lnaining errors are the sole responsibility of the author. ii LIST OF LIST OF CHAPTER I. II. III. IV. TABLE OF CONTENTS TABLES ......................... FIGURES . ....................... INTRODUCTION ...................... Problem Statement .................. Purpose ....................... Organization of the Thesis .............. A CONCEPTUAL MODEL OF NATURAL, MARKET AND POLITICAL RESOURCES ...................... Characteristics of Natural Resources ......... Allocation of Natural Resources ........... Natural Resource Planning .............. THE MODEL APPLIED TO WATER RESOURCES AND THE ROLE OF INFORMATION .......... . . ........ Characteristics of Water Resources .......... The Economic and Political Costs of Improving Water Quality ................... Means of Reducing Nonpoint Pollution ......... The Role of Information . . . . ........... METHODOLOGY ...................... Natural Experiments ................. Sample Frame . . . . ................. Survey Instrument .................. Variables and Hypotheses ............... Case Studies ..... . ............... DATA ANALYSIS AND RESULTS ............... Response Rates and Nonresponse Bias ...... . . . Analysis of Differences Between Organizations . . . . Within Groups Analysis ........... . . . . . Analysis of State Agencies ............. Analysis of Regional Agencies . . . . ....... iii VII. REFEREE Conclusions ..... . ............... Limitations of the Analysis ............. VI. SELECTED CASE STUDIES ................. Introduction ..................... New York State Department of Environmental Conservation .................... Central Upper Peninsula Planning and Development Commission ..................... Discussion ...................... VII. POLICY RELEVANCE OF THE RESEARCH AND FUTURE RESEARCH CONSIDERATIONS ............... Structure of Section 208 Planning .......... The Role of Information Subsidy ........... Time Horizon for Section 208 Planning and PL 92-500 . Future Research Needs ................ Measuring the Output of a Planning Process ...... Design of Future Resource Planning Programs ..... REFERENCES .......................... APPENDICES .......................... A. PL 92-500 SECTIONS: lOl-lO4 AND ZOl-2l2 ........ B. COVER LETTER, SURVEY AND FOLLOW UP LETTER ....... C. SELECTED CHARACTERISTICS OF SECTION 208 AGENCIES . . . . iv 64 65 68 68 69 73 78 TABLE 10. LIST OF TABLES Differences Between Jurisdictional Boundaries of State and Regional Section 208 Planning Agencies ....... Differences in Administrative Functions ........ Public Participation Hearings, Public Participation Budgets and Public Participation Staff Per County and Per l,OOO Population .................. News Releases Per County and Numbers and Percentages of Agencies Preparing News Releases ............ Agencies Making Special Presentations to the Governor or Governor's Staff, State Legislators, County Legislators and City and Township Legislators ........... Number of Special Presentations Made to the Governor or Governor's Staff, State Legislators, County Legislators, and City and Township Legislators ........... Correlation Between the Number of Counties in the Planning Area and Information Subsidy Variables fbr State Agencies with an Implicit Linear Relationship Correlation Between the Number of Counties in the Planning Area and Information Subsidy Variables for State Agencies with an Implicit Exponential Relationship ...................... Correlation Between the Number of Counties in the Planning Area and the Information Subsidies for Regional Agencies with an Implicit Linear Relationship . .............. . ...... Correlation Between the Nunber of Counties in the Planning Areas and Information Subsidy Variables for Regional Agencies with an Implicit Exponential Function ........................ Page 47 49 51 52 54 55 59 61 62 63 11. 12. 13. 14. 15. Selected Characteristics of Section 208 Agencies Responding as "Designated" ............... Selected Characteristics of Section 208 Agencies Responding as "Non-Designated“ ............. Selected Characteristics of Responding County Section 208 Agencies .................. Selected Characteristics of Responding City Section 208 Agencies ...................... Selected Characteristics of Responding Multi-State Section 208 Agencies .................. vi 119 120 121 122 123 LIST OF FIGURES FIGURE 1. Easton's Model of the Political Economy ........ 2. Modification of Easton's Model Including Jurisdictional Boundaries ....... . ............... 3. Matrix of Political and Dollar Costs at the Federal Level ......................... 4. New York State: Designated Areas & Agencies Funded Under Section 208 of PL 92-500 ............. 5. Central Upper Peninsula Planning and Development Region ......................... vii 3O 71 74 3 st dOCu Shar an a C0nsi me Of 0! trans the S f1r5t CHAPTER I INTRODUCTION Problem Statement It merits particular attention in this place, that the laws of the Confederacy, as to its enumerated and legitimate objects of its jurisdiction will become the SUPREME LAW of the land; to the observance of which all officers, legislative, executive and judicial, in each state will be bound by the sanctity of an oath. Thus the legislatures, courts, and magistrates, of the respective members, will be incorporated into the operations of the national government as far as its just jurisdiction and constitutional authority extends; and will be rendered auxiliary to the enforcement of its laws*'(Hamilton, Madison, and Jay, l96l, p. 22l). ’ The Federalist Papers established the philosophical base for a strong federalized system of government. At the same time this document recognized that the sum of all governmental powers would be shared between several levels of government. In a federalist system an anomaly exists when, in order to further some national interest, the consent of the units of government at some level below the federal level is required. This paradox is similar to the problem in economics of organizing a bid for a high exclusion cOst good. The problems of transaction costs, free riders, and strategic bargaining are virtually the same in each case. i: . .Federalist paper number 27 attributed to Alexander Hamilton, first published circa l787. -—4 CO CO; Wi‘ In trying to obtain the national clean water goals set forth in * the Federal Water Pollution Control Act, Amendments of 1972 (PL 92-500) such a situation arises. These goals state that: (l) it is the national goal that the discharge of pollutants into the navigable waters be eliminated by l985; (2) it is the national goal that wherever attainable an in- terim goal of water quality which provides for the pro- tection and propagation of fish, shellfish, and wildlife and provides for recreation in and on the water be achieved by July l, l983 (PL 92-500, Sec. lOl(a)). In order to achieve or even approximate these goals both point sources and nonpoint sources of pollution will have to be controlled. It is the control of nonpoint sources of pollution which presents us with the federalist's anomaly. It is almost axiomatic that nonpoint sources of pollution arise from the manner in which land is used. To reduce nonpoint pollution, the way literally millions of people make land use decisions will have to change. The anomaly exists because fOr various historical and political reasons the power to control land use lies principally with local governments (cities, towns, counties and villages). It will be the consent of these units which will be necessary to achieve national water quality goals. Purpose The purpose of this thesis is to examine this anomaly in the light of an institutional economic framework. In particular it will attempt to measure the performance of various institutional arrangements which have evolved from PL 92-500. In doing so, special emphasis will * . See Appendix A. be pl level to in devel resou this econo Drovi the j outpu' be placed on the relationships between the federal, state and regional levels of government and local governments, and their relative abilities to interact and exchange information. Organization of the Thesis In undertaking this endeavor it will first be necessary to develop a conceptual model of the process by which rights to natural resources are allocated. A model will be developed which looks at this process in the light of both a market economy and a political economy. It will also be shown that planning plays a key role in the provision of improved quality or quantity of natural resources and that the jurisdictional boundary of the planning activity will affect the output of that process. Next this model will be applied to a specific case, water re- sources and improvement of those resources through reduction of nonpoint sources of pollution. PL 92-500 will be used as a vehicle to allow an examination of the effects of jurisdictional boundaries on planning activities. Through section 208 of the act, which calls for pollution reduction on a nationwide scale, a natural experiment will be conducted to test how information exchange variables, which gauge the ability of planners and decision makers Dainteract, will be affected as the juris- dictional boundaries of planning areas change. The data used to test the hypotheses generated from the con- ceptual sections of this thesis will be gathered by a survey of all the relevant population of section 208 planning agencies in the country. This data will be analyzed and comparisons will be made concerning the differences in information exchange activities between the two major oi CL P!" re ce PE pr OR of DH an. Dre types of section 208 planning agencies. The extent that variation of jurisdictional bounds affects each class of agency will also be dis- cussed. Because a multitude of other factors enter into this process two brief case studies of section 208 planning agencies will also be presented. These case studies are designed to further acquaint the reader with some of the nonquantifiable factors that affect this pro- cess. These factors include agency credibility, history, and self- perception. In the final chapter policy prescriptions based on the material presented in this thesis will be presented. These include observations on the nature of the planning process, its time perspective, the role of information in changing public policy, and the structure of the present section 208 planning process. In addition, in the spirit of any academic research a future agenda of research needs will be presented. syst P850 resol miner diffi PESoU COmmO SOCle Charac CHAPTER II A CONCEPTUAL MODEL OF NATURAL, MARKET AND POLITICAL RESOURCES Every society faces the problem of allocating natural resources to a multitude of potential users. In this chapter some of the salient characteristics of natural resources will be discussed. Two systems of allocating rights to those resources will be described, and then the relationship of the natural resource planning process to these two systems will be discussed. In so doing a model of the rights to natural resources allocating process will be developed. Characteristics of Natural Resources Most people have an intuitive idea about what a natural resource is. A list of such resources might include water, air, timber, minerals, land, wildlife, and possibly energy. While it is probably difficult to find a true common denominator that links all natural resources they do appear, to greater or lesser extents, to have some common characteristics. They are the basic raw materials which any society must draw upon to engage in production processes and indeed for human life. Natural resources may be either renewable or nonrenewable. When they are renewable, improvements in their quality or quantity are characterized by natural processes which occur over long time periods 5 na pr wi Pd de of du.