|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 VALUING FOREST RESTORATION AND RE CREATIONAL BENEFITS OF A NATIONAL PARK IN ANDEAN COLOMBIA By SERGIO ALVAREZ A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008
2 2008 Sergio Alvarez
3 To my father.
4 ACKNOWLEDGMENTS This thesis would not have been possibl e witho ut the leadership, patience, and encouragement of Dr. Sherry Larkin, my advisor and committee chair, to whom I will be forever indebted for her commitment to this study at each and every stage. I al so want to thank my committee: Drs. Maria del Pilar Useche and Laila Racevskis for their help in the completion of this project. Their constructive inputs were tremendously valuable. Other faculty members at the University of Florida were also very important in the development of this study. I would like to thank Dr. Clyde Kiker, who inspired me to become an economist. Dr. Karen Kainers inputs in the be ginning stages of this study were incredibly helpful. Conversations and cour ses with Dr. Brian Child were in strumental in developing a more concise idea of protected areas and their problems. I would also like to thank Mr. Mario Moreno and the personnel at Concesion Nevados, as well as Mr. Jorge Lotero and Mr. Jairo Villanueva from UAESPNN, all of which were instrumental in the successful completion of the fieldwork and data collection stages. The fieldwork for this study was made possi ble by financial support from the Tropical Conservation and Development Program at the University of Florida an d the Tinker Foundation. To them many thanks.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........7LIST OF FIGURES.........................................................................................................................8ABSTRACT.........9 CHAPTER 1 INTRODUCTION..................................................................................................................11A Brief History of Protected Areas........................................................................................11Benefits of Protected Areas....................................................................................................14Threats to Protected Areas..................................................................................................... .17Economic Valuation of Protected Areas................................................................................. 21Valuation of a National Park in Andean Colombia................................................................ 252 LOS NEVADOS NATIONAL PARK................................................................................... 28The History of Los Nevados National Park (LNNP).............................................................. 28Management of LNNP............................................................................................................31The System of Protected Areas of Colombia......................................................................... 32Geographic Characteristics of the Area.................................................................................. 34Historical Visitation.......................................................................................................... ......36The Future of LNNP............................................................................................................. ..373 NON-MARKET VALUATION: TH EORY AND TECHNI QUES....................................... 41Utility and Choice: The Basis of Economic Value................................................................. 41Non-Market Goods and Services............................................................................................ 42Total Economic Value............................................................................................................44Measuring Non-Market Values.............................................................................................. 46Overview of Non-Market Valuation...............................................................................46The Travel Cost Method (TCM)..................................................................................... 47The Contingent Valuation Method (CVM)..................................................................... 51Using TCM and CVM for LNNP...........................................................................................554 EMPIRICAL APPLICATION................................................................................................59Survey Design and Methods...................................................................................................59Data Collection................................................................................................................59Demographics..................................................................................................................60Awareness and Preferences for Environmental Amenities and Services........................ 62
6 Park Visitation.................................................................................................................64Travel Cost Responses.................................................................................................... 65Contingent Valuation Responses..................................................................................... 66TCM Analysis and Results..................................................................................................... 68Recreational Demand......................................................................................................68Consumer Surplus Estimation......................................................................................... 70CVM Analysis and Results.....................................................................................................71Closed Ended: Dichotomous Choice (WTP) Model....................................................... 71Discussion of Explanatory Variables.............................................................................. 73WTP Model Estimation Results...................................................................................... 75Open Ended CV: Maximum WTP Estimation Results................................................... 78Discussion of Empirical Results............................................................................................. 815 CONCLUSION..................................................................................................................... ..96Implications................................................................................................................... .........96Policy Prescription..................................................................................................................97Caveats and Limitations of the Analysis................................................................................ 98Future Directions....................................................................................................................99Non-Market Valuation in Los Nevados National Park................................................... 99Non-Market Valuation in Devel oping Country Protected Areas....................................99APPENDIX..................................................................................................................................101Interview Script....................................................................................................................101LIST OF REFERENCES.............................................................................................................105BIOGRAPHICAL SKETCH.......................................................................................................109
7 LIST OF TABLES Table page 1-1 List of protected area categor ies and their description (IUCN) ......................................... 274-1 Per capita travel cost information and 7-month visitation rates by state........................... 914-2 Travel cost model results.................................................................................................. .914-3 Consumer surplus estimates by st ate, in thousands COP and USD................................... 924-4 Summary statistics of variables in the contingent valuation analysis................................ 934-5 Results of the dichotomous choice CV experiment, model 1............................................ 934-6 Results of the dichotomous choice CV experiment, model 2............................................ 944-7 Results of the open-ended willingness to pay models....................................................... 95
8 LIST OF FIGURES Figure page 2-1 Organizational chart of the UAESPNN............................................................................. 392-2 Budget of the UAESPNN in nominal COP, 2004-2007.................................................... 392-3 Map of Los Nevados National Park (LNNP).....................................................................402-4 Annual visitation to LNNP (2000-2006)........................................................................... 403-1 Example of utility function that shows diminishing marginal utility................................ 573-2 Total value of 100 units................................................................................................... ..583-3 Components of total economic va lue of Los Nevados National Park............................... 584-1 Income of respondents by socio-economic strata.............................................................. 834-2 Educational leve l of respondents.......................................................................................844-3 Age distribution of respondents.........................................................................................844-4 Respondents awareness of the eco logical services provided by LNNP........................... 854-5 Concern for environmental issu es in Colombia among respondents................................. 854-6 Location of LNNP (in red) and numbe r of visitors in sample by state.............................. 864-7 Distribution of visitation data and survey respondents by state........................................874-8 Dichotomous choi ce responses by card............................................................................. 874-9 Summary of WTP responses from the open-ended inquiry............................................... 884-10 Scatter plot and trend line of mean travel costs agains t observed visitation rates............. 884-11 Predicted values for travel cost model (in COP) for specification 1................................. 894-12 Predicted values for travel cost model (in COP) for specification 2................................. 894-13 Predicted values for travel cost model (in COP) for specification 3................................. 904-14 Predicted values for travel cost model (in COP) for specification 4................................. 90
9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science VALUING FOREST RESTORATION AND RE CREATIONAL BENEFITS OF A NATIONAL PARK IN ANDEAN COLOMBIA By Sergio Alvarez December 2008 Chair: Sherry L. Larkin Major: Food and Resource Economics Protected areas provide several important goods and services to society, yet these goods and services are usually not traded in markets and their economic value is unknown. The nonmarketable nature of protected areas creates a problem of underv aluation, which can result in the apparent superiority of alternate land uses re lative to conservation. The travel cost and contingent valuation methods allow the estimation of recreational demand from survey data, thus shedding light on the economic value of non-market goods and services such as those provided by protected areas. This study uses the travel cost and conti ngent valuation methods in a complementary manner to yield estimates of consumer surplu s from recreational use of Los Nevados National Park (LNNP), located in the Andean region of Co lombia. Park visitors were also surveyed regarding their willi ngness to pay (WTP) for ecological restor ation of areas affected by wildfires that took place in 2006 and degr aded 2,500 hectares (ha) of paramo The analysis consists of four travel co st models corresponding to four functional specifications of recreational demand, two models analyzing re spondents WTP for ecological restoration using a closed-ended dichotomous choi ce format and increased entrance fees as the
10 payment vehicle, and two models analyzing re spondents maximum WTP for restoration using an open-ended follow up to the dichotomous choice exercise. Consumer surplus for recreati onal use of LNNP wa s found to be large relative to the budget of the Colombian parks service, which is responsible for 52 pr otected areas throughout the country. Respondents WTP for ecological restoration was found to be positive and of a modest but significant magnitude. These results suggest that LNNP is a significan t contributor to social welfare in Colombia, and that protection of land for conservation a nd recreation are sound soci al investments. An increase in the entrance fee or some other fund collection mechanism could also provide important funds for improved manage ment and restoration in LNNP.
11 CHAPTER 1 INTRODUCTION A Brief History of Protected Areas Hum ans have been protecting areas of land for different purposes for several centuries and for several reasons. In India, land was firs t set aside to protect na tural resources about 2,000 years ago. Other examples of early protected areas are the Himas or protected grazing grounds of the Arabian Peninsula and the royal hunting grounds of Europe (Eagles et al., 2002). The modern protected area idea, however, did not emerge until 1872, with the designation of Yellowstone National Park in the American West The British quickly followed suit with the establishment of other protected areas in what were colonies at the tim e but would soon become emerging nations: Canada, South Africa, Australia and New Zealand (Eagles et al., 2002). This initial wave of protected area establishment throughout North America and the British colonies followed policies with a distinct exclusionary purpose, namely to exclude local and generally non-western communities in order to ensure a purely natural landscape. The designation of Yellowstone, for instance, brought about the e xpelling of the Shoshone people from their ancestral land (Carey et al., 2000). Paradoxically, many of the newly established parks took the name of the ethnic group that was be ing excluded from their tribal land. This Yellowstone model of fortr ess conservation was the rule rather than the exception for conservation approaches during the rest of the 19th and most of the 20th Century. The international community began recognizing the importance of pr otected areas early in the 20th Century. In 1933 the London Convention, otherwise known as the 1933 Convention Relative to the Preservation of Fauna and Flora in their Natural Stat e, included clauses that called on signatory nations to establish protected ar eas. The New World c ounterpart to the London Convention, the 1940 Convention on Nature Prot ection and Wild Life Preservation in the
12 Western Hemisphere, also called for signatory nations to establish protected areas as a means to conserve biological diversity (Mulongoy and Chape, 2004). The London Convention was later replaced by the 1968 African Convention on Nature and Natural Resources, which contained similar clauses. The idea of protecting large tr acts of relatively undisturbed natural areas would soon find its way into nearly every count ry on the planet. The number of protected areas grew almost exponentially during the 20th Century and by 2002 there were n early 44,000 officially designated protected areas covering near ly 10% of the Earths land su rface (Eagles et al., 2002). International non-governmental organizations (NGOs) such as the World Conservation Union (IUCN), the World Wildlife Fund for Nature (W WF), and United Nations agencies such as UNESCO have been extensively involved in the establishment and expansion of protected areas and protected area networks around the world. Toda y, protected areas are se en as an essential component of any biodiversity conservation stra tegy as well as maintaining some important social, cultural, and economic values (Carey et al., 2000). Recently, the international community has tende d to mobilize at a global rather than continental level. Inte rnational treaties such as the World Heritage C onvention and the Ramsar Convention have developed lists of sites that signatory governments have made commitments to protection under the convention. Most impor tantly, 189 nations signed the Convention on Biological Diversity (CBD) in 1992 during the Ri o de Janeiro Earth Summit thereby committing to protect biodiversity within their borders using protected areas. The CBD, which operated under the United Nations Environment Programme has established the Global Environment Facility, a major source of funding for conserva tion with a component that directly involves protected areas worldwide (Mulongoy and Chape, 2004).