‘ 90: the ”Oi Ddt and/or extremely costly investments by man. Finally it should be recognized that the users of many natural resources are highly inter- dependent. That is, in the economist's jargon, for many potential uses they exhibit high exclusion costs. Allocation of Natural Resources The marketplace is one forum where the rights to certain natural resources are transferred. However, the high exclusion cost property of many natural resources suggest that demands for certain uses of those resources will go unannounced in the~marketplace. Goods with high exclusion costs are those goods which if purchased by one party may still confirm benefits to another nonpaying party. National defense is probably the most common example of such a good. Typical of high exclusion cost goods, there is little incentive for indivi- duals to go into the marketplace and exchange their resources for that good if they know: (1) they could benefit if someone else purchased that goodinputs political . ees outputs-— system & actions Support — — -> (Easton, 1965, p. 32) Figure 1. Easton's Model of the Political Economy Most systems which make authoritative allocations of rights to resources are strategied into a series of dominant and subordinate de- cision units. Each decision unit will have a corresponding jurisdic- tional boundary, which is the area in which that decision unit may make 11 authoritative allocations. Dominant decision units contain, as sub- sets of their jurisdictional bounds, the jurisdictional boundaries of subordinate decision units. At the same time differing groups will find that they have differing relative resources with which to express political demands at different jurisdictional levels. Thus, the rules of dominance between decision units becomes an important factor when trying to predict the resolution of a political demand. That is if organization A dominates B then if contradictory decisions are made with respect to the allocation of a given resource in those organiza- tions, A's decision will be the one that is enforced. Of course, things are never that simple. More likely A has precedence in some areas and B in others, and where conflict exists there is a good deal of room for compromise. It should be remembered that decision makers, at least as con- sidered in this paper, function in such a way as to maximize support. Bartlett has described this in a democratic context as vote maximiza- tion (Bartlett, 1973). Any decision is a two edged sword. Those who benefit, at least in theory, ought to give support to that decision and those negatively affected ought to deny support. Thus it is possible to conceive of a calculus of political benefits and costs. We may then speak about a decision maker's opportunity set as the matrix of political costs and benefits that are perceived with respect to any given decision. In this decision making calculus, decision makers will attempt to reduce political costs and increase political benefits. One of the ways in which political costs may be reduced is to get a subordinate decision unit to do the dirty work, thus insulating the decision maker 12 from some of the negative impacts of a decision. This process of shedding political costs is intimately related to the rules of juris- dictional dominance. The shedding function may be described as the process by which dominant decision makers may effect the opportunity sets of subordinate decision makers. This nay be an over simplifica- tion since so-called subordinate decision units also have indirect means of effecting the opportunity sets of the so-called dominate decision unit. All of the elements of the model of a political economy have been specified, and Figure 2 diagramatically represents their inter- relationships. This model is essentially an adaptation of Easton's systems model of a political economy. The environment, represented by the box furthest to the left, consists not only of existing resources and technological relationships but also of the social relationships which define the rights and limitations of individuals and organiza- tions. Changes in that environment can have a direct impact on the welfare of individuals and organizations. Individuals and organizations provide the basic inputs into the political economy. They make demands of decision makers and provide them with support. They use these political resources at various jurisdictional levels, represented by the dotted lines in Figure 2. The jurisdictional dominance pattern is such that jurisdiction one (J1) dominates jurisdiction two J2 . . . dominates jurisdiction n (Jn). Each jurisdiction has a corresponding set of decision makers, 0], D2 . . . On and each decision maker has an opportunity set which consists of the matrix of political costs and benefits with respect to any potential 13 mowgmuczoa chowpowumwean acwuapocw Pmuoz m.:opmwm mo cowpmuwewuoz APV ucaom chowpowumwgzw fi1 IIIII 14 _ _ A.Vq 1vh n . :GA......ANGA_6 mucmcwsoo _a:.po.u...=q mpmm zumcngoaao swmga use mgmxme :owmwumv ucmmmcgmg o mmmxz .N mgzmwu a1. mun fi11 1.1n.1w.1w. 1n.1u.1n.1u m a. IIIIIIIIIIIIII _ > __|l __A111||| z -_ 11 : _ucoqaam o 11a meowmwomo o _a H co _ .2Mucmsmo % _ _ r1. IIIIIIII 1.1.1. N __ 7 __ H __ m __ a P No 11mL cwocmzm __ o m _ _ m _b _ r$1111111 o z . 1.... N __ z _ . . o o _ a o _ T a __ __ m SH, <__11111111¥111L_ ._ z < _ mcwuumcm _ 2 m _ a _ M. F _ mcowumuoPF< yam rA 1. > a_ m>wpmuwgocp=< xuwczp _ugoaa=m H .1\ -Loaao o _ 11 .mcowmwumo F _n z _ o .mucasmo a < _ _ r 11111111111111 11.— »socoum FMUPHPFon op mpzacw am will 14 decision. The decisions which are the outputs of a political economy are authoritative allocations of rights with respect to resources. These allocations change the environment facing individuals and organizations. The shedding function is represented by arrows which point from dominant decision units to subordinate decision units where the dominant unit has the ability to change the opportunity sets of subordinate units. This model is essentially static because it does not deal with uncertainty. Before this limitation can be diminished it is necessary to understand the role of resource planning. Natural Resource Planning The preceding discussion, has suggested that natural resources are characterized, if they are renewable, by regeneration over long time periods and/or with large capital outlays. If they are non- renewable then it would be reasonable to suggest that they need to be managed or allocated over long time periods. Because of the long time periods involved and the potential for large capital investment, future states of natural resources are highly uncertain. Both the private and public sectors attempt to manage uncertainty by planning. A comparison between the planning process of these sectors may shed some light on what the outputs of those processes are and how they differ. Firms seek to expand their time horizons and in so doing, manage uncertainty. This is particularly important for large firms in which the production process may take several years from start to finish (e.g., automotive, aerospace, or energy industries). These firms may try to circumvent the workings of an uncertain market by vari dive fin: acti futu may this indu work choc Doss In 5 the t0): and 15 various means including influencing tastes, contractual arrangements, diversification and market control (Gailbraith, 1967, pp. 22-34). The firm has an objective function which guides its profits and planning activities. The essence of this process is that a firm chooses a future state of nature that it would prefer and a means by which it may influence the intervening events in that direction. Of course, this does not mean they will succeed. Even the most controlled industrial systems have not been able to snuff out the invisible workings of the market economy. The key point is that the firm does choose, and in so doing, consciously or unconsciously rejects other possible choices, known or unknown of its preferred state of nature. In so choosing, the firm seeks to reduce uncertainty with respect to the future. The other side of this coin shows that if uncertainty is to be diminshed, opportunity must likewise be reduced. When trying to conceptualize the output of a governmental planning process we may borrow considerably from the model of the firm. When a government plans it does so because it too has chosen a preferred state of nature and strives to achieve that state. The planning process then describes a choice of the means to realize that state of nature. It would appear however, that for better or for worse government lacks the clear objective function of the firm. There is no bottom line for government which is as unambiguous as profits are to the firm. Clearly there are some generally agreed upon goals and some acceptable and un- acceptable means. However, most of the policy process centers upon a system in which one group's goals are given precedence over another's, and the proposed means which serve one set of ends, not necessarily 16 consistent with the chosen goals, dominate the means which serve another clientele. It would seem that the output of a governmental or public sector planning process is the choice of a set of means which will be used to obtain a chosen state of nature. There are, however, some differences between the planning efforts of the private and public sectors. The private sector has the advantage of a clearer objective function, while private corporations may be complex bureaucracies the opportunity sets of those who plan and those who implement are more closely alligned. This may only be because the distinction between the bureaucrat and the decision maker is less formal in the private sector. In the public sector, then, those who draw up the agenda of means to achieve a goal (bureaucrats) are for the most part not the same group of people who decide if that agenda of means will be ratified. Thus we may divide the planning process into two parts, the choice of an agenda of means and the ratification or rejection of those means. It seems fairly obvious that if an agenda of means is not implemented then, all things being equal, the existing state of nature will not become the preferred state of nature unless it would have occurred in the absense of that planning and implementation process. Implementation then becomes a critical veto point in the resource planning process. If those who draw up plans do so in a fashion which does not take into account the political realities facing those who must implement them then the planning process has little chance of passing the paper stage. It makes sense to suggest that the more closely alligned the opportunity sets of those who choose an agenda of means and those who ratify that 17 agenda are, the higher the probability that the chosen agenda of means will be ratified. One major component of the opportunity sets of both the parties is the jurisdictional boundaries of the structure that chooses an agenda of means and the structure that implements those means. The more the two boundaries of the planning process diverge the lower the probability that a plan will be implemented. It is just this point that the balance of this thesis will be designed to demonstrate. CHAPTER III THE MODEL APPLIED TO WATER RESOURCES AND THE ROLE OF INFORMATION In the preceding chapter, a model of the political economy, which makes authoritative allocations of resources was described and the models relationship to the planning process was examined. The purpose of this chapter is to relate that model to water resources. In so doing specific reference will be made to the Federal Clean Water Act Amendments of 1972 (PL 92-500). In particular the role of information in decision making processes and planners ability to provide that in- formation as a function of jurisdictional boundaries will be discussed. Out of this discussion a natural experiment will be designed to test the author's assertions. Characteristics of Water Resources The first step in discussing the characteristics of water re- sources is a definition of water resources. For the purposes of this paper water resources will be defined as water bodies and water courses. This definition is designed to include streams, lakes, rivers, swamps, bogs, oceans, and ground water systems. It is designed to exclude water systems such as municiple drinking water systems. This definition is somewhat arbitrary since the systems in both classifications are interdependent, however, it is useful. 18 19 Most of the fbllowing discussions will center on conflicts between uses that make waters dirtier and those that require cleaner water. It may be useful to try and define what is clean and what is dirty. Aside from analytically pure water, however, which is not normally found in nature, there is no objective manner of defining clean and dirty. If one tries to define clean water as the quality of the water in its natural state, there are problems. A case can be made for the Cuyahoga River, being natural even while it was burning, since its state was a product of human evolution in the environment. There is, simply, no objective base in the past from which changes in water quality may be measured. Thus it may be more meaningful to refer to water quality in terms of the resolution of demands for cleaner and dirtier water and the means by which these demands are articulated. Water resources are high exclusion cost goods. Any one economic actor who pays for some measure of clean water cannot exclude others from the benefits of that clean water. This is true because water doesn't stay put, it goes places. If it goes downstream and upstream users have made it dirty, it affects those downstream as well. Typical of goods characterized by high exclusion costs there is very little incentive for individuals to pay for that good if they know that: (1) they could benefit if someone else purchased the good or (2) if they themselves purchased the good others could not be excluded from enjoying it. While the consumers or users of cleaner water might have some willingness to pay fOr cleaner water there is little incen- tive for them to individually go into the marketplace and exchange their resources for it. 20 At the same time, those who make uses of water resources that degrade those resources are able to impose costs on other users of water resources. They are able to achieve a personal gain from a use of water resources that reduces the quality of that water. It may be that those who have a willingness to pay for cleaner water could compensate those who might pollute for not doing so but the organiza- tional barriers that exist preclude any such bid being made. The cost structure of water resources, like other natural resources, is such that those who would reduce the quality of the water gain a unique private benefit from doing so. Meanwhile those who might demand an improved quality of the resource find that they must be willing to endure free riders when they actually demand that good in the marketplace and consequently those who degrade the water resources are over exploiting those resources because their worth is under-valued in the market. A simple example may make this point clearer. Let A hold the rights to pollute a body of water and he/she will sell that right to the highest bidder. There are two types of bidders. The Amalgamated Pollution Company is willing to bid up to $2,500 for the right to pollute. Beyond that it is cheaper for them to clean up their wastes. The second group is composed of 3,000 clean water lovers who each have a $1 willingness to pay for clean water but only 1,500 of them are willing to bid. The other 1,500 want to ride free. The water lovers total willingness to pay is $3,000. With that sum they could buy the right to pollute and never exercise it and in so doing preserve the water. But only $1,500 will be actually bid because of the free 21 riders. The Amalgamated Pollution Company will bid $1,501 and get the right to pollute, even though the individual water lovers had a willingness to pay $3,000 if the free rider problem had been solved.f Because of the free riders, the Amalgamated Pollution Company was able to make a use of a resource which was under-valued in the marketplace. The Economic and Political Costs of Improving Water QuaTity It should also be recognized that improvements in water quality are very expensive and with some means a fairly long period of time must intervene befbre they are effective. When discussing specific means of improving water quality it is first useful to divide the sources of pollutants into two categories. We may then look at the processes that reduce water degradation from point sources and from nonpoint sources. Point sources of pollution are in effect "end of the pipe" sources of pollution. If we submit the contents of these pipes to some production process in which part of what makes that material dirty is removed then we have point sources of cleaner or less polluted water. The actual technical processes for removing many (although not all) of the materials that make these waters dirty is reasonably well known and much of it has been in practice for years. The economic characteristics of the production of clean water from point sources are such that the rational provider is a natural monopoly. With respect to optimum size and pricing policy the *Even if the free rider problem were to be somehow magically solved the problem of organizing a bid among 3,000 potential benefitors would involve a significant transaction cost. 22 economics of the situation dictate that there be only one provider for any one area. The following evidence will bring us to this conclusion. First for any given clean water plant (sewage treatment plant) the mar- ginal cost is decreasing or at least constant. As such the marginal cost must be less than the average cost, and the economist criterion for efficiency of price equaling marginal cost will lead to losses by any firm producing clean water. This characteristic alone is suffi- cient to characterize the production of clean water as a "natural monopoly." As with most "natural monopolies" this production process is characterized by high capital costs relative to operating costs. These capital costs are large and represent an investment in a treat- ment plant and collection system which will have to yield services over a large number of years if its cost is to be amortized. In other words, the time horizons of such projects are quite distant. In addition to having decreasing or constant costs for another user of a given plant, as plant size increases to a certain point the cost of treatment declines. The production of clean water is charac- terized by economies of scale over a certain range of provider size. This relationship, however, is normal in that past some geographic size the costs of the production of clean water will increase. Thus an argument can be made for the existence of an "optimal" geographic or demographic size for the production of clean water. There is no particularly good reason to believe that this optimum size will coin- cide with any existing political boundary (Bish & Ostrum), with the possible exception of special sewer districts. 23 Even between clean water provider units certain economies may be had by coordination of their production of clean water. This is true because water flows, and has an assimilative capacity. Water quality decisions made upstream may have an impact on the cost of pro- viding a certain quality of water downstream. It is possible then that coordination of the production of clean water may further reduce the total cost of a given level of clean water, but this may not.be true for the individual provider. Changes in water quality are not just a function of the changes in the degree of treatment of point sources of pollution. The materials which make water "dirty" also enter into the aquatic ecosystem from nonpoint sources. A few examples will make this point obvious. Silt enters water from agricultural land, construction projects and erosion in general. Pesticides and herbicides enter from agricultural runoff, forestry runoff, and suburban sources. Nitrogen and phosphorus enter from agriculture and suburban runoff. Oil and litter enter from stormrwater runoff. This list is far from exhaustive. The important point is that they do exist and they are significant contributors of the materials that make water dirty. Thus even if all the point sources of pollution were made pristine, there would still exist some material in the water making it "dirty." While the technology of point source control is reasonably well known it is expensive. The cost of achieving the next unit of clean water, by point source production of clean water, is greater than achieving the last unit of clean water. There are increasing costs associated with the production of point sources of clean water. 24 From an economists standpoint this factor begs the question of whether the cost of the next unit of water quality attained by point produc- tion is greater than the costs of a unit of water quality attained by the control of nonpoint sources. Several other factors need to be examined before it is possible to speculate on this question. First the technology of nonpoint control is not well developed. Much of what we do know is information concern- ing the control of soil loss. Most of this information and related techniques were derived not fbr controlling water pollution but for preventing the erosion of land. There should be a logical link between these activities but exact or even approximate knowledge of that linkage does not appear to exist. While it may be possible to estimate the cost effectiveness of some nonpoint clean water production processes those estimates may be biased and their variances large. If this is the case then it is probably very difficult to make meaningful comparisons between the costs of producing clean water by point production vs nonpoint production. However, there are some general things which may be said about nonpoint pollution. It is almost a fundamental property of nonpoint sources of pollution that they are generated by the way in which people use land. It would seem to follow that if any change is to be made in the level of nonpoint pollution there will have to be a change in the way people use land. As was discussed in the preceding chapter, those who have a demand for cleaner water have other avenues of redress than the market- place. They may choose to use their political resources by demanding that the government reallocate the rights to certain uses of water 25 resources. These demands have been expressed at all levels of govern- ment but the level at which they have met with the most success has been the Federal government. These demands were of sufficient magni- tude to persuade Congress to pass the Federal Water Pollution Control Act Amendments of 1972 (PL 92-500) which establish the goal of zero dis- charges into the nation's waterways by 1985. They also suggest an interim goal of fishable and swimmable water where possible by 1983. While the author believes these goals have more value as rhetoric than policy, clearly the intent of the act is to improve the quality of the nation's waters on a scale that can only be called ambitious. The act is probably the most comprehensive water pollution control act yet passed. It is comprehensive because it makes specific references to both point source and nonpoint source pollution activities. The amount of political benefits garnered at the Federal level ought to be a function of the performance of anti-pollution standards. More realistically, rhetorical perfbrmances may yield the same utility as substantive ones. Thus, we may begin to think about the means chosen to achieve or approximate these goals as the resolution of the political benefit-cost matrix. Means of Reducing,Nonpoint Pollution Governments face several options when they wish to change human behavior. They will choose to attempt and rearrange individuals' opportunity sets in a fashion which will result in a desired performance. The means by which a government might do so include taxation, subsidy, regulation and purchase. Let us examine the possibilities of each of 26 these alternatives in light of the existing political and market economies. The purchase option is obviously, from a dollar standpoint, very expensive. Even if government were to purchase just a few of the proverbial "bundle of rights" associated with land in the hopes of reducing nonpoint polluting activities these costs would still be astronomical. If the purchase of development rights plan of Suffolk County, New York is any sort of guideline we cannot expect