13 The fortress conservation model of the Yellowstone era has given way to a more comprehensive approach, where the role and contributions of indigenous and other local communities are valued and acknowledged. This recent reformulation of conservation ideology has resulted in the emergence of a great arra y of Integrated Conservation and Development Programs (ICDPs) worldwide. The push for involvement of local communities in the management of natural resources can be traced back to efforts in Southern Africa to involve locals in the management of w ildlife in communal areas, which in turn was brought about by the development of policies that emphasized the de volution of use rights for wildlife and other resources to dwellers of communal areas (Murph ree, 1998; 2004). Many ICDPs actively involve local communities in the manageme nt of protected areas, as is the case with Brazils extractive reserves (Mittermeier et al., 2005). In practice, however, mo st protected areas exist in a limbo between the fortress and the ICDP approach, w ith some protected area managers choosing to focus on one or the other. International NGOs a nd donor agencies are also partially responsible for setting the agendas of protected areas and they will commonly push for more of one or the other depending on their own goals or philosophy. The academic debate between these two approaches still endures, and the mixed resu lts of ICDPs have given proponents of the Yellowstone model additional validation (Schwa rtzman et al., 2000; Terborgh, 2000; Wilshusen, 2002). Some nations in the developed and developing world have found a source of national pride and economic opportunities in their systems of protected areas American intellectuals in the late 19th and early 20th Century, for instance, relied on the American Wests scenic beauty and the nations emerging system of protected areas as a much needed proof of national greatness, comparable and maybe ev en superior to any European country (Runte, 1987). In the
14 developing world, Costa Ri cas protected area system, considered to be one of the best in the world (Tangley, 1988), has served as a catalyst for an increase in international tourism and the portrayal of Costa Rica as the prime eco-touri sm destination in the Western Hemisphere. But even in places such as the United Stat es and Costa Rica, where protected areas are lauded as sources of pride and economic opportun ities, many of these parks are riddled by a myriad of problems including (but not limited to ) low funding, lack of demarcation, lack of policing and enforcement of rules and regulations, lack of local recognition, and lack of outreach and educational activities. Prot ected areas in the developing wo rld are increasingly experiencing people-park conflicts or en croachment by local people whos e livelihoods depend on the wood, land, fisheries, or other resources within the park (Tangley, 1988). In some cases where there is a severe mixture of people-park conflicts a nd lack of enforcement, the paper parks phenomenonparks that exist only in paper an d lack management or enforcementhas emerged. Benefits of Protected Areas Protected areas (PAs) provide som e important goods and services, thus making them an important contributor to social well-being. They are the cornerstone of all initiatives for biological diversity conservation worldwide, as they provide a space were ecological and evolutionary processes can continue, and ar e in many cases the only places were viable populations of large animals still remain. Some PAs are also inhabited by indigenous and local people whose traditional lifestyles can be maintained due to the existence of the protected area. Forested PAs provide important hydrologica l services such as flood mitigation, soil conservation, and provision of water for irrigation and consumption purposes. Marine PAs serve as spawning grounds for an astound ing diversity of speci es that are harvested for subsistence and commercial purposes. Most PAs are also open fo r tourism and recreatio n, providing the public
15 with a space to relax and interact with nature (C arey et al., 2000). The im portance of PAs is also reflected in the many international agreements and initiatives that call for the effective protection of ecosystems through their designation as protected sites. Today, biodiversity conservation is perhaps the most widely publicized benefit stemming from the creation of PAs. Given the expa nsion of the worlds human population and the corresponding increased pressure on land, water and other resources, many animal and plant species have become completely or at least he avily reliant on PAs for the provision of suitable habitat. The existence of large and relatively und isturbed tracts of land is also required for the maintenance of ecological and evolutionary processe s, as well as for the preservation of genetic variation within species, which can act as a buffer against extinction (McNeely, 1995). Protection also ensures that the productiv e capacity of ecosystems is maintained. Many indigenous groups also li ve inside PAs, where thei r traditional lifestyles are preserved and their autonomy is respected. The historic and cultural features of importance to them are maintained inta ct through the designation of the site as protected. Brazils indigenous extractive reserves and Indias sacred lakes are good examples of PAs where biodiversity conservation and preservation of cultural and religious heritage are complementary. Without the protection offered by the establishment of park s and reserves, many i ndigenous cultures would face possible disappearance as a consequence of resource pressures and assimilation into the mainstream culture. Continued breakthroughs in biotechnol ogy also depend on the maintenance and preservation of the genetic ra w material that is present in the worlds PAs. Continued improvements in crop yields depend at least in part on the availability of genetic material from the wild counterparts of our modern day crops. Advances in the field of medicine are also
16 dependent on the availability of wild plants and animals th at create special compounds with medicinal qualities. Protected areas are essentially in situ gene banks that ensure the availability of the genetic diversity that powers improvement s in agriculture, medicine, and other fields (McNeely, 1995). The sustainable use of wild plants and an imals is made possible by the provision of habitat suitable for reproduction and for comple tion of the life cycles of these organisms. Protected areas around the world offer this suitable habitat and are thus in part responsible for the feasibility of the sustainable us e and exploitation of many plants and animals. Marine PAs, for instance, serve as spawning grounds for seve ral species of fish, thereby ensuring the sustainability of some of the worlds fisheries. Forest and mountain PAs also pl ay a vital role in the provision of fresh water for human consumption. A disproportionate amount of the worlds drinking water co mes from PAs, and a third of the worlds 100 largest c ities draw a substantia l portion of their dr inking water directly from PAs (Mulongoy and Chape, 2004). The forest cover present in many parks and reserves protects watersheds from phenomena like soil erosion and sedimentation. Large tracts of forest, such as those present in parks and reserves, also create stable microclimates with regular rainfall and predictable temperatures. In coasts and estuaries the presence of mangrove forests and salt marshes, common in coastal PAs, can mitigate the impacts of severe weather events such as hurricanes and sea-level rise (Mulongoy and Chape, 2004). Forested areas near rivers can also prevent or mitigate floods. PAs also mitigate the effects of global climate change by sequestering atmo spheric carbon in the form of biomass. But PAs also provide intangible benefits, such as national pride a nd human inspiration. The landscapes and wildlife contai ned within parks and reserves enrich the human experience
17 through their beauty. They also provide opportunities for comm unity development, education and training, scientific research, recreation and tourism (McNeely, 1995). Some of the benefits of PAs are non-consum ptive use values, which, like recreation, can be enjoyed by many people without diminishing the quality of the protected area. Carbon sequestration and hydrologic regulati on create indirect-use values, as people benefit from these services even if they are not in contact with the pa rk or reserve that creates these services. In the intangible realm, individuals may value the optio n of maintaining resources in a protected area for future use. Others may feel a benefit in th e maintenance of a protected area to be enjoyed by future generations. Some individuals may even value the conservation of a protected area simply to know it exists, even if they do not plan to use or visit it at any point in time. PAs are common pool resources under public ownership that provide a wide array of products and services that contribute to human welf are in different ways. Threats to Protected Areas The IUCN defines a pro tected area to be an area of land and/or sea especially dedicated to the protection and maintenance of biological di versity, and of natural and associated cultural resources, and managed through lega l or other effective means (C arey et al., 2000). Under this umbrella, IUCN defines six broad categories of PAs mostly on the basis of human presence and uses of the area in question. The six categories are summarized in Table 1-1. These categories illustrate both the diversity of the worlds pr otected area system and the wide range of objectives that th ese parks and reserves are inte nded to meet. While the IUCN classification system is used by governments a nd organizations worldwide, governments are in no way bound to use this system and in many cases sovereign governments will use different systems. The United Nations also designates so me PAs as World Heritage Sites or Biosphere Reserves, and international agreements such as the Ramsar Convention on Wetlands and
18 Migratory Waterfowl also use a di fferent set of criteria for defi nition of PAs. There is also considerable academic debate regarding the ad equacy of the IUCN classification system. Despite the growth in the number of PAs and their geogr aphical extent throughout the world, relatively few are considered success stor ies. Conservation organizations around the world have recognized the overwhelming impor tance of these enti ties in biodiversity conservation, as is reflected in the conser vation literature and in funding initiatives of international NGOs and multilateral donor agencies. Yet all these actors r ecognize that protected area management has fallen quite short of success; instead, most parks and reserves are plagued by encroachment from agricultu re and mining, oil exploration and extraction, wildlife poaching, and fuel-wood gathering among other problems, resulting in environmental degradation and biodiversity loss (Reed, 2002). Lack of adequate funding for the enforcement of PAs is often cited as the major difficulty that PA managers face (Carey et al., 2000; Sc hwartzman et al., 2000; Reed, 2002; Mulder and Coppolillo, 2005). This lack of funding is reflec ted in the understaffing of protected area authorities and the under-investment in proper infr astructure and community outreach activities. Given the wide breath of the goals and objectives of PAs, the lack of proper funding and staffing almost unavoidably results in dy sfunctional and inefficient park agencies full of well-meaning and determined employees, all caught in a quagmir e of problems and thr eats that are unsolvable due to their lack of institutional capacity. Parks and reserves are in th e public trust and are financed with government funds. Yet there is a very high opportunity cost on public expenditures, not only in the developing world but in the developed world as well, and conservatio n is usually low in th e list of government priorities. Furthermore, the nature of some of the goods and services provided by these PAs
19 makes them virtually un-tradable in existing markets. Since governments cannot readily quantify the value of PAs, low funding on them is easily justified. Since these natural areas provide mostly non-market values, those areas which are not under protection are generally lost to other forms of development (Barzetti, 1993). Areas that are under official protection but are lacking of enforcement can, however, also be sus ceptible to transformation into other land uses. Nearly all PAs in the world, however, are under some form of threat. Some of these are minor or remote threats that do not necessarily endanger the existence or viability of the protected area in question. Other parks and reserves, however, are under major or more accentuated threats that could potentially undermin e the existence of the protected area and the resources contained within it. WWF has identified four major categories of threats to PAs (Carey et al., 2000). These general categories do not relate specifically to events or activities, but rather to the overall effect that an event or a conglomer ate of activities may have on a protected area. The first major threat category is the removal of individual elements (plants, animals or other resource) from the PA without causing a major alterations in ecosystem structure or function. Some examples of practices that can be included in this ca tegory are traditional hunting by local communities, extraction of plants for use in the preparation of meals or traditional medicines, fishing and fuel wood co llection. Given the close proximity of human settlements to many PAs, these th reats are expected due to thei r long history of use and do not represent a serious threat to th e existence or sustainability of the area. However, population increases due to immigration or improvements in health care can intensify these practices and turn them into more serious threats.
20 The second major threat category identified by WWF is the overall impoverishment of the ecology of the area, which is generally the result of long term interactions between human settlements and PAs. Air and water pollution, for instance, can diminish the quality of a protected area over time. Agricultural encro achment and urbanization, as well as constant poaching pressure, can also diminish the quality and overall ecological in tegrity of a park or reserve (Carey et al., 2000). A recent study has shown that many Amazonian vertebrates, for instance, are easily driven to extinction in loca l areas when both subsistence hunting and habitat fragmentation are present (Peres, 2001). Ecologi sts have also describe d what they call edge effectsradical changes to the structure and function of ecosystem s at the edge of a continuous expanse of forest, which can result in a different composition of plants and animals, decreased humidity, and a higher incidence of fires. Some species of birds are known to avoid the forest edge and are unable to cross even small road clearings (Lau rance et al., 2004). Major conversion and degradation of the ecosystems due to the removal of vegetation for the construction of roads or major human settlements, and the introduction of mining activities to a protected area represent the third major threat category. The promise of increased revenues often pushes governments toward the opening of na tional parks and reserves to oil exploration and drilling, as was the case in Ecuador with the opening for drilling of Yasuni National Park, the largest protected area in the country. In th e United States, the opening of the Artic National Wildlife Refuge (ANWR) in Alaska for oil expl oration and drilling has been mentioned by the media and policymakers at several points in time, only to be strongly opposed by environmentalists (Waller, 2001). The construction of major infrastructure and the removal of a significant amount of vegetative cover from a prot ected area can imperil the existence of such area not only through the short-term destruction, but also through the long term dynamics that
21 follow the initial event. The construction of roads or settlements may open the way for increased immigration to the region, fragment the forest into smaller patches, and increase human access to areas that were previously inaccessible, thus increasing human pressure on the ecosystems. Forest fires can also dramatically change or co mpletely remove the vege tative cover of an area, also opening the way for activities such as li vestock grazing and farming, which will likely further reduce the quality of the protected area. The fourth and most dramatic threat is the isolation of PAs by major changes in surrounding land or water use. Isolation works very slowly and usually ta kes a very long time to show the dramatic changes brought about by events such as fire or removal of vegetation. Isolation prevents the exchange of genetic ma terial that is the ba ckbone of evolutionary processes. Some species of anim als, specially the larger ones, al so require very large ranges to maintain viable populations. Isolation may also increase inbreeding among animals, which will in the long-term result in higher incidences of gene tic predispositions to certain diseases. In the long term, some species will disappear from PAs that are not la rge enough or lack connectivity to other areas where other populations of that species are present (Mulder and Coppolillo, 2005). The effect of isolation on the integrity of ecosystems is the subject matter of Island Biogeography, an emerging field within conservation biology. Economic Valuation of Protected Areas The economic value of national parks and other PAs, especia lly in developing countries, is often overlooked in public debate. In m any cases, conservation has consisted of conserving that which conserves itself, not by being financially sustai nable but by being economically marginal and having little utility and a redu ced economic value (Wyngaarden 2002). In other words, areas that have been designated as protecte d often provide little if any losses in terms of perceived opportunity costs accruing to those who decide whether or not to designate the
22 protected area. Economic valuati on of PAs is, therefore, an impor tant source of information both for park managers and for society in general. Economically efficien t resource management requires knowledge of the flow of park benefits and costs, and valuation can be used to measure the benefits derived from the existence of the park (Mathieu et al., 2003). The existence of public benefits derived from a PA in the form of environmental amenities and ecosystem services implies that the park contributes to publi c welfare and loss of the pa rk or decline in park quality could result in a loss in welfare (S hah, 1995). Luckily, economists have designed empirical methods to estimate the economic values associated with PAs and other environmental amenities and services (e.g., Mathieu et al., 2003; Maharana et al., 2000; Whittington, 1998; Navrud and Mungatana, 1994). The Travel Cost Method (TCM) and the Contingent Valuation Method (CVM) are widely accepted for non-market valuation of natural system s and amenities. The TCM uses travel costs to place minimum values on visits to sites for re creational purposes. The CVM asks the public to evaluate hypothetical scenarios involving changes to the environm ent that include the estimated costs. The CVM has been approved by the US Department of the Interior for implementing regulations under the Comprehensive Environmen tal Response, Compensation and Liability Act (CERCLA) of 1980 and its amendments of 1986. The CVM has also been accepted by US courts since a 1989 decision that allowed the in clusion of nonuse values generated through the CVM in the assessment of environmental damage s from the Exxon Valdez oil spill in Alaska (Van Kooten and Bulte, 2000). The TCM was esse ntially sanctioned by th e US government with a study on eliminating recreational use of the Hells Canyon by Resources for the Future; based on the results of the study, Congr ess voted to prohibit further development of Hells Canyon for generating electricity from hydropower (King and Mazzota, 2000).
23 To date, numerous travel cost and continge nt valuation studies have been conducted on areas in the developing world to measure a di verse array of environmental benefits. In developing countries, studies init ially focused on issues related to valuing water supplies and sanitation, recreation, and tourism, but the areas have expanded to include surface water quality, and biodiversity conservation (Whittington, 1998). Several studies concerning PAs in the developing world have estimated benefits and valu es accrued to foreigners rather than locals (Horton, et al., 2003), probably due to their highe r incomes and consequently higher valuation estimates. While there have been numerous TCM a nd CVM studies to measure environmental benefits in developed countries and some in deve loping countries, there a ppear to be very few studies that have used both nonmarket valuation methods part icularly on protected areas in developing countries and that have tried to capture valued held by its citizens. But using both methods provides an estimate of the recreational value of a site and the st ated preferences of visitors with respect to changes in the site. The use of information from visitors ignores values held by non-users but also is expected to provi de more realistic esti mates since users have experience and knowledge of the site in question. Navrud and Mungatanas (1994) valuation of flamingo viewing in Kenya was one of the milestones of non-market valuation in the develo ping world. The study consisted of a survey of a random sample of 185 visitors to Laku Nakuru National Park in Kenya, a protected area that serves as a bird sanctuary and supports 1.4 milli on flamingos. The researchers interviewed 58 Kenyan residents and 127 foreign vi sitors to the park. Their qu estionnaire queried them about their travel costs and asked them some continge nt valuation questions. The mean observed value per visitor per day was between $68 and $85 for the Kenyan residents and between $75 and $79
24 for non-resident tourists. The c ontingent valuation exercise yi elded a willingness to pay (WTP) for a higher entrance fee for park manageme nt of $53.25, a WTP for flamingo protection of $19.44, a WTP for the setup of a WWF flamingo fund of $21.88, and a willingness to accept (WTA) of $86.97 for visiting if there were no flami ngos in the park. The statistically significant variables with a p-value of 0.10 or lower were travel cost, income education, and age, as well as an additional household income vari able for Kenyan residents only. Maharana et al.s (2000) study of the sacredness value of a lake in the Sikkim Himalayan region of India consisted of surveys given to 50 members of the local community, 140 local pilgrims, 95 Indian tourists, and 75 international tourists. The m ean observed value for visitors to the lake was $3.87 per visitor pe r day. The contingent valuati on exercise yielded a mean WTP for the non-market commodity benefits of the lake of $0.88 for members of the local community, $2.16 for local pilgrims, $2.51 for Indian touris ts, and $7.19 for non-resident tourists. The significant variables at a p-value of 0.10 or lower were travel cost, distance, age, sex, education, occupation, and income. Mathieu et al.s (2003) study of internationa l tourists willingne ss to pay to enter a marine park in the Seychelles consisted of a CV su rvey administered to 30 0 tourists at different hotels in the Seychelles who had not yet visited a marine park. The researchers tried to link expectations and motivations of the visitors to their WTP. The results showed that the average WTP for entering a marine park was $12.20 per pers on. Given that the exis ting entrance fee is $10, there is a $2.20 consumer surplus per pers on, which would amount to a total annual consumer surplus of $88,000. The significant variables at a p-value of 0.10 or lower were the diving motivation, the prevention of ecosystem destruction motivation, the conservation
25 principle motivation, country of origin, engagement in snorkeling, and visits to other parks in the Seychelles. Valuation of a National Park in Andean Colombia While the environm ental valuat ion literature is strong, info rmation on the economic value of PAs in developing countries is scarce, especia lly for residents in the developing country. The goal of the proposed study is an economic valu ation of LNNP by Columbians using both the TCM and the CVM. The park, which is descri bed in detail in Chapter 2, suffered wildfire damage in July 2006. The objectives of this st udy include: 1) estimati ng the recreational use value of the park for visitors and 2) estimating their willingness-to-pay (W TP) for restoration of the areas affected by the 2006 fires. More specifically, the stu dy will discern which demographic variables affect vi sitors WTP and which park attr ibutes most affect peoples values. The study will also test whether informing visitors of the ecologica l functions of the park affected their WTP. The underlying issue, howev er, is whether Colombians in general accrue a welfare gain (more benefits than costs) due to the existence of th e park, which is information that could help the Ministry of the Environment jus tified continued expenditures on the park system. This thesis is divided into five chapters. Chapter one introduced the concept of PAs by first outlining the history behi nd the establishment of park s and reserves, followed by a description of the benefits that society receives from the existen ce of these parks and the threats that modern parks and reserves now face. Some institutional problems of PAs were also identified. The importance of economic valua tion in parks and reserves throughout the world were then discussed, followed by a brief overview of valuation studies th at have used both the TCM and CVM to value PAs in a developing count ry, including those by its own residents. Chapter 2 offers an introduction to Los Nevados National Park (LNNP) the site of this case study. The chapter begins with the history of LNNP, based mos tly on the legislative documents
26 that created the park and the government agency in charge of its management. A summary of the management of the park and the park system as a whole is then provided. A short geographic description of the area is then gi ven, including a descri ption of the recreati onal visitation to the park. A section describing the future of the park in terms of main threats faced by LNNP is also discussed. The chapter ends with a brief ove rview of the financial future of the park. Chapter 3 introduces the economic theory behi nd non-market valuation of natural assets. The first section introduces the c oncepts of utility and choice, cen tral to all economic theory. A brief explanation of non-market goods and servi ces is then followed by an introduction to the concepts of willingness to pay and total economic value. The travel cost and contingent valuation methods are then explained in more detail. Chapter 4 describes the empirica l application of the travel cost and contingent valuation methods in LNNP. An overview of the survey de sign is then followed by the descriptive results of the survey. A section describing the use of th e travel cost methods a nd its results is then followed by a similar section describing the use of the contingent valuation method in this study. Chapter 5 offers some concluding remarks, be ginning with the implications of the results of the analysis of the travel cost and continge nt valuation data. Some policy prescriptions are then offered, followed by a brief overview of the caveats and limitations of this study. The possible directions that this st udy may take in the near future are then discussed, followed by a brief overview of the future of non-market valuation of PAs in the developing world.
27 Table 1-1. List of protected area cat egories and their description (IUCN) Cat. Definition Description I (a) Strict Nature Reserve; (b) Wilderness Area (a) An area of land and/or s ea possessing some outstanding or representative ecosystems, geol ogical or physiological features and/or species available primarily for research and/or environmental monitoring; (b) Large ar ea of unmodified or slightly modified land and/or sea, retaining its natural character and influence, without permanent or significant habitation, which is protected and managed so as to preserve its natural condition II National Park A natural area of land and/or s ea designated to (a) protect the ecological integrity of one or more ecosystems for present and future generations; (b) exclude ex ploitation or occupation inimical to the purposes of the area; and (c) provide a foundation for spiritual, scientific, educational, recreational and visitor opportunities, all of which must be environmentally and culturally compatible III Natural Monument An area containing one or more speci fic natural or natural/cultural feature which is of outstanding or unique value because of its inherent rarity, representative or aesthetic qualities or cultural significance IV Habitat/Species Management Area An area of land and/or sea subject to active intervention for management purposes so as to ensu re the maintenance of habitats and/or to meet the require ments of specific species V Protected Landscape / Seascape An area with coast and sea, as ap propriate, where the interaction of people and nature over time has prod uced an area with significant aesthetic, ecological and/or cultu ral value and often with high biological diversity. Safeguarding th e integrity of this traditional interaction is vital to the protec tion, maintenance and evolution of such an area VI Managed Resource Protected Area An area containing predominantly unmodified natural systems managed to ensure long term protection and maintenance of biological diversity while providing at the same time a sustainable flow of natural products and serv ices to meet community needs
28 CHAPTER 2 LOS NEVADOS NATIONAL PARK The History of Los Nevados National Park (LNNP) The em ergence of protected areas (PAs) in Colombia has a utilitarian and grassroots basis: the first areas for cons ervation in the country emerged in the first half of the 20th Century as a means to protect the headwaters of rivers that fed municipal aqueduct systems. In 1938 a portion of the watershed of the Cali River was d eclared a Forest Reserve Zone and in 1943 an area surrounding the Mua River near Bogota was declared as protected and fishing and hunting were banned. Other municipalities followed suit, declaring Areas of Flora and Fauna Conservation in the headwaters of the rivers and streams that fed their aqueducts (Fundacion Gabriel Piedrahita Uribe, undated). The national system of protected areas first emerged in 1948; the Serrania de la Macarena a low-rising massif east of the Andes that is geologically part of the Guianan Shield (Franco et al., 2007) was declared a Natural Reserve in order to protect its biodiversity, which is markedly different than the biodiversity of the Eastern Andes. The history of National Parks in Colombia begins in 1959 with the Law on Forestry and Natural Renewable Resources, which created the institution of National Natural Parks with the objective of conserving the national flora and fauna. In article 13 of what is known as the 2nd law of 1959 it is stipulated that the National Government through the Ministry of Agriculture and with the consent of the Colombian Academy of Exact, Physical and Natura l Sciences was from that point on granted the authority and responsibility of establishing protected areas through ex ecutive decrees. The law allowed the establishment of PAs that included prohibitions on the granting or sale of lands for use as hunting, fishing and all industrial ac tivity that involves livestock or agriculture. A paragraph at the end of this article declares the Nevados and the areas around them National
29 Natural Parks. It is unclear whet her the legislators intended to cr eate what is today known as Los Nevados National Park (LNNP), or whether they in tended to create National Natural Parks in all areas where Nevados, or glacial peaks, are present. Arti cle 14 of this law declares the areas designated as National Natural Park s as areas of public utility and allows for the expropriation of lands or private infrastructure that exists in those lands. The official establishment of the first nati onal park did not come until 1960 when Cave of the Guacharos National Park was designated. The parks main features are extensive Andean forests and a system of caves and underground form ations that serve as nesting grounds for the Guacharo, a species of bird. More importantly, the park protects the headwaters of the Magdalena River, the most economically important waterway of Colombia. The establishment of Cave of the Guacharos National Park portrays the dual role of PAs in the Andean region of Colombia as reserves of flora and fauna and hydrographic reserves to protect the countrys waterways. Even though the legislative branch mandated the protection of the glacial peaks since 1959, the establishment of Los Nevados National Parkcomprised of five glacial peaksdid not take place until 1973. The establishment of the park was preceded by the establishment of the Institute for the Development of Renewabl e Natural Resources (INDERENA) as an agency under the Ministry of Agriculture in 1968 that w ould consequently manage and create National Parks and other PAs. In March 1973 INDERE NA designated a 38,000 he ctare area as Los Nevados National Park (LNNP), with the objective of conserving the flora, fauna and natural scenic beauties, and with scientific, educa tional, recreational, and aesthetic uses. In the same decree used to create the par k, INDERENA delineates the borders of the park using trails, roads, creeks, and rivers as markers. The language used in the border delineation
30 sheds some light into the arbitrariness of the proce ss, as it seems that the only instrument used to establish the borders of the park was a topographic ma p of the area. It also seems that at the time INDERENA was not worried about the landholders th at would be affected. The decree uses the same language prohibiting activities in the areas designated as Nationa l Park that was used in the 2nd law of 1959. Today, LNNP is comprised of 58,300 ha, over 20,000 ha more than its initial size. The growth of the parks size has been the result of consequent designations of more protected land. The written record for these designations is, however, lacking. The statute of reservations of the National Parks System, established in 1971, delineates the following eight general obj ectives of the park system: Technical oriented management and use of the reserves that make up the system. Protect and study the wild fa una and flora of the nation. Conserve and administer the na tural values of the country. Reserve representative areas that allow th e perpetuation of the respective primary ecosystems. Establish natural gene banks. Promote the development of new and improve d techniques for the conservation and use of natural renewable resources. Restore wildlife. Research the values of the natura l renewable resour ces of the nation. Given that INDERENA was establishe d to fulfill these objectives th rough the creation of PAs, it is reasonable to assume that the protected areas that were established after the creation of INDERENAsuch as LNNPhad the same objectives Also, even though not mentioned in the official decree of 1971, the Colombian Parks Unit pr ides itself in protecting a majority of the hydrographic centers of the country (Franco et al., 2007) and the protection of the headwaters of many rivers could be taken to be another one of the Parks Systems objectives.