any govern- ment to be able to afford this alternative. Lesher and Eiler point out that Suffolk County was prepared to spend $21 million to obtain the development rights to 3883A. of land. Even if we assume that this high price is a function of high land values in this case (which it is, land values in this case were about $7500/A.) and that the purchase of "pollution rights" would cost less, the amount of land that would have to be involved would make the dollar costs tremendous (Lesher and Eiler, 1977). While political and dollar costs are not the same thing, clearly they are related. If it is decided to buy those "pollution rights" the billions of dollars used to purchase them must come from somewhere. Ultimately they must come from taxpayers and the political costs of increasing taxes enough to cover these costs would be large. Taxation is a second alternative. The basic notion in designing a tax to control some activity is to tax that activity in such a way so as to reduce the difference between social costs and private costs. In the case of nonpoint pollution, to be effective, a tax program would have to levy tolls against a vast number of activities. Even if it were possible to effectively delineate and tax these activities the political costs would be high. These taxes would have to be levied 27 against huge numbers of people and specific groups of people with political clout. These might include farmers, foresters, builders, agricultural chemical producers, and others. Thus even if a taxation policy were to yield revenues to a government there would be large political costs to overcome. A subsidy is essentially a bribe for carrying out some activity. In another sense it is the opposite of a tax. In this case subsidies are offered to the potential producers of nonpoint clean water to hope- fully reduce the divergence between the social and private costs of preventing water degradation from nonpoint sources. Because they are the opposite of taxes very specific groups may support them (i.e., those which will receive them) and, depending on the source of that subsidy, there may be little opposition to such a program, particularly if the source is the state or federal government because any subsidy makes very little difference in an individual's tax bill. The experiences of the Soil Conservation Service (SCS) and the Agricultural Stabilization and Conservation Service (ASCS) may be the most instruc- tive in this case. The SCS provides farmers with subsidies of tech- nical information (e.g., farm plans) and the ASCS provides cost saving subsidies for soil improvement practices. Farmers (clientele) provide both the $05 and the ASCS with the political support necessary to assure their survival. The recipient farmers have an interest in maximizing their subsidies and the SOS and the ASCS have an interest in giving out as much subsidy as possible (Bartlett, R., 1973, pp. 70-75). Individual tax payers have no incentive to oppose this relationship because the transaction costs of opposition are much higher than any 28 individual benefit they might achieve. This is a natural outgrowth of the fact that such lobbying is a high exclusion cost good. Such subsidy programs have been suggested as a means of producing nonpoint sources of clean water. Indeed the SCS is often proposed as one of the land agencies in this respect. Subsidies do have the problem that they may be expensive, as does government purchase, and some political costs may flow from these high dollar costs. Finally there exists the option of regulation. Regulation in effect means that certain activities are declared illegal or at the minimum requires some form of governmental consent prior to undertaking that activity. The dollar costs to government of this option are relatively small. For the most part they involve policing costs. The dollar costs to society may be much greater depending on the nature of the activity which is regulated. On the other hand the political costs of such actions may be quite high. As pointed out, any signifi- cant change in the production of clean water from nonpoint sources will probably have to come from a change in the manner in which people use land. This in effect means some form of land use controls. When discussing land use controls, at least in the context that they are thought of in the United States, there are two distinct characteristics of the existing political economy which must be considered. First, for the most part the only unit of government that is considered to have a legitimate right to adopt land use controls is a local (county, township, city, or village) government.* For *Several states have also adopted land use controls to control development near specific resources such as shorelines and parks. 29 whatever reason, these are the only units of government which have successfully adopted comprehensive land use controls. (The only exception that the author is aware of is the state of Hawaii which has adopted a statewide system of land use controls). The second fact to be considered is that land use controls are less popular in rural areas than in urban areas. If we are to consider land use regula- tion as a viable alternative to achieve the goal of reduction of non- point sources of pollution, we must recognize that those decisions will probably have to be made at the local level. In addition we must recognize that the political costs involved will probably have to be borne by local politicians, and that those costs will be differentially assessed between urban and rural decision makers. It is not realistic to believe that any one of these techniques will be "the" technique that will be the most effective in changing the amount of nonpoint pollution that occurs. If one buys Lindbloom's argument that most policies never change more than incrementally (Lindbloom, 1959) then it follows that incremental changes in all of these policy options will be more successful in producing clean water than an incremental change in only one. We can now address the matrix of political and dollar costs of various measures which might be initiated at the Federal level to improve water quality. Figure 3 is a diagramatic representation of such a matrix. Implicit in this matrix is the idea that the structure of the demand for clean water is fixed and must be met and that it cannot be met simply by controling point sources. The matrix suggests that there will be increasing political costs as any one 3O Dollar Costs Political Costs Point Sources: Treatment Plants High and increasing Low, increasing Nonpoint Sources: 1. Purchase Very high High and increasing 2. Subsidy High Low and increasing 3. Tax Low or Negative High and increasing 4. Regulate Low Very high if done directly by Federal Govenment Figure 3. Matrix of Political and Dollar Costs at the Federal Level method of control is used to a greater and greater extent. The lone exception to this is the regulation option. In this case, the political costs of changing the authority to regulate land use are so great that the direct federal regulation of land use will not enter the picture. The rest of the matrix suggests that to minimize political cost, decision makers will probably choose a mix of means to achieve a water quality goal. Indeed one might be so bold as to suggest that water quality will be improved in such a way that the Marginal Political Cost of Means A equals the Marginal Political Cost of Means B and so on. The calculus is not very important because these functions cannot be specified in any exact manner. But what about regulation? This is where the concept of shedding becomes important. If the federal government can get local levels of government to regulate land use then the political costs of regulation 31 at the federal level are much lower and regulation enters the picture as a viable alternative. The only question then that remains is how to convince local governments that they should control land use for the purpose of reducing nonpoint pollution. One way is to tie the flow of sewage treatment construction dollars, which have many positive side effects in a given community in terms of jobs, to the adoption of land use controls. In rural areas, where resistance to land use controls is higher and the impact of construction grants smaller it may also be necessary to offer subsidies-bribes, particularly to agricultural and forestry interests which have traditionally opposed land use controls, and may be well organized. In the provision of point source and nonpoint source control of pollution, the planning process plays an important role in deter- mining the least political cost combination of programs. It has already been suggested that, in the area of point source control, the planning process helps provide for control of capital expenditure and coordina- tion of control activities in a watershed. As is the case with point sources of clean water planning plays a crucial role in the cost effectiveness (both in dollar and political terms) of nonpoint programs. Similar to point sources, planning for nonpoint provision of clean water must try to integrate watersheds and jurisdictional boundaries, identify problems and propose alternatives, and suggest investment strategies. However, these functions turn out very differently in nonpoint control. For example where point source planning can identify a problem in terms of a pipe, nonpoint 32 planning must deal with land parcel by parcel. Where point planning deals with a reasonably well developed highly capital intensive technology, nonpoint planning deals with a tremendous diversity of techniques which are less capital intensive and represent a somewhat lower level of technological development. In both point and nonpoint planning the planning process is also a forum in which political compromises are made. For point sources the planning process to some extent formalizes the process by which politi- cians "bring home the bacon." With nonpoint sources the salient trade- offs may involve choices of techniques of nonpoint control and managing urban-rural conflicts. Finally it should be remembered that the planning process is not necessarily the reflector of the values that society holds. It may in fact be intimately involved in the process by which values and political perceptions are changed (Lientoff, 1974). PL 92-500 recognized, as other clean water legislation has, the importance of the planning process in the provision of clean water. The act and subsequent court rulings have mandated that clean water planning process (Section 208 Planning) encompass the entire United States and its territories. To accomplish this objective, approximately $225 million have been allocated to various agencies around the country. As with the provision of actual control of pollution, the jurisdictional boundary of the planning process is an important con- sideration. There should be some coordinational economies to be gained as the jurisdictional bounds of the planning process increase, parti- cularly in the management of point sources. On the other hand, many 33 of the implementation powers have been delegated to local units of government (county, city, town and village) and many of the decisions to be made with respect to nonpoint sources must be done on a parcel by parcel basis. It has already been suggested that implementation is the crucial step in the planning process and that the probability of implementation increases as the opportunity sets of those who draw up agendas of means and those who ratify those agendas converge. Again the jurisdictional boundaries are key elements in the two respective opportunity sets. Section 208 planning is principally done by two types of organizations.