31 Management of LNNP In 1993 INDERENA disappeared with the creati on of the Ministry of the Environm ent, a result of the new constitution of 1991. The Speci al Administrative Unit of the National Natural Parks System (UAESPNN) was created under the Mini stry of the Environmen t with the task of managing existing PAs and designating new ones. A 1996 decree from the Ministry of the Environment emphasized that the UAESPNN also assumed the power of expropriating and negotiating land for the establishment of PAs. A 1995 decree from the same Ministry allows the UAESPNN use of police powers to protect national parks and other reserves. The park includes an area commonly used for livestock grazing and some of the areas that today are under protection used to be gr azing grounds utilized by the same group of people who own the land around the park and who currently use the buffer z one that surrounds the park. This creates a conflict that sometimes results in cattle being found within the core area of the park. Pastoralist fires outsid e the park and in the buffer z one are very common (Wyngaarden and Fandino-Lozano, 2002). Therefore, the UAESP NN personnel in the park spend considerable time and effort driving cattle out of the core area and monitoring the area for fires. Typically, however, there are less than 10 UAESPNN employees covering the 58,300 hectares of the park. Several of the people employed by UAESPNN in Los Nevados are locals that have been working at LNNP for a long time and their commonly held vi ew is that the arbitrar iness of the process by which the park was declared is the source of much discontent among those living near the park (Jairo Villanueva, pers comm.). In 2006 the management of the park experienced a drastic change. A concession contract was signed between the UAESPNN and a union between several private and semi-private entities that includes a tourism agency for the management of the eco-tourism aspects of the park. The contract included a clause that ma ndates the investment of $600 million pesos (about
32 $300,000 USD) in infrastructure that would become the property of the park at the end of the 10 year agreement, if it were not to be extended. The investment plans incl ude the construction of new cabins for tourists and an aqueduct that would supply drinking water to the old and new cabins. It also allows the limited personnel of UAESPNN to concentrate on park management instead of looking after the over 60,000 annua l visitors that the park receives. The main LNNP office, which houses the park di rector and the rest of the executives and office staff, is located in Manizales, the clos est major urban center. The park has about 25 employees, of which 10-13 are housed in the Manizales office and very few of them visit the park on a regular basis (Jorge Lotero, program director for LNNP, pers. comm.). The rest of the employees work in the field and are housed inside the park during the peri ods in which they are working. Generally, field employees work three consecutive weeks and get one week off before their next work period. This implies that at any given point in time there may be somewhere between 8-12 field employees stationed inside the park. The System of Protected Areas of Colombia The Colom bian Protected Area System (Unidad Administrativa Especial del Sistema de Parques Nacionales Naturales de Colombia UAESP NN) is currently composed of 53 protected areas that include both Nationa l Natural Parks and Sanctuari es of Flora and Fauna. The parks system gets its funding from two main s ources. Most of the budget is covered by direct transfers from the Ministry of Environment, wh ich covers both operational costs and some funds for capital investment projects. Investment funds primarily come from FONAM, the National Environmental Fund. FONAM is an independently managed system of bank accounts set by the Ministry of Environment designed to comple ment environmental projects included in the National Development Plan or projects taking pla ce inside the areas that make up the Colombian PA System. The external (forei gn) funding that the parks receive comes through the Ministry of
33 Environment first, then to the central office of the Protected Area System and then to the individual parks, but is usua lly earmarked for specific projects rather than for normal expenditures. All proceeds from entrance fees an d other funds collected by the individual parks are sent back to the Ministry, which then allocates the money back to the parks. Every year the central office of the park s service presents compiled lists of all expenditures made by all of the 53 PAs. This pr ocess implies that there is little autonomy in financial decision making at the field leve l as spending petitions must go through the bureaucratic chain, usually taking a long time to be accepted or may even be denied. Furthermore, the organizational structure of the UAESPNN (Figure 2-1) dictates that the finances are completely separated from the executive directors, which are under the Territorial Directors. In Figure 2-1, the LNNP manageme nt would be under the Norandina Territorial Directorship (Direcciones Territoriales), while all the finances of the UAESPNN are managed under the Administrative and Financial Sub-dire ctorship (Subdirecci on Administrativa y Financiera). Given that the PAs systems finances are cen tralized, an individual financial assessment of LNNP is not possible. However, an analysis of the expenditures of th e system as a whole is useful. Figure 2-2 shows the trends in expenditu res for the entire park system from 2004 through 2007. As shown, 2007 is the first year in which total investment in the parks will exceed operational costs. Not only is 2007 a great improvement in the fi nances of parks in Colombia with the increase in investment relative to operationa l costs compared to the recent past, or in the adjudication of concessions to manage the eco-tourism aspects of certain parksincluding LNNPbut also in the prospects of an in creasing budget for 2008. Next year, UAESPNNs
34 budget will increase from about 13 billion pesos to about 53 billion pesos. This drastic increase seeks, among other goals, higher presence of the parks service in all PAs and more effective management. Nowadays, many of Colombias PAs do not have any presence of the UAESPNN at all, and are instead sites of violent conflic t between illegal armed groups and the Colombian army, which routinely sprays roundup in an attempt to cut down on the production of illegal drugs (Carranza, 2005). While most of the increase in the budget will very likely go to parks in conflict zones, some improvements in the finances at LNNP can be expected as it still remains one of the most socially important protected areas in the country. Geographic Characteristics of the Area Tropical Andean ecosystem s are the richest bi otic communities on Earth, topping the list of the worlds biodiversity hotspots (Myers et al ., 2000). In Andean ecosystems, diversity of ecological communities is enhanced by distinct altitudinal gradients over relatively short distances, with changes in elevation of over 5,000 meters (m) over distances of less than 50 kilometers (km), which are common at equa torial latitudes (Van Der Hammen et al., 1983). The Central Cordillera or mountain rangeof Colombia is a prime example of the extremely diverse biotic systems that the A ndes can harbor. On the Western slope, the Cordillera receives the moist oceanic winds streaming in from the Pacific, resulting in ample rainfall throughout the year. On the Eastern sl ope the predominant winds are continental in character, resulting in a much drier climate with pronounced wet and dry seasons (Van Der Hammen et al., 1983). Los Nevados National Natural Park is representative of the Central Cordillera both from an economic and an ecological stan dpoint (Figure 2-3). Several rivers originate within LNNP, among them the Otun, Quindio, Campoalegre, an d Chinchina on the Western slope and the Totare, Lagunillas, Recio, Guali, Cajones, and Guarino on the Eastern sl ope (Van Der Hammen
35 et al., 1983). Within and around L NNP all possible altitudinal variat ions can be observed: from the year-round glaciated areas above 4,600 m, the paramos above 3,800 m, to the warm equatorial savannas and valleys under 700 m. The predominant ecosystem in LNNP is the paramo a high altitude grassland that is unique to th e neotropics (tropics in the Americas). The most notable species in the park include the Andean bear and the Andean condor, but other tropo-montane flora and fauna are present. Land use around LNNP is characterized by exploitation of the altitudinal a nd topographic gradients. In th e valleys of the Magdalena and Cauca rivers, mechanized agricu lture of rice, cotton, sorghum, a nd sugarcane is predominant. The countrys most important and best know n coffee area is located between 1,000 m and 2,000 m. Above 2,000 m, livestock and tuber crops dominate land use (Van Der Hammen et al., 1983). The four urban centers in the area (Manizales, Pereira, Armeni a and Ibague), with populations between 200,000 and 500,000 each, are located in the coffee belt, betw een 1,000 m and 2,000 m. Given that LNNP falls within the historic al range of the Andean condorthe national bird of Colombiabut had been driven out of the area in the earlier part of the 20th Century, the UAESPNN, along with CORPOCALDAS (the regional environmental authority) and with the support of other national and international organi zations, initiated the A ndean Condor Project, with the goal of establishing a healthy population of condors in th e park. The project began in 1989 and ended with the freeing of 14 cap tive-bred condors in 1997 (CORPOCALDAS, undated). The UAESPNN and CORPOCALDAS continue monitoring the birds today but the success of this initiative is still uncertain. A recent study by Franco et al. (2007) addres ses the success of LNNP with respect to biodiversity conservation. The study compares th e potential ranges of avifauna with the actual records of avifauna occurring within National Parks in Colombia. The study found that more
36 than 80% of the bird species whose potential ra nge includes LNNP are actually found within the park today. This is the second highest presence of potential species in National Parks in Colombia, after the Sierra Nevada de Santa Marta National Park. Historical Visitation The recen t history of LNNP is marked by the November 13th, 1985, eruption of the Nevado del Ruiz Volcano. Previous to this volcanic erup tion the main glacier in the park, which stretched from the summit of Nevado del Ruiz at 5,389 m to areas as low as 4,500 m, provided opportunities for snow skiing and similar activities. A majority of the glacier melted upon contact with the hot pyroclastic flows erupting from the summit of the mountain, creating a disastrous mudflow that levele d the town of Armero, Tolima, resulting in over 20,000 deaths. Due to the extent of the disaster, perhaps the wo rst in Colombian history, authorities decided to close LNNP to all visitors. The park reopened some years later under the management of INDERENA, but entrance to the park required special permits. Visitation during this time was relatively low, as INDERENAs permit processing was costly and tim e consuming for prospective visitors. With the creation of the Ministry of the Environment and the UAESPNN, the permit requirement was eliminated. Visitation to LNNP after 2000 has more than doubled, from 23,593 visitors in 2000 to 58,659 in 2006, peaking in 2004 with close to 80,000 annual visits (Figure 2-4). A map of Colombia showing the states of origin of the ma jority of visitors is shown and discussed in Chapter 4. Tourism in LNNP is seasonal. Its most important attraction, Nevado del Ruiz is located less than an hour drive from Manizales, one of the major urban centers in th e area. The majority of visits occur in the month of January, when th e world famous Manizales City Fair is held. Visitation also peaks in the month of April wh en many Colombians take a week off work to
37 celebrate Easter. In general, visitation is hi ghest during the months of June, July, and August when many schools and universities are not in se ssion. The highest daily visitation in 2007 was recorded on July 20th, Colombias Independence Day, and th e most important national holiday. Visitation has been increasing dramatically sinc e the turn of the Century and it is likely that the annual number of visits will increase in the future. The increase is in part due to president Alvaro Uribes Love Colombia, Trav el Around Her program, which advertises incountry road travel as a key strategy in the fight against the Fuerzas Armadas Revolucionarias de Colombia (FARC) guerrillas. It is very likely that this program will continue for several years. The Future of LNNP In the past, field em ployees were responsible for the management and operation of all aspects relating to the park, incl uding tourism. The situation ch anged drastically in 2006 with the signing of a concession contract that gave the eco-tourism rights to a semi-private entity to be managed as a private business. This freed up th e human resources of the park, as park rangers can now devote their time to park management instead of tending to tourists. It is important to note that any given day, the semi-private conc ession, which is devoted only to the eco-tourism aspects of the park and is not involved in any activity occurr ing outdoors, has more employees inside the park than the actual park manageme nt, since the number of concession employees at the park ranges from 10 to 15 for w eekdays and weekends, respectively. The ability of a government agency to comp ile the expenditures made by all of the 53 PAs in any given year provides a basis for accoun tability in expenditures, but also reflects a heavy centralization in spending by the agency. Employees at LNNP have expressed their dissatisfaction with the difficulties of getting the funds to carry out minimum expenditures like changing light bulbs or fixing the motorcycles th at the rangers use for transportation (Jairo
38 Villanueva, pers. comm.). These difficulties reflect the inefficiencies that result from having long bureaucratic chains designed to create accountability in spending. Wildfires, which start off as pastoralist fi res within or around the park, have become a constant threat to the ecological integrity of the park, and fire scouting has become a daily activity for park rangers. These fires are a common practice in the paramo where locals use it to boost the productivity of the grass and graze more livestock. This is because many locals and absentee ranchers view the park as a commons and ignore the legal pr otection of the park, choosing to graze their cattle within the borders of the park. In a survey with park rangers, fire was identified as the main threat to LNNP (Wyngaarden and Fandino-Lozano, 2002). In July 2006, a series of wildfires swept through the park, burning close to 2,500 ha of paramo It is very likely that thes e fires began as small pastoralist fires but got out of control. The event was heavily covered in the regional me dia and a program for the recovery of the area affected by the fire was proposed by the Minist ry of the Environment. Park managers now emphasize fire prevention as one of th e most important management goals. The future promises a four-fold increase in funding to the entire system of 53 parks in Columbia with the majority to combat illegal pr oduction activities. While this effort may reduce the chance of future wildfires, information is needed on how Columbians value their parks and park restoration. This is because national source s of funding are scarce and parks in particular need ways to justify the conti nued expenditures, incl uding any increases that might improve the ecological services they provide.
39 Figure 2-1. Organizationa l chart of the UAESPNN 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 12,000,000 14,000,000 2004200520062007*1000s of Colombian Pesos Operational Costs Investment UAESPNN Investment FONAM Total Investment Figure 2-2. Budget of the UAESP NN in nominal COP, 2004-2007
40 Figure 2-3. Map of Los Nevados National Park (LNNP) Source: UAESPNN, undated. Available online at http://www.parquesnacionales.gov.co/ 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 2000 2001 2002 2003 2004 2005 2006Number of Visitors Figure 2-4. Annual visitation to LNNP (2000-2006)
41 CHAPTER 3 NON-MARKET VALUATION: THEORY AND TEC HNIQUES Utility and Choice: The Basis of Economic Value Central to ec onomic theory is the notion that peoples choi ces are made in a way that maximizes their happiness and well-being, or ut ility, as Bentham and Mill would refer to it (Farber, 2002). Carl Menger, founder of the Aust rian School of Economics first proposed that the intensity of desireor utilityfor one addi tional unit of a good declines with consumption or use of successive units (Farbe r, 2002). This diminishing marg inal utility can be illustrated with a hypothetical situation in which a person is consuming chocolate-coated strawberries. The first strawberry brings a larg e amount of satisfactionor utilit y. The second strawberry still brings more satisfaction, but not as much as the first (Figure 3-1). By the time the person has eaten the seventh strawberry the point of satiation has almost be ing reached, and eating an eight strawberry brings hardly a ny extra satisfaction at all. More recently, economists have come to re cognize that people receive utility from consumption due to the characteristics of goods, rather than by the goods themselves. Multiattribute utility theory holds th at total utility is a function of the characteris tics of goods and services (Farber, 2002); hence th e utility (U) received by the consumption of chocolate coated strawberries is a functio n of the freshness of the strawberries, S, the quality of the chocolate, Q, and the temperature of the chocolate, T, such that: U = f (S, Q, T) (3-1) In order to maximize utility, individuals mu st allocate their use or consumption among several goods in such a way as to equate their marginal utility for consumption or use of each good (Farber, 2002). If individuals are interacting in a well-functioning market and are facing an income constraint (I), they will allocate their expenditures among different goods such that:
42 Max U = f (X1, X2, Xn) subject to Pi Xi I (3-2) where Xi is a good that can be bought for price Pi. When consumers interact in a market they choose to spend some of their income in the purchase of a good or se rvice and so they are expressing a positive preference for it. This allo cation of income towards the purchase of goods which bring consumers utility forms the basis of the economic value of these market goods. The solution to the problem expressed in equation 3-2 is characterized by a situat ion in which the last unit of each and every good purchased brings them the same utility. In other words, the marginal utility of Xi just equals the marginal utility provided by Xj (where i j). Mathematically, this is expressed as: (3-3) The marginal utilities are just the change in utili ty, which is measured as the slope of the utility function. Non-Market Goods and Services Most, if not all of the m arkets in existen ce today, are trading mechanisms for private goods and services. However, not all the goods and services consumed by individuals are private in nature. Public goods and services, su ch as environmental service flows, are enjoyed by all individuals, yet there is no effective mechanism for allocating them. In this case, people may value a clean environment such as clean air or, more specifically, air with lower levels of different types of harmful particulates. So me public goods and services, such as national defense, which are produced by human enterprises, have costs that are readily available. In the case of private goods there is an ec onomic value over and above the price paid to acquire the good, which is known as consumer surplus. In this manner, the total utilitarian value of a good is the sum of its priceor paid be nefitsand its consumer surplusor unpaid
43 benefits. For example, if 100 units of strawber ries were sold for $10 each, total revenue is $10 times 100or the area of the darkest box in Figur e 3-2. When all individuals are represented, these areas under the demandor willingness to pay (WTP)curve are readily identified. In other words, the total WTP is the sum of the total revenue from all sales and the associated consumer surplus from these individuals whose willingness to pay exceeded the amount that was actually paid (Pearce and Turner, 1990). In Figu re 3-2, this is the shaded area under the demand curve up to 100 units. The same is true for e nvironmental goods and services, whose demand is not expressed in markets, but through the politica l process. The extent to which such demand gets represented, however, is a function of the regul atory and democratic processes that facilitate their expression (Kiker and Lynne, 1997). Even though consumer surplus gives us an easily understandable measure of welfare or well-being, when it comes to measuring it in a non-market scenarioas in the provision of public goods like environmental service flows a nd amenitiesit is no more than an ambiguous measure (Van Kooten and Bulte, 2000). Howeve r, two similar concepts, compensating surplus and equivalent surplus, are valid when it comes to measuring the value to consumers of certain goods and services not traded in markets. Th e difference between these two measurements is that one measures losses in value due to decr eased levels of the public good while the other measures gains in value due to increased leve ls of the same good. Compensating surplus is measured by finding the amount that individuals would be willing to pay to increase the level or quantity of the public good in question. Convers ely, equivalent surplus is measured by finding the amount that individuals would be willing to accept as compensation to forgo the higher level of the public good or to accept th e reduction in level or quantity of the good (Van Kooten and Bulte, 2000). These two measures allow for the tracing of a demand or willingness to pay curve.
44 What is sought in benefit measurement is then a measure of the areas under the demand curves (Pearce and Turner, 1990). Total Economic Value The willing ness to pay or demand curve arises due to the existence of economic values held by the individuals who intera ct in the market. More explicitly, consumers demand a certain good or service because of the economic values they place on the given good or service. In the case of environmental goods a nd services, there are two categ ories of economic values: use values and passive-use values, which are sometime s referred to as intrinsic or non-use values. Use values derive from actual use of the environment and can take the form of consumptive or direct use values (e.g., farming, fishing, hunting), non-consumptive or indirect use values (e.g., wildlife viewing, ecosystem services ), or option values. Option values are complex in that they represent a preference toward preservation for a non-specific use (consumptive or nonconsumptive) sometime in the future (Pearce and Turner, 1990). Passive-use values are more problematic in th at they do not involve direct use or benefit by the individual, yet the individual exhibits a positive WTP towards the good or service in question. There are two main categories of pass ive-use values: bequest values and existence values (Sharp and Kerr, 2005). Bequest values re late to the idea of leav ing a supply of natural environments to ones heirs or to future genera tions in general, hence peoples willingness to contribute towards land trusts that will protect land from development indefinitely, even if they do not plan to see or use the land in any way. Existence values relate to knowing that the resource simply exists and will continue to exis t, hence peoples willingness to contribute to the protection of blue whales, even thou gh they will probably never see one. In thinking about protected areas as economic entities, it is helpful to define them as a conglomerate of various resource s and ecological proce sses that provide a variety of goods and
45 services, rather than as one individual public good. An individual recei ves different kinds of benefits from these different re sources and processes and values them accordingly. However, the individual only values those be nefits of which he is aware of Therefore, the individual will value those goods and services that are used direc tly in a more coherent manner than those which are used indirectly or not used at all. This can be described as a problem of tangibility of benefits. Direct-use of a resourcefishing fo r instanceis easily pe rceived by the individual who is fishing, and so the value for this direct-use is very tangible. Le ss direct benefits of protected areas that accrue to individuals at a global scalesuch as carbon sequestrationare more difficult to perceive by the individual, hence the valuation process occu rs in a less coherent manner. Values for indirect-use or passive-u se of resources are th erefore less tangible. Most direct-use values are accrued by indivi duals at the local level and are expressed in markets for commodities like timber, food, and recr eation. The existence of markets for these goods and services imply that they are mostly priv ate goods. In the case of LNNP, park visitors accrue most of the direct-use values in the fo rm of recreation, but th e park management only captures a portion of that value via entrance f ees. Capturing refers to receiving a monetary compensation for the supply of a good or service; hence those who sell private goods are able to capture some of the value of th at good. When the good is based on a natural resource, the value above the cost, where value is represented by the demand (willingness to pay) and the costs are measured as opportunity costs, is referred to as the resource rent. Th e para-statal concession that has recently acquired the tourism-development rights of LNNP also captures a portion of the recreational value of the park by offering tour packages to the park and by running small shops within the park. Some locals also capture a po rtion of that value (i .e., resource rent) through sales of food and souvenirs along the road that l eads to the park entrance. Recreation is very
46 important for protected areas because it is a nonconsumptive service that can be provided to private individuals without diminishing the quality of the park. Other direct-use values that accrue to locals from consumptive activities include grazing, hunting, and fishing within the borders of the park, though both grazing and hunting inside the pa rk are considered illegal. Indirect-use values transcend the local realm by providing bene fits at a regi onal, national and global scale. Most of the rivers in the coffee region of Co lombia originate in LNNP, hence the hydrological services of LNNP offer a stream of benefits at a regional scale. The value of the hydrologic services of the park is widely recognized and is one of the factors that influenced the creation of the park. However, there is no exis ting mechanism for capturi ng these benefits. By acting as a carbon sink, LNNP offers benefits at a global scal e as well but the park management receives no compensation for the provision of this service. The problem that emerges with goods and services that benefit individuals at scales grea ter than the local level is that the value derived by these benefits is almost never captured by th e park management; this phenomena is referred to as lost rents or rent dissipation if improper management resu lts in the failure to generate or capture resource rents. Specific use and passive use values that have been identified as being provided by LNNP are summ arized in Figure 3-3. Measuring Non-Market Values Overview of Non-Market Valuation Modern societies are forced to m ake choi ces regarding the use and protection of the natural environment on a regular basis. The publ ic decision-making proce ss is inherently linked to societys valuation of that environment (Costanza, 1997). Monetary valuation of these ecosystems is, however, rarely made explicit as mo st of the goods and services provided by them are public or common in nature and not traded in markets. This probl em often leads to undervalued natural systems (i.e., a market failure to economists) whose protecti on is often seen as a
47 decrease in national economic growth, and loss of these ecosystems through inefficient public decision-making is a common outcome. This is es pecially true in deve loping countries, where environmental protection is at the bottom of a list of government priori ties that includes such necessities as public housing, hea lth, food security and education. Economic theory conceptually provides several methods to quantify the value that individuals place on these natural systems in moneta ry terms. In the case of protected areas, which are often used as outdoor recreation sites, the idea of measuring the economic value of recreational use through the expend itures in complementary market goods, such as transportation to the site, was first proposed by Harold Hotelling in a letter to the director of the U.S. National Parks Service in 1947. This use of observable choices to estimate a monetary measure of changes in individuals utility th at would come with a change in a non-market amenity is now known as the Travel Cost Method (Smith, 1996). There is an alternative appro ach, the Contingent Valuation Method (CVM), that can also be used to elicit values for non-market goods. The CVM is the only approach that can be used to capture non-use values. The CVM elicits indi vidual behavior by developing a hypothetical market (such that all observed behavior is contingent on the ficti onal market) and having individuals participate in that market. This ap proach can capture the WTP for a whole suite of non-market values. If applied with a TCM fr amework, this approach would entail use of individual information on travel costs and hypothetical choices. Each modeling approach will be discussed in turn. The Travel Cost Method (TCM) As was noted above, the TCM consists of us ing individuals observable choices in the for m of their actual payment of some of their income for environmentally-based recreation as a basis for estimating the value of a non-market good if it had been available in a well-functioning
48 market (Clawson and Knetsch, 1966). In the case of recreation in public areas such as National Parks, users usually pay a minimal fee or no fee at all; hence park revenue is a poor indicator of recreationists WTP. This is be cause outdoor recreationists usually incur costs in travel and time that could have been used for other purposes (Clawson and Knetsch, 1966). A recreation demand curve can conceptual ly be developed using observations on visitation and expenses incurred during these visits, that is, by determining how visitation rates vary with total travel expenses. Conceptually, vi sitation is then a function of travel costs, such that: Vj = f (TCj) (3-4) where Vj is the visitation rate of zone j and TCj is the average travel cost from zone j to the recreational site. Given that entrance fees to th ese parks are minimal or non-existent, most of the variation in travel expenditures will be due to the distance trav elled, as individuals travelling from farther away can be expected to spend more on their travel than i ndividuals travelling from areas closer to the park in ques tion. The opportunity cost of peopl es time must also be factored into the travel cost calculations as time spen t in recreational activities implies an income tradeoff, since the individual has chosen recreatio n instead of work. Theo retically, as distances and incomes (i.e., travel costs) incr ease, less people will travel to the site or they will do so less often. Differences in peoples incomes are also an important source of variation since people with higher incomes are expected to spend more on recreational activities. Hence, it is assumed that recreation is a normal good. Th is variation in travel cost a nd visitation rates forms the basis of the downward sloping outdoor recreation demand curve (Clawson and Knetsch, 1966). Assuming a linear functional form, e quation 3-4 could be estimated as: Vj = 0 1 TCj + j (3-5)
49 which shows that at zero travel costs, visitation would equal 0 and for each increase in one dollar of travel costs, visitation will fall by 1. If the error term is well-behaved (i.e., no heteroskedasticity due to differences between zones), then equation 3-5 can be estimated with an ordinary least squares procedure that minimizes the sum of the squared variance to find the parameter estimates. The zonal TCMalso known as the Clawson and Knetsch approach (Bowes and Loomis, 1980)divides the visitation observati ons into zones or areas base d on their location relative to the site and uses historical records of visits per capita within each zone as the measure of visitation to the site. While the zonal approach has several difficulties, the estimates can be improved by extending the analysis to include more comprehensive factors affecting values, such as population characteristics. The opportunity cost of time can be included in the calculations by either collecting a large sample of surveys that include detailed inform ation regarding peoples income or by including the socio-economic census data from each zone as a proxy for a sample of detailed individua l information. Given that many outdoor recreation areas keep track of the number of visitors they receive and their places of origin, the zonal TC M can be used to easily quantify the economic benefits of these areas (Clawson and Knetsch, 1966). Thus, its use is wider in scope than more complex and costly (albeit more precise) valuat ion techniques (Bowes and Loomis, 1980). This is especially true in developi ng country contexts, where the ne cessity of conducting non-market valuation studies is ignored due to the prohibitive costs of more comprehensive studies. National Parks in particular are known for their visitor record keeping such that the use of this relatively simple technique in these PAs can be a fruitful area of research and an important source of information for public policy makers. The TC M is particularly well-suited to valuing
50 recreational areas that draw inte rnational visitors and those that visit multiple times each year. This is because in order to cons truct a demand curve the data needs to include a wide variation in the travel cost a nd visitation numbers. Some problems with the zonal TCM approach have been discussed in the literature. Bowes and Loomis (1980) point out that the diffe rences in population size across the different zonal definitions naturally results in heteroscedasticity due to the grouping of observations. When the recreational demand curve is estim ated using Ordinary Least Squares (OLS) regression, the estimated number of visits at a zero entrance f ee and the observed number of visits at a zero entrance fee are different. To solve this problem they propose an estimation of recreational demand using the Gene ralized Least Squares (GLS) regression instead such that the GLS estimators will have the same variance and possibly improve the statistical properties of the estimates. Moeltner (2003) also points out that aggrega tion of the zonal observations will ignore the natural heterogeneity present within the populatio ns, as it is assumed that income within each zone is constant. This problem makes zonal TC M models susceptible to aggregation bias and makes their results questionable. Some of th e heterogeneity of indi vidual visitors can be captured by incorporating distribut ional information on per capita income within the zones, usually available from census sources. The incorporation of income distribution within zones in addition to using GLSthus reduces the lik elihood of aggregation bias and increases the reliability of the model and resulting predictions. One of the notable limitations of the TCM approach in measuring the non-market benefits of environmental amenities is its limite d scope. Since it relies on observations of choices incurred through the use of a recreational area, it is limite d to measuring only direct use
51 values and ignores the rest of the areas econom ic values, which could include passive use, option, bequest, and or existence values. While some of these benefits will accrue to park visitors and they are likely to have a positive WT P to maintain or improve these benefits, their travel expenditures will not provide a measure of that positive WTP. Thus, the TCM will produce conservative estimates of park benefits if non-use values by visitors and non-visitors are positive. This shortcoming can, however, be addressed in part through structur ed conversations in which hypothetical markets and situations are cr eated and visitors (and potentially non-visitors) are asked to evaluate them. This practice, know n as the Contingent Valu ation Method (CVM) is the subject of the following section. The Contingent Valuation Method (CVM) As m entioned earlier in this chapter, indivi duals accrue utility by the consumption or use of goods and services (via their characteristics) and th ey are expected to allocate their limited income among different goods in such a way as to maximize total utility. The use or enjoyment of some goods and services does not require income expenditures by individuals, yet these individuals do receive u tility through the enjoyment or us e of these public or common goods. Since the use of such non-market goods and servi ces does not require income expenditures, there is no discernible behavior al trail left by their consumption or use (Smith, 1996). This does not mean that individuals would not be willing to spend some of their income or make some tradeoffs towards the eventual purchase of th ese goods or services. Under this premise, economists have devised structured conversations or surveys in which a h ypothetical scenario is set in such a way as to presen t the respondent with the choice of paying for the maintenance or improvement of a good or service th at is currently freethat is a non-market good or service. Conversely, respondents could be asked for thei r willingness to accept compensation to forgo a
52 benefit for which they currently have a right. Conceptually, an individuals WTP for an environmental good or service is a function of the price of the good or service (P), the individuals income (I), and othe r characteristics (O) such that: WTP = f (P, I, O) (3-6) Assuming a linear functional form, an estimable equation would be as: WTPi = + Pi + Ii + + i (3-7) where i identifies each individual; and the j s (associated with different characteristics of the individual) are parameters to be estimated; and i is the error term. This method, known as the Contingent Valuation Method (CVM) is perhaps the most widely used of all the non-market valuation m echanisms, and its application in the literature ranges from public projects in urba n areas to the protection of vast and isolated ar eas of tropical forest (e.g., Whittington, 1998; Horton et al., 2003). When it comes to the question of restoration of ecological goods or services, the majority involve va luing the restoration of river flows and riparian ecosystems (e.g., Loomis et al ., 2000; Ojeda et al., 2008). The restoration of native land ecosystems has also been investigat ed in the United Kingdom (MacMillan and Duff, 1998). However, only one of the restoration st udies (Ojeda et al., 2008) was conducted in a developing country. Similar studies have, however, estimated values for improving the conditions of a park or the protection of a species in a developing country using the CVM approach (e.g. Navrud and Mungatana, 1994; Math ieu et al., 2003; Jabarin and Damhoureyeh, 2006). Several problems with the CVM have been identified throughout th e literature and most of them result from the hypothetical nature of the approach. The idea of placing a monetary