* The first and most frequent is the multi-county region. These regions typically are associations of l to 10 counties. The second major type of organization that does section 208 planning are state level departments of Conservation, Environment, Public Health, or Natural Resources. In general, though not always, they conduct section 208 planning exercises in the areas not covered by multi- county regions. These areas generally include more counties than a multi-county region. As a general rule then the physical jurisdictional boundary of the state level section 208 planning agency is larger than the multi-county agencies jurisdictional boundaries. If this is the case, we may perform a natural experiment to test whether the size of the planning boundary has an effect on the probability of implementa- tion. The question then arises, how can we measure these probabilities. *Section 208 plans are also, to a lesser extent carried out by city and county governments, and multi-state planning organizations. Each of these organizational types are so few in number as to not lend themselves to statistical comparison. Appendix C lists the various organizational forms and some selected characteristics. 34 The Role of Information What we are trying to measure is the relative ability of different institutional structures to influence the decisions of a set of governments. This situation is highly analogous to one found in consumer theory. Consumers rely on infbrmation in making their choices in the marketplace. Information about a set of goals allows the con- sumer to assess the utility he or she would expect to receive if that goal were purchased. In the pure theory perfect information is assumed, thus there is no uncertainty with respect to the utility of a given purchase. In reality this assumption holds about as much water as a leaky sieve. Not only is information not perfect it is expensive to acquire. Any decision which is made by the consumer will be made on the basis of partial information. Producers may lower the relative cost of acquiring information to the consumer by providing that information to the consumer. This is in essense an information subsidy. This is, of course, not an altruistic act on the part of the producers. The information they provide will try to portray their product in a favorable light so as to induce the consumer to purchase that product. This process is usually called advertising. Because the consumer is uncertain, this information sub- sidy process does more than just infbrm the consumer. It is also intimately involved in the formation of tastes and preferences. The structure of this information subsidy process effects the amount and type of the subsidized information provided. Two factors determine this. First, the information provided is a high exclusion cost good. When a firm advertises,that advertising may have positive 35 effects on the consumption of close substitutes and complements, may also increase. The producers of those close substitutes and complements may benefit; in essense they are riding free at the expense of the firm which is advertising. An example may make this clearer. When Wisconsin cheesemakers promote Wisconsin cheese, New York cheesemakers (a close substitute) benefit as well from a general increase in cheese consumption. Not only do New York cheesemakers benefit but cracker makers benefit as well since cheese and crackers are complements. Firms will be willing to endure these free riders as long as benefits they receive from advertising outweigh the costs. As the firm size expands they will be able to internalize more of these benefits and as such provide more infbrmation. This factor explains why individual dairy farmers don't advertise but the Wisconsin cheese industry does. No individual farmer has a significant share of the market but the Wisconsin cheese industry does. The second factor has to do with the type of the information subsidy activity. If it is done through a mass market then the informa- tion subsidy process is a joint impact good (Schmid, 1977). Joint impact goods are ones which have a marginal cost of zero over a certain range. When an advertiser pays fer an ad on television it costs the same whether one person or 1,000 people see it. The marginal cost of subsidizing information to the next viewer is zero. Infbrmation subsidy activities can also be done on a "normal" cost function basis. The use of salesmen may be a good example. The marginal cost of information subsidy activities in this case decreases over some range and then increases and as such is "normal." The information needs and wants, 36 then, of the consumer determine the degree of jointness that the information subsidy activity displays. When the consumer's information needs and wants are relatively simple (e.g., breakfast cereals and automobiles), infbrmation subsidy activities may exhibit a high degree of jointness. Where the information needs and wants of the consumer are more demanding (e.g., Avon cosmetics and industrial machinery), informa- tion systems which cater to the unique demands of the consumer must be designed.* Thus the structure of the information subsidy system is determined by the consumers infbrmation wants and needs and the producers market share. Randall Bartlett applies a very similar set of arguments in discussing the behavior of political systems. He suggests that politi- cians seek to maximize votes and operate under uncertainty with respect to the impact on their constituencies of the decisions they make. Consequently various groups may subsidize information to the decision maker in hopes of influencing his/her decision because he/she operates under uncertainty. Bartlett also describes bureaucrats as security maximizers. One of the means they may use to increase security is subsidization of information to decision makers (Bartlett, 1973). Politicians or decision makers are analogous to consumers. They are uncertain about how to go about maximizing the votes or utility they want to maximize. Planners are similar to producers in that it is in their interest to subsidize information to decision makers. *This in no way implies that the underlying product must change just the way information is provided about that product. Any good sales- man knows that different people need different pitches. 37 Section 208 planners are essentially monopoly water quality plan producers within a given jurisdictional boundary. If they engage in some information subsidy process that convinces people that they want more water quality planning the agency that did the convincing is probably the one that will do the planning and hence directly benefit from that infbrmation subsidy activity. Their objective, then, is to convince local decision makers that the plan they have drawn up is good for them and ought to be implemented. To accomplish this goal they will set out to subsidize information about their planning process. Politicans or decision makers represent a group with a fairly high level of information needs. This is particularly true when non- marginal policy changes are involved. In addition each decision maker or body of decision makers have, to a greater or lesser extent, differ- ing information needs. This suggests that the information subsidy system needs a "salesman" approach instead of a "mass audience" approach. As such the cost functions of such a program are Ynormal." The principle reason we may expect this cost function to increase beyond some point has to do with the transaction costs of an information sub- sidy process. As the jurisdictional boundary of the planning process increases the physical distances between the planning agency and decision makers increase. Since the information subsidy system pro- bably calls for a great deal of personal contact between planners and decision makers, as the jurisdictional boundary increases planners will have to spend more time traveling between the agency office and decision makers' offices, telephone bills will be higher, and coordination will 38 probably be more difficult. All of these increasing transaction costs will tend to reduce the agencies ability to subsidize information to decision makers. Thus, it is expected that as the jurisdictional boundary of the planning process increases the ability of planners to subsidize information to decision makers will decrease. If the probability of implementation is a positive function of the amount of information subsidy activities undertaken then, all other things being equal, the probability of a plan being implemented declines as the jurisdictional boundary increases. The question then becomes how can this information subsidy pro- cess be measured? There are several ways this may be accomplished. Section 208 planning has as part of its process public participation requirements. If formal public participation meetings are viewed as iterative, then they may be viewed as an information subsidy activity. That is, a public hearing allows the planner to extoll the virtues of a given set or sets of agendas of means, and in so doing sells the public and decision makers, as well as gathers community inputs into the planning process. Budgets for public participation should also include special presentations to decision makers, whether the agency writes press releases and if so how many, and size of public participa- tion staff. All of these variables appear to gauge some aspect of the information subsidy process and are quantifiable. If this is the case then these variables may be used to measure the differences in the performance of alternative 208 water quality planning organizations, and the effects of jurisdictional boundaries within each organizational type. CHAPTER IV METHODOLOGY In the previous chapter, it was suggested that the section 208 planning agencies whose boundaries were closer to local govern- ments would be better adapted to carry out information subsidy activi- ties to that level of government. This chapter will discuss the impor- tant factors in the design of an experiment to validate or invalidate this contention. In so doing a brief discussion of the nature of such an experiment will first be undertaken. Then the design of the experiment, nature of the data to be collected and hypothesis to be tested will be described. Natural Experiments It is the objective of this experiment to determine the effect that the jurisdictional boundaries of the organizations in question have on the performance of those organizations. Because in the course of testing the proposed hypotheses many intervening conditions will not be strictly or even remotely controllable this experiment is best classified as a natural experiment. A natural experiment is one in which the experimentor exhibits little or no control over the events undergoing observation. That is, the experimentor may only validate or invalidate the hypothesis on the basis of observations of some externally occurring event or events. This is contrasted by a normal 39 4O experiment where the experimentor has control over the occurring events. Though the distinctions between the two experimental types are useful there is probably no objective way to distinguish between the two. The functional distinction lies in whether the experimentor can control the factors which he/she perceives as relevant to the research he/she is conducting. This should not be misconstrued to suggest that strong statements cannot be derived from a natural ex- periment. At the same time the reader should be aware of their limita- tions. Sample Frame If data on the infbrmation subsidy process is to be gathered it is first necessary to define the relevant population from which it is to be gathered. For the purposes of this exercise the relevant popula- tion is all the agencies with 208 planning responsibilities in the 50 United States. Specifically excluded from this population are Washington, D.C., Guam, and Puerto Rico. In addition the 208 agency in the Little Rock, Arkansas area which covers two counties was excluded because of an arrangement by which their funding was provided by the Army Corps of Engineers. This left a total population of 217. The actual sample frame was the Environmental Protection Agencies Water Quality Management Directory, which listed all the addresses of the 208 agencies, and the names of the directors of each program. Because the size of the population was relatively small a census of the agencies will be attempted. 11 p1. nu 65 CO 1131 at‘ Phj CET of the the the 41 Survey Instrument A mail questionnaire* was chosen as a survey instrument for two reasons. The first reason was economy. Mail surveying is generally recognized as the most cost effective means of gathering data if reasonable rates of response are expected. The second reason why a mail questionnaire was used concerned the nature of the information that was to be gathered. Some of the questions asked would probably require the respondent to look up information or consult another party. This limitation effectively precludes any sampling technique, such as phone interviews, which require relatively quick responses. The questionnaire was composed of an agency identification number and three types of questions. Two consistency questions were asked. These were questions to which the author already knew the correct answer and were used to judge the knowledge of the respondent. The second type of questions were the control questions which attempted to measure some aspect of the agency's administrative or physical jurisdictional boundaries. The final set of questions con- cerned the agencies information subsidy activities. The survey instrument was mailed to the agencies on the first of February, 1978 and a follow-up letter was sent to nonrespondents on the twenty-fourth of February 1978. No attempt was made to pretest the questionnaire. Variables and Hypotheses Two types of variables were constructed from the data gathered by the questionnaire. The first type were classification variables. These *See Appendix B. DO: 695 eXC Sta Val 1:10: ther and 42 came from the control questions. They were designed to classify agencies and describe certain demographic and administrative charac- teristics of those agencies. The second types were the experimental variables which were divided into two types. The standardized informa- tion subsidy variables measure the information subsidy activities with respect to some reference area described by the classification variables. The dichotomous information subsidy variables gauged whether a particular information subsidy trait was or was not present in a given organization. On the basis of the discussion in the preceding chapters, it has been suggested that an agency's information subsidy activities to local decision makers will decrease as the jurisdictional boundaries of the planning process increase. It has been assumed that these information subsidy programs are positively correlated with the probability of a plan being implemented. In the between organization analysis the hypothesis which will be tested is the one-tailed proposition that regional agencies, because they have smaller average jurisdictional boundaries, will have larger information subsidy variables than state agencies which have larger jurisdictional boundaries. The only exception to this will be when the information subsidy activities of state level agencies are aimed at state level decision makers. It is hypothesized that in this case the state level agencies will have higher values than the regions because of the coincidence of their jurisdic- tional boundaries. In the within group analysis it is expected that there will be a negative relationship between the planning area size and the information subsidy variables. 43 Case Studies Because a great number of factors, which are not quantifiable affect the section 208 planning process, nonstatistical methods may also be useful in examining the planning activities. Two brief case studies of section 208 agencies will be presented which will address some of these less quantifiable aspects. These case studies represent a judgment sample and as such, because they are nonrandom, statistical comparison of the agencies is not valid. They are, however, useful in that they allow a descriptive presentation of factors which might otherwise go unnoticed. of an. St. 1.111 EX If wh be bl CHAPTER V DATA ANALYSIS AND RESULTS In this chapter, survey results will be analyzed. The effects of the response rate and response bias will be discussed. Finally an analysis of the effects of jurisdictional boundaries within and between state and regional agencies will be conducted, and the limitations of the analysis discussed. Response Rates and Nonresponse Bias In any experiment where the response of the subject of the experiment is optional nonresponse bias is an important consideration. If the sample is divided into two strata, those who respond and those who do not, to the extent that the potential responses of the non- respondents differ from the responses of the respondent class bias may be introduced into the analysis. That is, the values garnered from the respondents will differ from the true population mean. Formal bias may be described as: U'l-U where: U1 = the estimate of the population parameter from the respondents U = the population parameter. 44 2C 91' CO Cd DO. 351‘ 45 This in turn is equal to: Ul'U=U1'(R1Ul+R2U2) U1 - U = R2 (U1 - U2) where: R1 = response rate R2 = nonresponse rate = l - R1 U2 = the estimate of the parameter from the nonrespondents (Moser and Kalton, 1972, p. 167). Thus the lower the nonresponse rate, all things being equal, the lower the potential from nonresponse bias. A total of 217 surveys were sent out which represented a census of the relevant population. Of these 179 were returned which yields a total response rate of 82.5 percent. After examination a total of 163 were found usable which yields a usable response rate of 75.1 percent. Surveys were rejected for one of three reasons. First they may have failed the correspondence test. Two questions were asked which the author already knew the answers to. The first question concerned the size of the agencies section 208 budget. The second question asked the respondent to give the date the agency was chosen as a section 208 planning agency. If the answers given on the survey differed significantly the survey failed the correspondence test and was rejected. The second reason for rejection came from obvious errors in the questionnaire. If the value reported in the survey was wildly out of line with what was expected, where possible these values were cross referenced. As an example, one survey said the agency was preparing a plan for 157 counties. The author did ne di ha CO 130; Val val 5111 of - SUcl 46 not believe any state had that many counties but found he was wrong. The final criterion fbr exclusion was the arrival of the survey too late to be used in the statistical procedures. The effective nonresponse rate was 24.9 percent. Even if the values of the nonrespondents differed from the respondents by a factor I of 50 percent this would bias the results by only 12.5 percent. The author believes that this nonresponse rate is low and does not signifi- cantly bias the results. Inevitably, however, this is a decision the reader must make for his/herself. Analysis of Differences Between Organizations Clearly, if the two types of organizations (state and regional) are to be used as independent variables in this section the first criterion to establish is that there is a difference in these two agencies. It has been hypothesized that a difference in jurisdictional bounds will lead to a difference in information subsidy efforts. It is necessary, therefore, to establish that there is a difference in juris- dictional bounds. Jurisdictional boundaries have been described as having both area and administrative components. Data on both these components was collected from the survey. Table 1* shows the results of the tests of jurisdictional boundaries conducted with respect to area criteria. The dependent variables are state and regional organizations. The independent variables are number of counties in the planning area, number of square miles in the planning area and population in the planning area. *Where applicable these tables provide information on degrees of freedom, T-values and chi square values for those interested in such figures. O. 1 1 ‘1”1. ,.F I“~.-n‘.wm an mappgoamgco: mo mmsmumn ugmzczoz ummmwa m. mcamww mwgha ooo.o~F.P mem.mpm., mm meowmam mmm. omp mm. mwm.omo ¢m_._oo.m mm mwpmpm 8.8 9:553 mg» :_ cowpmpzaoa FFN.m opm.m mm m=02omm coo. FF, _N.m _wm.e mm~.ss mN emmpaum 3.8 @5533 93 E mmpz= mgmzcm mo Langsz mm.o Fa.¢ mm mcopmmm coo. mmp mo.~_ mm.m mq.mo mm «poem amen mcwccopa an» cw mmwuczou mo cmaszz mungwm .2 .2... . am“? e... z mmwucmm< acmccmpa mom cowuumm Pacowmmm new macaw mo mmwgmucaom chomuumnmwgzn cmwzumm mmucmcmmmpa .F mpnmh 48 The tests performed are one tailed tests because as has already been suggested it is expected that the state planning units will be larger. The results from Table 1 show that the state units plan for more counties and a larger planning area. It is somewhat less clear that the populations in the states' planning areas are greater than the regions' planning area. This may be because the state units tend to plan in the so called “nondesignated” areas which tend to be more rural and hence have lower population densities. In the areas of administrative functions, eight questions were asked about different responsibilities the agencies had. These varia- bles included responsibility for nonsection 208 water quality planning, land use planning, enforcement of land use regulations, enforcement of water quality standards, areawide A95 review, statewide A95 review, and enforcement of environmental laws. Table 2 lays out the results of this analysis. In all but two cases the results were significant to the ten thousandth's decimal. Thus on the basis of the statistical analysis there is strong evidence for there being a difference in the administrative functions of the two groups. Indeed this really represents a case of the empirical validation of the obvious. Even a casual observer of the roles of state and regional agencies could have predicted that their roles would be different. The conclusions that can be made are that there is a difference in the physical and jurisdictional boundaries of the two organizations. In trying to measure information subsidy the variables may be broken into two general classes, those that aim at widespread audiences Table 2 ————-J Respons amrd 51 Re lespons use p11 5 R Respon land u S l Respor water l RESpoi OfNPl 1 RESpor A95r S R. Respon1 Hide Ag St Rei ReSDOns; 49 Table 2. Differences in Administrative Functions 2 Signifi- Yes No N X * cance Responsible for non-208 water quality planning States 27 5 33 9.68 .0019 Regions 43 42 85 Responsible for land use planning States 8 23 31 56.05 .0000 Regions 83 5 88 Responsible for enforcing land use regulations States 7 25 32 2.66 .1026 Regions 7 75 82 Responsible for enforcing water quality standards States 28 4 32 78.30 .0000 Regions 2 76 71 Responsible for issuance of NPDES permits States 18 14 32 35.31 .0000 Regions 4 77 81 Responsible for area wide A 95 review States 8 22 30 34.02 .0000 Regions 75 13 88 Responsible for states wide A95 review States 23 8 31 29.63 .0000 Regions 13 63 76 Responsible for enforcing environmental laws States 25 6 31 63.07 .0000 Regions 4 77 81 *Chi square values were corrected with the Yeats continuity corre- lation formula. 