53 value on environmental goods or services is unf amiliar to respondents as they seldom think of these goods or services in monetary terms. Respondents may also perceive the same good or service very differently, and th eir conceptualization of the good or service in question may be divergent from that of the researcher. When th ese differences in conceptualization exist, the information given by the researchers via the qu estionnaire may affect respondents answers, resulting in information bias. Furthermore, re spondents have difficulty in valuing parts of an ecosystem separately from the entire ecosyst em, thus creating an embedding effect. If respondents are asked to value items from a list, the order in which the items are listed can also affect the responses. The c hoice of payment vehicle may re sult in protest responses as individuals may be unwilling to pay not because they do not value the good or service but because they dislike the payment vehicle chosen (e.g., taxes). Some respondents may also feel good when answering positively to contingent va luation questions, which results in their WTP being overstated, a phenomenon known as the w arm glow effect. Respondents may also answer positively to questions, th inking that positive answers are wh at the interviewer expects. If respondents are prompted for their WTP through a bidding game, the initial bid will also affect their final response (King and Mazzota, 2000). Another difficulty with the CVM is the diffe rence between the WTP and WTA values for the same goods and the same individuals. Th is phenomenon, known as cognitive dissonance, has been exposed empirically by psychologists who have concluded that individuals get attached to the positions that they hold at the beginning of the experiments, as if that position belonged to them by right. Consequently, their willingness to pay to secure a be nefit is usually not as high as the willingness to accept to forego the same benefit, and the willingness to pay to prevent a loss is usually not as high as the w illingness to accept to to lerate that same loss (Pearce and Turner,
54 1990). While all these problems are still sources of concern for researchers, improvements in questionnaire design have partially or comp letely addressed most of these problems. In the more recent CVM studies, for exampl e, a single bound dichotomous choice format has been used to address some of these concer ns. With this approach respondents go through a dichotomous choice experiment in which they are asked whether or not they would be willing to pay a given amount for the good or service in qu estion, followed by a single bound experiment in which they are asked to elicit their maximum willingness to pay (Ojeda et al., 2008). This format allows researchers to first confront the su rvey respondents with a situation similar to the one they usually encounter when purchasing market goods in which a price is given by the supplier and the respondent has the choice to either make the purch ase or not. The process also elicits the respondents maximum WTP, which w ould be a more precise estimate of each individuals true WTP, or the point on the de mand curve for the non-market good in question. In the closed-ended dichotomous choice portion, differe nt respondents are confronted with different prices, which allows for variation in the price levels. Conceptually, this process should result in a downward sloping demand curve as more respondents are expected to respond positively to low prices and vice versa for the normal non-market good in question. The common practice to analyze the data from these CVM studies is to use a maximum likelihood estimation method for the probability th at a respondent answers the dichotomous choice portion in a positive manner (Jabarin and Damhoureyeh, 2006; Ojeda et al., 2008). This is usually accomplished using a logit or probit model. For example, a probit model is defined as: Pr( Y = 1 | X) = (X ) (3-8) which specifies that the probability that Y = 1 (in this case that an individual said yes to being willing to pay the proposed fee incr ease) is contingent upon the vector of explanatory variables,
55 which is here generically represented by X. Furt hermore, equation 3-7 also states that this probability equals the standard cumu lative normal probability distribution, and X is the probit score. The resulting general log-likelihood function for a probit is ln L = ln (X ) + ln(1(X )) (3-9) The result is an S-shaped curve that runs from zero to one, which is usually very similar to that produced by the logit procedure. The probit procedure is preferred usually because it can be used to more easily extract the aff ect of each variable on the probability. The follow-up maximum WTP question is analyzed using a least squares regression model since the dependent variable is quantitative and continuous in nature (Ojeda et al., 2008). This was shown in equation 3-7. The choice of estimation technique will depend on the assumptions made about, or results of te sts on, the properties of the error term. Using TCM and CVM for LNNP The hum an experience is one filled with deci sion-making moments in which individuals make choices that imply tradeoffs. The free mark et has become an efficient and widely used forum for the expression of human wants and n eeds and the interaction between the wants and needs of many individuals. These choices are motivated by individuals preferences, which are in turn molded by their values and value systems (Farber, 2002). While non-market valuation is not a panacea fo r all the problems and threats faced by protected areas, the relation between many of these problems to the chronic underfunding that plagues them hints at the possibility that re searching their economic values will provide an important piece of information. In the specific case of LNNP, the people-park conflict also indicates that the park is not only underfunded, but that its co nservation value is perceived by some local peasants as inferior to its value as grazing grounds for cattle. Furthermore, some of
56 the objectives delineated by the Colombian government for LNNP and the other PAs of the nation (Chapter 2) may be fulfilled in part by a non-market valuation study of LNNP. Traditionally, the CVM and TCM models have been viewed primarily as substitute valuation techniques. In more recent times, however, the CVM and TCM have been successfully combined to estimate welfare measures for eco system valuation. This has produced a more comprehensive picture of preferences than what would be available from using either method separately (Kling, 1997). This is the appr oach that is used in this paper. The zonal TCM in particular fits the contex t of LNNP, as park users only visit once a year on average. The existence of secondary data on visitation also provides an excellent source of information for a zonal TCM analysis. Unwillingness on the part of Colombians and foreigners visiting Colombia to share information regarding their income also make an individual TCM analysis very difficult to complete. Wh ile the zonal TCM approach may have some problems, the specific context of LNNP make s it not only feasible but also appropriate. The wildfires that took place in 2006 provide an excellent opportunity to value visitors WTP for ecological restoration th rough the use of a CVM framewor k. The possibility of having a self-funding mechanism in which funds are tran sferred from visitors into an ecological restoration program provides an ex plicit and believable scenario in which park visitors can be asked to elicit their WTP for restoration. Use of the TCM and CVM in LNNP is theref ore relevant and feasible. The following hypotheses are then set forth as th e base of the non-market valu ation study in LNNP using the CVM and TCM: LNNP has significant recreational value. Visitors to the park will exhibit a positive WTP for restoration of the damaged areas.
57 Visitors with knowledge of the environmenta l and hydrological services produced by the park will exhibit a higher WTP for restoration. Demographic and socioeconomic characteristics of the visitors are largely responsible for the differences in thei r WTP for restoration. Respondents who express pro-environmental pr eferences (donations and membership to environmental organizations, visits to other environmental amenitie s, etc.) will have a higher WTP for restoration. These hypotheses will be tested using the empi rical results of the TCM and CVM analysis, which is conducted in the following chapter. 0 20 40 60 80 100 120 140 160 180 0123456789 Number of chocolate coated strawberriesUtility Figure 3-1. Example of utility function th at shows diminishing marginal utility
58 Figure 3-2. Total value of 100 units Figure 3-3. Components of total economic value of Los Nevados National Park
59 CHAPTER 4 EMPIRICAL APPLICATION Survey Design and Methods Data Collection Both prim ary and secondary data were used in this study to value th e recreational benefits of LNNP and visitors demand for restoration following wildfires that took place in 2006. The primary data was obtained from 66 interviews with park visitors during th ree weeks in July and August 2007. The interviews were completed at the parks main entran ce and all interviews were conducted by the author. All visitors to LNNP are required to attend a mandatory session in which park rangers explain the possible health hazards of high altitudes and low temperatures. Visitors then purchase their ticket to enter the pa rk. The visitors were approached while they were waiting for the beginning of the information session and were asked to participate in a joint study between the University of Florida and th e Colombian Parks service. All of those approached either accepted the in vitation to participate or aske d another member of their group to participate. However, from a total of 69 attempted interviews, only 66 were completed as three respondents were foreign visitors that we re reluctant to disclose information concerning their income or their travel costs. Of the remaining respondents, only two were international visitors. Given the extremely small number of interviews completed with foreign visitors, they were also excluded from the anal ysis leaving 64 completed survey s all of which were Colombian visitors. The travel cost section of the interview asked respondents to list their expenses for transportation, lodging, equipment rental, and gu idance, and respondents were given the option to answer either as an individual or as a gr oup. Group responses were la ter divided by group size to determine travel costs per person. The contin gent valuation section of the interview began by
60 asking respondents how familiar, if at all, they were with the wildfires that occurred during 2006. Then, they were informed about the magnitude of the wildfires (i .e., number of acres burned) and were confronted by a scenario in which the park authorities are considering an increase in the park admission fee to cover the e xpenses of restoration of the ar eas affected by the wildfire. A card with the proposed fee increas e (one of four values) was th en shown to the respondent, followed by the question: Please consider how much you spend on recreati on each year and for this trip. Would you have been willing to pay this extra amount per person during this visit? The four fee increase levels were 25%, 50%, 75%, and 100% increases of the existing entrance fee of 8,000 Colombian Pesos (COP, which equals approximately $4USD). Regardless of their response, respondents were then asked to identif y their maximum willingness to pay an increased fee as an open-ended question. The questionnaire also included a section th at asked respondents about their knowledge of the ecological services provided by the park. This portion of the questionnaire was administered before the CV exercise in half of the interviews and after the CV exercise in the other half. Respondents were also asked about their knowledge or awareness of ecological services provided by the park, regu lar visits to natural areas or other national parks, membership or donations to environmental organizations, and general concern for the condition of the environment in Colombia. The rest of the quest ionnaire consisted of que stions regarding past visits to the park, the current visit to th e park, and other demographic information. Demographics Of the 64 interviews com pleted with Colo mbian nationals, 40 (62.5%) were completed with men and 24 (37.5%) were completed with women. While this gender distribution is not reflective of the overall population of either Colombians or visitors to LNNP as the proportion of
61 males is too high, it is the result of heads of household (which ar e predominantly male) being the ones responding to the interview. Similar male-bias has been en countered in other non-market valuation surveys conducted in Latin America (e.g., Shrestha et al., 2002). Since Colombia is the site of a decades-long military and civil conflict in which kidnappings and extortion are commonplace, some re luctance on the part of Colombian nationals in disclosing their incomes or expenditures was expected. To avoid th e possible reluctance to answer the most important portions of the que stionnaire, the Colombian governments socioeconomic stratification system was used as a pro xy for income. Under th is system, residential properties are grouped into differe nt strata for taxing purposes, w ith stratum one representing the lowest tax rates (poorest neighborhoods) and st ratum six representing the highest tax rates (wealthiest neighborhoods). Respondents were, therefore, asked in which stratum their home was located instead of directly asking them for their annual income. Even though socioeconomic strata may not be as accurate in reflec ting income differentials as full disclosures of annual income, and the calculati ons of the opportunity costs of respondents time cannot be calculated on an individual basis, using the stra tification system was considered the most reliable way of asking about incomea critical piece of information when estimating non-market valuesin a country like Colombia. While respondents indicated origins from a ll income strata, the highest frequency of response was associated with stratum 3, which ca n be considered lowermiddle class. It was surprising to observe the same number of visitors from the wea lthiest portion of the population as from the second poorest portion of the populati on, and that there was some visitation by the poorest segment of the population (Figure 4-1). This finding is in contrast to previous studies that have concluded outdoor recreation in develo ping countries is a luxury enjoyed only by the
62 wealthiest portions of the population and by inte rnational tourists (e.g., Shrestha et al., 2002; Navrud and Mungatana, 1994). All of the survey respondents had completed the equivalent of a high school degree, and a large number had completed college (46.9%) or graduate school (18.7%) (Figure 4-2). This is a bit surprising given that the inco me distribution discussed earlier shows that visitors are spread throughout the socio-economic landscap e; thus, income strata for tax purposes and education are not directly correlated among th e sample respondents in this study. However, the high educational attainment of park visitors may reflect the high literacy rates of the Colombian population in general. The majority of survey respondents were ei ther single (45%) or married (45%). The remaining respondents (10%) were divorced, widowe d, or in a free union (i.e., a Colombian legal term for non-marriage civil unions). The age dist ribution of the adult resp ondents (i.e., at least 18 years of age) shows that a large percentage of the visitors were rela tively young; 62.5% were less than 40 years of age and 60 years was th e maximum age among the respondents (Figure 43). While it is possible that outdoor recreat ion in Colombia is enjoyed by most by younger people, LNNP is not a good example of the aver age Colombian National Park since there is a health hazard posed by the potential for hypoxia in high altitudes that can prevent young children and older people or people with card iovascular ailments from visiting. Awareness and Preferences for Envi ronmenta l Amenities and Services To better understand visitors preferences towa rd PAs in general and LNNP in particular, respondents were asked a series of questions about their visitation to other natural areas and their concern for their environment. Respondents we re also asked how much they knew about the wildfires that had taken place the previous y ear, as well as about their knowledge of the environmental and ecological services provided by the park in particular.