110 1511' ”Ed 519 a cl 208 50 and those targeted at specific audiences. Those variables which address general audiences include number of hearings and meetings, public participation budgets, and news releases. The variables measuring information subsidy to specific audiences attempt to guage the information subsidy process to decision makers at various levels of government. This process is measured by trying to look at the number of special presentations to decision makers. In each of these tests a one tailed hypothesis is specified because it is expected that each measure of information subsidy will be weighted in favor of the regions. A word of caution is called for at this point. Many of these tables give mean summary statistics with respect to a certain activity carried out by a certain type of agency. These mean statistics should ngt_be considered standards by which any individual agency could measure the effectiveness of its public participation programs. While aggregate they are useful in comparing classes of agencies, they are not meaningful standards for individual agencies. Tables 3 and 4 depict the general information subsidy variables. All of the parametric tests suggest that the regional agencies responding were more active in the information subsidy process, as measured by the included variables. The variables included were all significant at the .05 level or better. One nonparametric test was done in this section. This test was a chi squared procedure in Table 4 to test whether there was a differ- ence in the number of agencies that prepared news releases on section 208 planning at the state and regional level. The analysis feund that Pul ti< con Pub tio 1,0 Pub? 1101 COUr Pub] tion 1,0-(M 51 Table 3. Public Participation Hearings, Public Participation Budgets and Public Participation Staff Per County and Per 1,000 Population Standard t- D F Significance N Mean Error value (one tailed) Public participa- tion hearings per county States 28 .67 .711 -2.05 111 .021 Regions f 85 15.46 38.022 Public participa- tion hearings per 1000 population States 28 .0251 .005 -l.66 109 .050 Regions 83 .2050 .063 Public participa- tion budget per county States 32 $12,900 9,311 -1.73 112 .048 Regions 83 $33,828 6,645 Public participa- tion budgets per 1,000 population States 32 $ 95.00 17.67 -3.36 109 .001 Regions 79 $215.89 21.69 Public participa- tion staff per county States 31 .1836 .114 -2.47 112 .008 Regions 79 .7196 .125 Public participa- tion staff per 1,000 population States 31 .0018 .000 -2.28 109 .012 Regions 80 .0065 .001 mm Am.av a Am.mmv em meowmma _mm~. KN.F mm Am.~.v a Am.amv mm maaapm mcwc:m_a mom cowpumm co mummmpmc mzm: mcwgmamca mmwocmm< 52 augmo.e.=m.m .mx 2 Aev oz hev mm> ¢©.N m¢.NF NR m=o_maa moo. co. ma.~- ma._ oo.~ om mmpapm sucaou 2mg mummmpmc mzm: eo gmnszz Aum_.ap acov . . - mucmuwmwcmwm m o mapm> u coccm ucmucmpm cam: z mommmpom mzmz mcwgmamca mmwocmm< mo mommucmocma new mcmae:z use apcsou cum mummmem mzmz .e «paw» 53 87.5 percent of the state organizations and 95.5 percent of the regional organizations prepared their own news releases. This differ- ence is significant to the .2581 level. It would be difficult to suggest on the basis of this test that there was a difference. How- ever, if one examines the number of news releases prepared per county (Table 4), it is easy to see that on the average the regional agencies prepare far more news releases per county (12.48) than the state agencies (2.00). This is a good example where the relative strength of a parametric test is useful in obtaining a more detailed analysis of the data. Tables 5 and 6 list the results of the analysis done on the variables which attempt to depict the information subsidy process to specific groups of decision makers. Those decision makers are the governor or the governor's staff, state legislators, county legisla- tors and city and township legislators. Again one tailed tests are used. It is hypothesized that the regions will do a better job sub- sidizing infbrmation to county, city and township legislators than the state agencies will. On the other hand, the jurisdictional boundaries of concern to the governors and state legislators are pro- bably closer to the jurisdictional boundaries of the state agencies. Thus it is expected that state agencies will do a better job of sub- sidizing information to these groups. The results in Table 5 show that these hypotheses are upheld. The state agencies made presentations to the governor or the governor's staff and state legislators (82.6 percent to 54.4 percent and 76.2 to 33.9 percent respectively) more frequently than regional agencies. 54 .auwchpcou com cowuumccou m.amm> mza saw: uwumznvm mp ovumwpmpm umcmzcm Page mm AN.“ v o Am.~mv as m=o_umm mm_o. mm.m Fm Ao.m~v o Aa._~v mp mmpapm mcogmpmwmmp avgmczop new xuwu ow mums mcopumucmmmcq meumam om Ae.m_v PF Am.omv mo meowmmm mmmo. ma.~ NN Am.2mv N Am.mov mp mauapm mgopmpmwmmF zuczou op mums m:o_popcmmwca meomam mm Ap.mov mm Aa.mmv om meowmmm omoo. am.m 2N Aw.mmv m AN.mNV o_ masapm mcomemwmmp mumpm op mums meowpmpcmmmcq pmwuwnm we Ao.mev _m A¢.emv em mcopmam memo. sm.e mm Ae.epv a Am.~mv m_ museum ewmpm m.coccm>om so coccm>om op mums mcowpmpcmmmcg Pneumam mo:mu.e.=m_m aux z ARV oz flay ma» mgomemmeJ awgmczo» ucm save ucm mcomemmeA xacaou .mcopopmwamg mpmpm .mwmum m.soccm>ow co coccm>oo an» op mcowuopcmmmca Pneumgm mcwxmz mwvocwm< .m mpnmh 55 opo.p ooN.o. mm. weapoom mpo. mo m~.N1 Nme. mom. Np mmpopm spooou coo mgopopmpmmp opcmczop coo zp_u op mcopp -opcmmmco Fopomom mo cmoaoz mom. mpo.m mm mcopmmm NNo. mo oo.m- Foo. «mm. mp mopopm apnooo coo mcopopmwmop zpcoou op mcopp uopcmmoco Fopomom mo consoz m_.N pm.m mo mcopmmm New. mm mm. o~.m mo.m Np mopopm mcopopmpmop opopm op mcoppopcmmoco _o_uoom mo gonzoz ope. Km._ mm mcopmmm omp. om op.F opp. nm.~ op mmpopm mwopm m.coccm>om Lo coccw>om mop op mcopp lopcomoco popuoom po cmosoz Romppop acov 1 coggm mucoupppcmpm no mopo> p ocoocmpm com: 2 mcopopmpmmp opcmczop oco xppo oco .mcopopmpmop xpcoou .mcopopmpmmp opopm .mpopm m.cocco>ow co Loccm>ou asp op moo: mcoppopcmmmco Popomom mo consoz .o epoch 56 At the same time the regional agencies made presentations to county and city and township legislators with a greater frequency than state agencies (86.2 percent to 68.2 percent and 92.8 percent to 71.4 percent respectively). All of these results are significant at the .10 level or better. Table 6 describes the tests of the number of special presenta- tions made to each specific group. In general the statistics behave as it was predicted they would although the tests involving county, city, and township legislators are much more significant than those involving the governor, the governor's staff and state legislators. Unfortunately, the author believes that in these tests nonresponse bias has biased the results upwards. This set of questions was the least answered on the entire questionnaire. 0f the state agencies, 36 to 42 percent filled in the item in question. The regional agencies responded at a slightly higher rate, 53 percent to 59 percent but again nonresponse bias is probably present. The author suspects that the values excluded were probably zero more often than not although he has no way to prove this contention. If this assertion is correct, then there is reason to suspect that the estimates of the population means given in Table 6 are biased upwards. This does not mean that no information can be gleaned from this table. While little faith may be attached to the absolute differences in the estimates of the population parameters, other information is available. If it is assumed that the direction and magnitude of the nonresponse bias in each case is about the same, then the means in Table 6 may still tell us something about the relative differences in the information subsidy 57 practices of each group. This case proves to be just the opposite of the first set of variables. Here the nonparametric tests have probably yielded more information than the parametric ones because the questions involved were simpler and consequently were answered with a higher frequency. On the basis of the evidence presented it is possible to draw four conclusions about the differences between the information sub- sidy processes of the two organizations. First, the regional agencies appear to provide more information to general audiences. The analysis also suggests that the regional agencies also subsidize infbrmation on a larger scale to county, city and township legislators. Since these levels of government are generally the ones that are perceived as the legitimate controllers of land use, the plans to control nonpoint pollution drawn up by the regions should have a higher probability of implementation than those which are drawn up by state level agencies. Finally the state agencies appear to do a better job of subsidizing information to state level decision makers. Within Groups Analysis In the between group analysis, a difference in the two groups was demonstrated. However, it was not possible to attribute this difference to physical or administrative boundaries of the organiza- tion. It is possible to control for administrative function. This may be done by examining the relationship between physical planning boundaries and the information subsidy variables within each group. By holding the group type constant, the administrative boundaries are to some extent controlled. 58 The test statistic used here is the coefficient of correla- tion (R). It measures the degree to which some linear relationship exists between two variables, with a value of zero for a relationship and plus or minus one for a perfect positive or negative linear relationship. It is hypothesized that as the jurisdictional bounda- ries of the planning areas increase, the values for the information subsidy variables will decrease. Thus it is expected that a negative relationship will exist between the two variables. In all of these cases the independent variable is the number of counties in the plan- ning area. The dependent variables are all in the class of the general information subsidy variables because of the nonresponse problems in the specific group information subsidy variables pre- viously discussed. In each case, the correlation between the variables has been measured using implicit linear and exponential functions. A plot of the values in each case suggested that an exponential functional form might explain more of the variability in the values than a linear functional form. Analysis of State Agencies In this procedure the correlation between the information subsidy variables and the number of counties in the state level agencies was calculated. Table 7 describes this correlation with an implicit linear relationship. In each case the sign of the coeffi- cient of correlation is negative and four out of seven are signifi- cantly different from zero at the .05 level or better. 59 Table 7. Correlation Between the Number of Counties in the Planning Area and Information Subsidy Variables for State Agencies with an Implicit Linear Relationship Dependent Variables R Significance of R Public participation hearings per county -.457 .0072 Public participation hearings per unit of population -.432 .0173 Public participation budgets per county -.201 .1346 Public participation budgets per unit of population -.337 .0296 Public participation staff per county -.219 .1187 Public participation staff per unit of population -.320 .0438 Number of news releases per county -.158 .2041 60 Table 8 looks at the same variables except that the implicit underlying relationship is exponential. All of the coefficients were significantly different from zero at the .0503 level or better. Five out of seven of the coefficients decreased (approached -1) when this implicit exponential farm was used. Thus it would appear that the exponential functional form more accurately describes the situation. Analysis of Regional Agencies The same procedure that was used in the analysis of the state level agencies was used with regional agencies. Tables 9 and 10 de- pict these relationships. With the implicit linear relationship (Table 9) two of the coefficients, public participation budgets per county and public participation staff per county, appear to have signs that contradict expectation. This may be the result of misspecification of functional form and the presence of outlying data points. When the exponential functional form (Table 10) was used the signs of all the coefficients became negative, although it would be difficult to suggest that the coefficients of the public participation staff per county and unit of population were significantly different from zero. In each case the coefficients of the variables that represent direct program outputs are all negative and significantly different from zero. These variables, public participation hearings per county, public participation hearings per unit of population and number of news releases per county all represent final products of the public participation process. The other variables represent intermediate steps or products in the public participation process. 61 Table 8. Correlation Between the Number of Counties in the Planning Area and Infbrmation Subsidy Variables for State Agencies with an Implicit Exponential Relationship Dependent Variables R Significance of R Public participation hearings per county -.455 .0044 Public participation hearings per unit of population -.474 .0096 Public participation budgets per county -.295 .0503 Public participation budgets per unit of population -.297 .0492 Public participation staff per county -,535 .0009 Public participation staff per unit of population -.380 .0377 Number of news releases per county -.329 .0377 62 Table 9. Correlation Between the Number of Counties in the Planning Area and the Information Subsidies for Regional Agencies with an Implicit Linear Relationship Dependent Variable R Significance of R Public participation hearings per county —.301 .0024 Public participation hearings per unit of population -.207 .0300 Public participation budgets per county .193 .040 Public participation budgets per unit of population -.351 .0007 Public participation staff per county .236 .0156 Public participation staff per unit of population -.221 .0244 Number of news releases per county -.207 .0402 63 Table 10. Correlation Between the Number of Counties in the Planning Areas and Information Subsidy Variables for Regional Agencies with an Implicit Exponential Function Dependent Variables R Significance of R Public participation hearings per county -.469 .0000 Public participation hearings per unit of population -.329 .0012 Public participation budgets per county -.252 .0112 Public participation budgets per unit of population -.322 .0019 Public participation staff per county -.O32 .3865 Public participation staff per unit of population -.105 .176 Number of news releases per county -.329 .0024 64 None of the coefficients are indicative of tremendously strong relationships. While this may be somewhat disturbing from a statis- tical standpoint, from a policy perspective it should be intuitively obvious that a multitude of other factors, some measurable but most not, operate to effect the dependent variables. 0n the other hand, in both types of agencies as the jurisdic- tional boudaries of the planning process increased a noticeable decline in the values of the information subsidy variables was noticed. It would seem that as the size of the planning area increased structural barriers were imposed on the information subsidy process. If we view information subsidy activities as transactions, then it is possible to suggest that as the size of the planning area increases, the trans- action costs of information subsidy activities also increase. While the analysis is probably not exact enough to specify the functional form of this cost function it is exact enough to suggest that it is an increasing function. Conclusions 0n the basis of the between-group analysis and the within-group analysis it is possible to draw the following conclusions: 1. Regional agencies provide more information to general audiences than state agencies. 2. Regional agencies subsidize information to local decision makers on a larger scale than state agencies. 3. State agencies subsidize more information to state level decision makers than regional agencies but this difference is not highly significant. 4. As the jurisdictional boundaries of both state and regional agencies increase their ability to subsidize information at the local level decreases. 65 If one assumes that the degree of information subsidy provided to local decision makers is positively related to the probability that a plan will be implemented then it can be inferred that: 1. In general, plans drawn up by regional agencies have a greater probability of implementation at the local level than plans drawn up by state agencies. 2. In both state and regional agencies as the jurisdictional boundaries of those agencies' planning area increase the probability of a plan being implemented at the local level decreases. Limitations of the Analysis It would appear that some strong conclusions can be made on the basis of this analysis. At the same time this analysis is not without limitations and those limitations ought to be made explicit. The first limiting factor concerns the observation that this analysis has only dealt with formal information systems. That is, there was no means of measuring informal exchanges that may have occurred between and among the various classes of actors in the planning process. In- deed it is almost a matter of definition that infbrmal information systems are not quantifiable. At the same time it should be recognized that they are extremely important. Any one who has ever worked in or observed bureaucratic or governmental behavior would not be taxed to conclude that many decisions are made on the basis of information which was exchanged far from the public eye. This analysis also tried to measure the information variables on a quantitative as opposed to a qualitative basis. While, fbr example, the number of public participation hearings was measured, no attempt was made to ascertain what was said at them. It should be 66 clear that what is said or not said at a meeting, and the manner in which it is said will impact very heavily on the way that information is perceived and discounted. The data, then, impute the same weight to a meeting in which a calm discussion of the pros and cons of non- point pollution control are discussed and one in which the moderator makes an impassioned plea for the dictatorship of the proletariate. Along the same line no difference was assigned to meeting whether they were attended by one or a thousand observers. The comprehensive measurement of such qualitative factors even it possible were beyond the resources available to the author. No measures were developed for the historical perspective or political climate facing the planner and the relative aides or barriers these factors may have placed in his or her path. In addi- tion, the information subsidy process was measured only as it applied to transactions between planners and various groups in the community. No measure was taken of efforts made by other groups to subsidize information which may have verified or opposed the planners position. Finally this analysis has assumed that the power to implement nany or most of the policies that will be necessary to control nonpoint pollution lie at the local level. In the short run this assumption probably stands on its own merits. The present equilibrium of public powers with respect to land use control are balanced toward local government. It is entirely possible if not probable that many local governments will not find it politically expedient to regulate land use to improve water quality. However, if enough local governments exercise their right not to control land use to further these ends 67 they may in effect create a regulatory vacuum into which the state or federal government may be drawn. In so doing they would usurp some of the local prerogativeS'hiland use control. This nay be a case where local governments' short and long run interests diverge. Indeed a second factor which may be exacerbating this process is the fact that state level planners do have certain economies of scale available to them when subsidizing infbrmation to state level decision makers. The analysis in this chapter has shown that state level planners are interacting with state governments and it is reasonable to assume that they are doing so to improve the probability that state govern- ments may take steps to act in the nonpoint pollution control area. This discussion of the limitations of this study are obviously not intended to discredit this piece of work. Rather they are designed to place this analysis in perspective and to offer some additional insights with respect to other factors which may affect the process. CHAPTER VI SELECTED CASE STUDIES Introduction This chapter will present two brief case studies of agencies with 208 planning responsibilities. One is a state level agency and the other a regional agency. Neither is intended to serve as an example of the normal or average agency in its class. Rather, these case studies are presented to address some of the factors which have limited the statistical analysis presented in the previous chapter. These factors include the agency's historical development, credibility, self-perception and perception of the planning process. Two principle considerations influenced the choice of these agencies. The first was the author's prior familiarity with each organization. Prior knowledge is an important consideration in constructing a judgment sample; it allows the researcher to select cases which can offer meaningful comparison. This need for a meaning- ful comparison brings out the second reason for choice of these two agencies. To an extent these two agencies represent the ends of the section 208 planning continuum. One is a large and complex state bureaucracy with a relatively large planning grant. The other is a much smaller regional agency with few formal powers and one of the nation's smallest section 208 planning grant. These two cases were 68 69 chosen to partially aquaint the reader with the range and diversity that exists in section 208 agencies. New York State Department of EnVironmentaTFConservation New York's Department of Environmental Conservation (DEC) is a large and diverse bureaucracy which is charged with preparing a section 208 plan for the majority of the state. The agency's respon- sibilities include enforcement of fish and game regulations, environ- mental regulations, park management, forest management and a host of related activities. The agency occupies its own five-story office building. The DEC was designated as a 208 planning agency in May of 1976. It received an initial planning grant of $1,959,000. The agency is charged with preparing a-plan for 44 of the state's 57 counties. This area includes three cities with populations in excess of 100,000. However, these cities do not include the state's two largest cities, New York and Buffalo. In addition the planning area contains 11 cities with populations of between 25,000 and 100,000. However, the majority of the area fer which the DEC must draw up a plan for could, by almost any standard, be considered rural. Only 13 of the area's counties lie in standard metropolitan statistical areas (U.S. Environ- mental Protection Agency, 1977). During the first year of its planning activity the agency developed a work plan and negotiated a five-year contract with the Environmental Protection Agency. The DEC divided the portion of the state it was to prepare a plan for into eight regions. 70 Figure 4 depicts these regions which are listed in the key as State Planning Areas. In three of these eight regions the DEC's regional office that serves that region is not located within the boundaries of that region. This fact begs the question of how effectively decentralized this planning process is. One would expect that the transaction costs of interaction between planners, decision makers, and the general public would be higher in these areas. The agencies public participation program has several elements. Each region has a Policy Advisory Committee of which the majority of members are elected officials. In four of the regions the Policy Advisory Committee is the existing Regional Planning Board and in the other regions the Policy Advisory Committees are indepen- dent organizations. According to the DEC's public participation coordinator the Policy Advisory Committees connected with Regional Planning Boards were not as interested in working on 208 planning issues as the boards the DEC set up (Hawkins, 1978). This must represent a difficult choice of whether to establish a new organiza- tion which must gain credibility in a community or to piggy back with an organization which may not see the DEC's goals as consistent with its own. Beyond the Policy Advisory Committees there are Technical Advisory Committees and/or Citizen Advisory Committees in each region. All in all there were, at the time of the author's interview, about 50 people in each region actively involved in this process. The county planners often tended to be the backbones of this process. Indeed it was recognized that the interest level of local political 71 e 2.5: (mu >Cu xmo> 2m: nxhzaou ”3...me uhmuz u2>13 uafllxflaH 320m Hzmzmogm>mo m az_zz<4n_ 1_