63 Overall, respondents were not too familia r with the 2006 wildfires. Although many respondents seemed to recall hearing about the pa rk in the news, most 76.5% did not know that a massive wildfire had occurred. Only 23.0% knew that a fire had taken place in the previous year. This was somewhat surprising, as the events were covered in the regional and national media. Only 6.3% of the respondents were familiar enough to be able to describe the fire in physical dimensions. Most respondents were knowledgeable of th e ecological services provided by the park. While most of the publicity conc erning the park emphasizes its ro le as a source of many rivers and one of the main hydrologic re gulators in the country, there was a higher awareness for the biodiversity conservation role of the park; 76.5% were aware of biodive rsity conservation and 70.3% were aware of hydrologic regulation (Figur e 4-4). Carbon sequestration awareness was the lowest among the three services that responden ts were asked about (68.7%), perhaps because most people associate large trees such as those in the Amazon w ith carbon sequestration rather than the small shrubs and grasses th at are present in LNNPs dominant paramo A large portion of respondents stated a relatively high level of concern about environmental issues in Colombia. More than two-thirds of survey re spondents (71.9%) have a concern of 8 or higher in a 0 to 10 scale where 0 represents no concern and 10 represents very concerned (Figure 4-5). While th ere were several lower responses it is important to note that there was only one response below a 5 or me dium level of concern, indicating that an overwhelming majority of visitors to LNNP are at least mildly concerned with the environment. It is likely that the warm glow phenomenon explains this result. Despite their reported concern for environmental issues, most survey respondents are not members or have not donated money to enviro nmental organizations (85.9%). Thus, this
64 variable might be a better reflection of environmental concern than asking respondents about their environmental co ncern directly. That said, paying memberships might not be a good indication either since although there are severa l international environmental NGOs active in Colombia, very few of them target Colomb ian citizens for membership or donations. Alternatively, a majority of re spondents visit national parks or other natural areas with some regularity (59.3%), perhaps indicating that frequency of visitation to natural areasand payment of a nominal entry feecould be a better indica tor of preferences for environmentally based non-market goods. Park Visitation To better understand th e patterns of visitation to L NNP, respondents were asked a series of questions about this visit and ot her visits to the park they had made in the past. Seventy three percent of respondents had never visited the park in the past and only 12% of respondents had visited the park in the previous year. The average number of annual visits to LNNP is 1.05, indicating that this is a site that is normally visited only once a year. The current visit to LNNP was the primary purpose of their trip for a majority of respondents (81.0%). A large proportion of respo ndents were visiting the park for only one day (87.5%). This is the common form of visitation to the park, given the lack of infrastructure and accommodations and the harsh temperatures and hypoxia. The average length of each visit in our sample is 1.21 days per visit; 3.1% visited fo r two days and 9.4% visited for three days, the maximum reported. The number of sampled visitors from each state is shown in Figure 4-6, where LNNP appears in red and only the names of the states wi th respondents are labeled. Given that the park routinely collects basic data on all visitors, it is possible to compare the or igins of visitors from the official visitation data with that from the sample of 64. Secondary data obtained from the
65 Parks Service and the Concession include detailed information on park visitors for October and November 2006, and January, February, March, April, and May of 2007, totaling 26,246 observations that include cities of origin from visitors in the 13 states modeled in this paper; these observations account for 94.5% of all record ed visits during that period even though the states included account for only 73.6% of the population of Columbia. Thus, the exclusion of the other states will not have a significant effect on the calculated visitation rate. Conversely, the exclusion of 5 of the 12 months is expected to downward bias the visitation rate by at least half given total visitation in recent years. Figure 4-7 shows a comparison between the zo nal visitation reported in the sample and that obtained from the secondary data. The graph shows that the vi sitor populat ion and the sample population contain the samerelatively la rge shareof visitors from the same four states. This similarity implies th at there is a high correlation between the proportion of visitors from each individual state in both the sample and the visiting populat ion. Given that the secondary data contains additional informati on on the visitor population, it is possible that additional analysis can be conducte d to further assess how representative the sample is of the population. Travel Cost Responses Respondents were initially queried about their group size, c ity of origin, and transportation m ethod used to reach the park. They were then as ked about their expenses in the current trip for transportation, lodging, equipment rental, and guidance costs, and all these values were aggregated to yield a total cost. Respondents were allowed to report their expenses either individually or by group according to their own pr eference. If the respondent had purchased an all-inclusive package, this was noted on the que stionnaire, and only the amount paid for the all inclusive package was reported. If the respondent had not purchased an all-inclusive package
66 and had chosen to report trip expenses as a group, the total cost reported was divided by the number of individuals in the group when the data was coded. If the respondent had chosen to report expenses individually, the total costs were not manipulate d. The mean travel costs per person (COP) from each of the 13 states with info rmation obtained from the surveyed visitors are reported in the first column of Table 4-1. Visitors to LNNP use a variety of trans portation methods, including small motorcycles, cars, tourist buses, and chartered buses. This characteristic of travel model creates high variability in travel costs within and among zone s as compared to similar analyses (e.g., Navrud and Mungatana, 1994; Mathieu et al., 2003; Fl eming and Cook, 2007), as evidenced by the relatively high standard devi ations (Table 4-1). The travel costs used in this study are biased downward since they exclude the opportunity cost of time. Future analysis can choose to augment the reported data with the average daily income associated with each stratum in each state, which would increase the consumer surplus estimates and the estimated value of the LNNP. Contingent Valuation Responses Respondents were read a statem ent that info rmed them of the fire events and the magnitude of the resulting devast ation to ecosystems in the park. The statement also informed them that the park management authorities were considering an increase in the existing entrance fee, and that the extra money would be used to rest ore the areas affected by the fires. They were then asked to consider their expenditures for this trip and for recreation in general, and were asked whether or not they would be willing to pay the amount shown in a card held by the researcher. The payment cards included fee increases of 25%, 50%, 75%, or 100% over the existing entrance fee of $8,000 COP. Overall, 81 % of respondents (81%) indicated they would be willing to pay that additional amount for the rest oration of the areas affected by the wildfires.
67 The responses by the level of fee increase are show n in Figure 4-8. As expected, the percentage of individuals that are willing to pay is reduced as the proposed fee is increased. Across all respondents, the average WTP for restoration was $3,969 COP ($1.98 USD). Several authors have emphasized the importa nce of information in the molding of preferences and the resulting WTP in contingent valuation experiments (e.g., Pearce and Turner, 1990; Smith, 1996; MacMillan and Duff, 1998; Ki ng and Mazzota, 2000; Loomis et al., 2000). To test whether or not the information given to respondents about the ecol ogical functions of the park influenced their WTP, two sets or versions of the survey were developed. In version 1, the portion of the questionna ire that asks respondents about th eir awareness of the ecological services provided LNNP was asked after the contingent valuation experiment. Conversely, in version 2, the portion of the questio nnaire regarding the ecological services provided by the park was asked before the contingent valuation experiment. Roughly the same number of version 1 (51.6%) and version 2 (48.4%) surveys were completed. In the follow-up question, respondents were as ked their maximum WTP for restoration as an open-ended question. Of the total respondent s, 89% expressed appositive WTP amount that can be used in a regression analysis. This highe r share of respondents in dicates that some who said no to the amount they were first presente d with were willing to pay a lower amount. Conversely, some who said yes, were willing to pay more. Results are summarized in Figure 49, which indicates that the WTP values ranged from 0 (those who declined th e first price increase and were not willing to pay anything in respons e to the follow-up, 12.5% of respondents) up to $30,000 COP ($15USD). Across all respondents, the average of the maximum reported WTP amounts was $6,638 COP ($3.32 USD), which represen ts a 67% increase in the WTP by using the open-ended question format following a closed-ended dichotomous choice format.
68 TCM Analysis and Results Recreational Demand Recall that the individual-based TC M involves data on the number of trips in the previous year and the cost per trip to calculate annual recreational demand. The survey data show a mean number of annual trips per respondent of 1.05, so there would be no tradeoff between trip cost and demand on an annual basis across individual s. Use of the zonal approach to capture visitation was, thus, necessary since this park is typically visited just once per year. With the zonal TCM approach, secondary vi sitation data (assumed to be a census of visitors with home location identified) is first used with populati on data to estimate vi sitation rates by zone. The political division of Colo mbia into mutually-exclusive states was used for the definition of distinct geographic zones. The latest census figures were used to obtain the population of each state included in the sample popul ation from the survey (13 in total) and the number of visitors from each state over a 7month period was calculated using the secondary data from the park; this information is summari zed by state in Table 4-1. Visitation rates were calculated using a zonal approach (Navrud a nd Mungatana, 1994; Fleming and Cook, 2007) with the following equation: Vj = (USERSj/POPj), (4-1) where Vj = visitation rate for zone j; USERSj = total number of users from zone j as obtained from the monthly visito r counts (N = 26,246); POPj = Population of zone j from the 2005 census. Using the reported visitor counts accounts for only a seven month period (and only for the 13 states represented by survey respondents) and, th us, generates a demand for only that period and those states covered in the study; the visitation rate (Vj) is an underestimate of the annual demand of all Columbians. It would be possible to extrapolate to an entire year using the reported 2006 overall visitation; it would be more difficult to extrap olate to the remaining states
69 since travel cost data are una vailable (although as mentioned earlier, the exclusion of the remaining states is less of an issue since the 13 states included account fo r nearly 95% of total visits). If travel costs (incl uding the opportunity cost of income ) are found to be correlated by distance, these states could be included in the future and/or the opportunity costs of time could be included directly. Economic theory suggests that the quantity purchased of any good is a function of its price. Given that the zonal TCM uses visitation ra tes as a proxy for quantity purchased and travel cost as a proxy for price, equation 3-5 pr oposed a linear demand specification. Prior to estimation, the mean travel costs were plotted ag ainst the visitation rates from each state (Figure 4-10). The scatter plot suggests a non-linear relationship between travel costs and visitation and suggests a logarithmic form may bette r fit the data. In order to iden tify the best fitting model, a series of models with different functional forms were estimated; notably the use of a logarithmic form as in previous studies. All the mode ls were estimated using a generalized likelihood method (GLM) as suggested by Bowes and Loom is (1980), thereby eliminating the natural heteroscedasticity inherent in the analysis of heterogeneous states. Th e results are summarized in Table 4-2. It should be noted that all models use the data at the state level from Table 4-1. This is because the relatively hi gh standard deviation of the travel costs within each state likely due to those individuals trav eling by tour bus precluded th e estimation of a model with acceptable explanatory value and st atistically signifi cant parameter estimates. Attempting to weight the observations by the variance estimati ons, as has been suggested, did not improve the models in this study due to a lack of observations. All four forms of the recreational demand estimation yield parameter estimates whose signs support economic theory. In all four cases the estimates for the intercept term 0 are
70 positive, indicating that the visitation rate will be positive when there are no travel costs. The demand curves estimated for models 1-4 are show n in Figures 11-14, respectively. Recall that the actual annual visitati on has ranged from more than 45,000 to nearly 80,000 visits for the last four years (Figure 2-2). The parameter estimates for the travel cost variation 1 are all negative, in dicating that as travel costs increase, fewer visits to LNNP are predicted. Even though all models follow the expectations of economic theorywhich indicat e that the recreationa l services provided by LNNP are a normal goodfunctional specification 3 e xhibits a better fit than the other three with the highest R-squared value. Model specifica tion 3 is selected as th e most accurate since it has the best fit and provides the most conserva tive estimate of recreational demand for LNNP and, thus, is used in subsequent analysis. Consumer Surplus Estimation The recreational benefits of an environm ental amenity can be estimated by quantifying the area under the demand curve and above the pr ice paid for purchase of the amenity, also known as the consumer surplus. Consumer surp lus is the willingness to pay over and above the price of the trip paid by the consumer (Maharan a et al., 2000). The zonal consumer surplus value can then be expressed mathematically as: (4-3) where CS is this the consumer surplus that is calculated from the summation of the consumer surplus from each zone j (i.e., the second summation that operates on the term in brackets). The consumer surplus from each zone is calcula ted in increments of $1,000 COP (i.e., $0.50 USD) beginning at the average total cost from each z one (i.e., ATCj as repor ted in Table 4-1) and increasing to the choke-off price (C), which is the total cost that produces the lowest demand
71 (i.e., zero). Above the choke-off price, which is different for each model, demand (visitation) is zero. In other words, the calculations measure the area above the average cost but below the demand using a step-wise approach. Note that th e calculations first predict the visitation rates (Vj) at each increment of travel costs using the four empirical models shown in Table 4-2, which are then multiplied by the population in each zone (POPj) to derive th e predicted number of visitors from zone j ( ). The estimates of consumer surplus for each model specification and each of the zones included in the analysis are shown in Table 4-3 in 2007 nominal values. The total consumer surplus accruing the citizens of the 13 states in cluded in the study due to the existence of LNNP ranges between a conservative 2.2 billion COP, th e equivalent of 1.1 million USD or about 3 cents per person, to a high estimate of 9.2 billion COP or 4.6 million USD or about 14 cents per person. It is possible that the unusually high esti mates from model specification 4 are the result of the existence of a horizontal asymptote at TC = 0. However, the consumer surplus estimation for model specification 4 was only carried out from TC = 1,000 COP ( 1,000 COP = 0.50 USD) as the starting point in order to reduce the pos sible overestimation produ ced by the presence of this mathematical pr oblem (Figure 4-14). CVM Analysis and Results Closed Ended: Dichotomous Choice (WTP) Model A linear m odel was specified for this problem in equation 3-8 and followed by a description of a probability m odel (equation 3-9) and resultin g log-likelihood function for a probit model (equation 3-10) to allow estimation of the probability that a respondent would be WTP the increased entry fee for restoration. This model is specified as follows: WTPi = 0 + 1 VERSIONi + 2 FIRSTVISi + 3 CHILDi + 4 PRIMPURPi + 5 FIREAWAREi + 6 PRICEi + 7 ECOAWAREi + 8 MEMBERi + 9 RVISITi +
72 10 AGEi + 11 INCOMEi + 12 GENDERi + 13 TOURBUSi + i (4-4) where WTPi takes on a value of 1 if the respondent an swered yes to paying a higher fee or a value of 0 if they were not willing to pay the ad ditional fee. The fee is one of four amounts and is included as an explanatory variable; PRICEi is the proposed increase fee that respondent i was asked to consider. There are a total of 11 binary e xplanatory variables in the model. Each of these variables takes on the value of one if the characteristic is present or zero if not. VERSIONi is a binary variable that takes a va lue of 1 if respondent i was asked about his/her knowledge of the ecological services provided by the park after the WTP question; FIRSTVISi is a binary variable that takes a value of 1 if th e current visit is respondent i s first visit to LNNP; CHILDi is a binary variable that takes a valu e of 1 if there was a child present in respondent i s group; PRIMPURPi is a binary variable that ta kes a value of 1 if respondent i s primary purpose of travel was the visit to LNNP; FIREAWAREi is a binary variable that ta kes a value of 1 if respondent i was somewhat or very aware of the fire events th at had taken place the previous year; ECOAWAREi is a binary variable that ta kes a value of 1 if respondent i was aware of the hydrological, biodiversity conservation, and carbon regul ation services provided by LNNP; MEMBERi is a binary variable that takes a value of 1 if respondent i was a member or donated money to an environmental organization; RVISITi is a binary variable that ta kes a value of 1 if respondent i visits national parks or natural ar eas in general with regularity; AGEi is a binary variable that takes a value of 1 if respondent i is older than 40; INCOMEi is a categorical variable that takes values between 1 and 6 according to the so cioeconomic strata in which respondent i s home is located; GENDERi is a binary variable that takes a value of 1 if respondent i is male; and
73 TOURBUSi is a binary variable that takes a value of 1 if respondent i traveled to the park in a chartered or other tour bus. Discussion of Explanatory Variables Respondents were asked how m any children they were traveling with. Given that some respondents were traveling in la rge groups (over 30 individuals), there was a high variability of responses. This created a problem in the estimati on of the relative importa nce of the presence of children for a respondents WTP for restoration. To solve this problem, the dummy variable CHILD was created to indicate the presence of at least one minor in the respondents group. The distribution of the age variable was also problematic as it was skewed toward younger respondents. To solve this problem, a dumm y variable was created to differentiate the young from older respondents. The threshold of 40 years of age was selected for two reasons. First, the oldest respondents in our study were in late fifties and ear ly sixties. Visits to LNNP can be dangerous for very old people and very young children, as the low oxygen conditions of high altitudes can exacerbate cardio-respiratory ailments, and park personnel ask older people and young children from abstaining to enter the pa rk. The age distributio n of respondents has a minimum of 19 and a maximum of 60, preventing the separation of young and old into adults and senior citizens. Second, th e threshold at age 40 then creates a potential differentiation between those who travel in mostly small groups of single and recently married individuals without children and those traveli ng in larger groups that contain one or several families. While this separation is somewhat arbitrary, it provided a solution to the problems inherent in the sample. The questionnaire also contai ned three possible answers to the question regarding awareness of the wildfires that had taken place the previous year. Theoretically, knowledge or awareness of the fires will influe nce a respondents WTP for restor ation of the areas affected by
74 those fires. Respondents were asked if they were very, somewhat or not familiar at all with the fire events. For analysis, the very and somewhat familiar responses were clumped into the dummy variable FIREAWARE, which differentiates whether the respondent had some knowledge of the fire events or none at all. A series of questions con cerning respondents awareness of the ecological services provided by the park were also asked. The re spondents were asked if they had previous knowledge of the hydrologic regulation services provided by the park, its importance for the conservation of specific endangere d and threatened speci es, and its role as in carbon storage and atmospheric carbon dioxide sequestration. The du mmy variable ECOAWARE was then created to differentiate respondents who had previous kn owledge of all three eco logical services from those who did not or who did not know of at least one of these services. The ordering of the contingent valuation exercise and the questions concerning the respondents awareness of the ecol ogical services provided by the park varied so as to determine the effect that the information given through the su rvey had on individuals WTP for restoration. Economic theory suggests that information helps mold peoples preferences, and the information given to respondents in the form of questions ma y well affect their WTP. The dummy variable VERSION indicated whether the CV exercise wa s carried out before the respondent was asked about his/her awareness of the ecolo gical services provided by LNNP. One of the most surprising things encounter ed during the data co llection period was the variability in transportation methods chosen by vi sitors to LNNP. The method of transportation inherently affects the group si ze of the respondent, as people tr aveling in tour buses will be traveling in larger groups than those traveling in privately owne d cars and motorcycles. The initial analysis included group size as an explanatory variable, but the large range of responses,
75 which included groups as small as a single individual and as large as an entire busload, created a problem for the analysis. The dummy variable TOURBUS was created to indicate whether the respondent was traveling in a chartered bus and hence was part of a large group. Summary statistics of all the variables used in the analysis are shown in Table 4-4. WTP Model Estimation Results The W TP responses were analyzed using two probit models in which the probability of responding positively to the dichotomous choice e xperiment was analyzed as a function of the independent variables identified ea rlier. Two models were estimat ed to account for two possible functional forms assumed by the PRICE variable. Given the small number of observations used and the relatively small variati on among prices (only four price le vels were used), there was no a priori information regarding the correct functio nal specification for these models. The first model assumed that the relationship between probability of answering positively to the experiment and proposed fee increase (PRICE) was linear, while the second assumed that the relationship is logarithmic, and the natural logarithm of PRICE wa s used instead. The results of the analysis are summarized in Ta bles 4-5 and 4-6, respectively. While not all the variables included in the analysis were statistically significant, there were several variables that were consistently st atistically significant in both models. Following economic theory, PRICE was statistically significant at a 5% level in bo th models, and both the parameter estimate and the marginal effect have a negative sign, implying that as the proposed entrance fee hike shown to surv ey respondents increased, they we re less likely to agree to pay the extra money for ecological restoration. The marginal effect shows the estimated change in the probability of responding pos itively to the WTP question w ith a 1 unit increase in the independent variable, starting from a correspondi ng value of the independent variable, shown in tables 4-5 and 4-6 as the column labeled x (Gujarati, 2004). Probit models are based on the
76 probability of a certain outcome, and changes in one independent variable will not only bring changes to the dependent variable but are also inherent ly accompanied by changes in the relative effect of the other independent variables. Marginal changes ar e therefore shown along with the base values with which they were calculated. The parameter estimate for the variable indicating whether the visit to LNNP was the primary purpose of the respondents trip, PRIMPURP was statistically signi ficant at a 10% level and had a positive sign, but its marginal effect wa s not statistically significant. This indicates that while respondents who indicate d that the visit to LNNP was the primary purpose of their trip have a higher probability of being WTP than those who did not, this effect is either weak or was weakened in the analysis due to th e small number of observations used. Contrary to expectations, those respo ndents who were knowledgeable about the ecological services provided by the park (ECOAW ARE) exhibited a lower probability of being WTP for ecosystem restoration, as indicated by negative and statistically significant parameter and marginal effect estimates at a 5% level. This seems counterintuitive as individuals more familiarized with the services provided by an eco system are expected to value that ecosystem more than those who are not. However, more knowledge regarding the se rvices provided by an ecosystem may also be an indication of more know ledge about this particular ecosystem or about ecology in general. Individuals familiarized with ecological science may have more understanding of the self-restor ation capacities of ecosystems and, hence, may exhibit a lower probability of being WTP than those who do not understand these processes since they expect nature to play its course and h eal itself. Also, those individuals more familiar with the services provided by LNNP may know more about LNNP and take into acc ount that the 2,500 hectares or so affected by the wildfires are a small area co mpared to the more than 58,000 hectares under
77 protection within the park, or they may know that fires are common in paramo ecosystems and are arguably a natural component of the disturbance regimes that maintain the in tegrity of the system. These more informed respondents may also be aware of the human dynamic of pastoralist fires and do not believe that a higher entrance fee would solve the problem, and may in fact believe that increased management activi ties could potentially resu lt in more conflict and more fires. Respondents who expressed that they are regular visitors to other National parks or other natural areas (RVISIT) exhibited a higher probabi lity of being WTP for restoration than those who did not, as the positive and statistically signi ficant parameter and marginal effect estimates indicate. Regular visits to natural areas may indicate environmentalist preferences, which can be expressed as willingness-to-pay for restoration of damaged ecosystems. Regular users of natural amenities also have more concrete direct-use valu es that stem from actual use and enjoyment of these amenities, as opposed to more esoteric passive-use, option, or bequest values held by individuals who do not use these amenities, and this is reflected in the WTP for restoration exhibited by regular users of environmental amenities. The negative and statistically significant pa rameter and marginal effect estimates associated with the GENDER variable show that men are less likely to respond positively to the WTP question. There is no obvious reason for the higher willingness by women to pay for restoration, and this issue deserves more attention. Survey respondents who were traveling in tour buses, most of whom were traveling in large groups, exhibited a higher probability of responding positively to the dichotomous choice question, as evidenced by the posi tive and statistically significant parameter and marginal effect estimates associated with the TOURBUS variable. It was initially expe cted that respondents
78 travelling in large groups would be less likely to be willing to pay an extra amount on top of the entrance fee, as they were told that the given amount would be char ged on top of the entrance fee of each of the members of the group, hence ma king it a large sum when the respondent was responsible for paying the surcharge for his entire family, for instance. However, tour travel may also reflect an individuals desi re to be part of a group that is interested and concerned with nature thereby resulting in a higher probability of being WTP. Open Ended CV: Maximum WTP Estimation Results A follow-up question to whether respondents w ould be W TP the fee proposed to them asked respondents directly to provide their maximum willingness to pay for restoration. The question was asked regardless of whether the res pondent had agreed to pay the amount shown in the card, but the format of the question changed to accommodate the previous response. If the response had been positive, then the question read: If your response was yes, what is the maximum amount you would be willing to pay for restoration of the areas affected by the wildfires? Conversely, if their response had been negative, th e question read: If your response was no, what is the maximum amount you would be willing to pay for restoration of the areas affected by the wildfires, if anything? The answers were given in Colombian peso s and coded as the MAX_WTP variable. They were analyzed using an Ordinary Least Sq uares regression in which the dependent variable was regressed against the same variables used in the dichotomous choice analysis, plus a variable to account for their initial res ponse. Four different models were regressed in order to accommodate different functional forms taken by the MAX_WTP and the PRICE variable. The models follow the following form: MAX_WTP = 0 + 1 VERSION + 2 WTP_YES + 3 FIRSTVIS + 4 CHILD + 5 PRIMPURP + 6 FIREAWARE + 7 PRICE + 8 ECOAWARE + 9 MEMBER + 10 RVISIT + 11 AGE + 12 INCOME + 13 GENDER + 14 TOURBUS (4-5)
79 Equation 4-5 shows the linea r specification used in model 1. In model 2 a semi-log specification was used in which the response variable was the natural logar ithm of MAX_WTP (Ln MAX_WTP) instead of a linear specification. In model 3 a linear-log specification was used where by the linear explanatory variable price, PRICE, was replaced with its natural logarithm (Ln PRICE). In model specification 4, a doubl e-log specification was used in which the dependent variable MAX_WTP and the explanator y variable PRICE were both included after taking the natural logs. The results of the analysis are summarized in Table 4-7. At a first glance it can be noted that the intercept terms in two of the models have positive signs while on the other two the sign is negative. This anomaly can be attributed to the use of the logarithmic expression as the response variable, as in the two models where the response variable is linear the intercept estimate is negative, while in the two models th at use the logarithmic expression as the response variable the intercept estimate is positive. The parameter estimate associated with the VERSION variable was positive and statistically significant in all f our models. This indicates that respondents who went through the contingent valuation exercise before they were asked about thei r awareness of the ecological services provided by the park exhibited a higher WTP. It was expected that respondents who were asked about their awareness of the ecol ogical services provided by the parkand thus informed or reminded about these servicesbe fore going through the contingent valuation exercise would exhibit a higher WTP, as they had more information and would value these services that they had just been informed a bout. The results, however, show the opposite, which are contrary to previous studies that have tested for information bias. Not only was information bias not a problem, but its possible that respond ents perceived that extra information about
80 ecological services as a blatant at tempt to increase their willingness to pay such that they resisted the notion of paying for what is commonly considered a public good. Both the PRICE and WTP_YES variables, which are included to account for the anchoring effect, were found to be positive a nd statistically significant in all models. Respondents who agreed to pay the amount s hown to them during the dichotomous choice exercise exhibited higher WTP valu es than those who did not. Highe r prices or amounts used in the closed ended exercise also helped elicit hi gher maximum WTP values, as respondents had a positive and non-zero baseline from which they were asked to go as high as they were willing. The variable indicati ng if the respondent was visiting th e park for the first time was associated with a positive and statistically signi ficant coefficient in all four models, indicating that repeat visitors experien ce diminishing marginal utility. Respondents who were traveling with children also exhibited a higher WTP than those who did not, as evidenced by the positive and statistically significant parameter estimate associated with the CHILD variable. This result could be an indicato r of bequest values. The coefficient associated with the MEMB ER variable was found to be positive and statistically significant in all of the models, indicating that individuals who are members or donate money to environmental organizations have a higher WTP for ecologi cal restoration than those who do not. This is not a surprising result as these individuals are currently expressing their non-market environmental pr eferences by another mechanism. Survey respondents younger than 40 were found to exhibit higher WTP values for restoration of the areas affected by the wildfires than older in dividuals, as evidenced by the negative and statistically significant parameter estimates associated with the AGE variable in all models. Environmentalism is a relatively recent phenomenon, especially in the third world, were
81 conventional wisdom places environmental concerns at the bottom of the list of policy priorities. Older individuals may also be more skeptical that the additional fees collected would be used for restoration of LNNP since that w ould require the funds to be give n back to the LNNP instead of retained and redistributed by the Ministry of the Environment. Income as proxied by the tax stratum of th eir home was also found to be positively related to maximum willingness to pay for restor ation, echoing the theoretical notion that higher incomes result in higher demand for goods and serv ices. Individuals with higher incomes have more disposable income that can be spent in r ecreation or ecological rest oration, for instance. The positive parameter estimates associated w ith INCOME do not reflect a new finding but rather reinforce the valid ity of economic theory and previous empirical results. Discussion of Empirical Results The per cap ita consumer surplus estimates asso ciated with reported visits over a 7-month period from October 2006 through May 2007 from the TCM analysis range from a low of 70 COP or 3 USD cents to a high of 293 COP or 14 USD cents, which are very small amounts. However, once aggregated over the entire population of the country these small individual amounts sum to a considerable total relative to park system expenditures. Recall that the UAESPNNs annual budget for 2007 (Figure 2-4) s hows that both investment in Colombias parks and reserves and operational costs for all parks and offices of the park service amount to 23.5 billion COP or about 12 million USD. The mo st conservative estimate of total consumer surplus, at 2.2 billion COP, is equivalent to 9.4% of the entire Colomb ian parks system budget. Since this figure was based on visi tation data from just 7 of 12 months (58.3%), it might seem reasonable to extrapolate the findings to an entire year (which is the traditional unit of measure) in which case the lowest value found was a pproximately 3.8 billion COP ($1.9 million USD).
82 Respondents were found to have, on average, a positive willingness to pay for restoration of the areas affected by the 2006 wildfires. The closed-ended dichotomous choice exercise received a positive response rate of 81%, indica ting that a wide majority of respondents were willing to incur the higher entrance fee that they were confronted with. The open-ended follow up exercise produced an even higher average WTP as several individuals that responded yes, provided even higher values and some that re sponded no provided lower amounts than they were originally confronted with (but higher than zero). The average willingness to pay for restoration elicited through the closed-ended dichotomous choice exercise was found to be 3,969 COP, or about $2 USD per person. The average WTP for restoration elicited through the open ended exercise was found to be 6,742 COP or about $3.50 USD. If these average values are representative of the visitor population and visitation in 2007 remained at 2006 levels, pa rk management authorities could have raised between 238 and 405 million pesos, or between $119,000 and $202,000 USD for ecological restoration in 2007. These figures indicate that eco logical restoration is valuable enough to the visiting public to be financed directly by visitors through hi gher entrance fees or some form of suggested donation mechanism. The overwhelming positive response shows that a wide majority of respondents are willing to pay for restoring the ecologi cal integrity of the park, and while the annual estimates of WTP may not be high enough to finance a one-time investment to restore 2,500 hectares of paramo, these sums could be used to finance an ongoing small-scale restoration program in LNNP.
83 Stratum 1 Stratum 2 Stratum 3 Stratum 4 Stratum 5 Stratum 6 Figure 4-1. Income of respondents by socio-economic strata
84 0 5 10 15 20 25 30 35 Less than ElementaryElementary SchoolSecondary School College Graduate/Professional SchoolNumber of Respondents Figure 4-2. Educational level of respondents Figure 4-3. Age distri bution of respondents
85 0 10 20 30 40 50 60 Hydrologic Regulation Biodiversity Conservation Carbon Sequestration Yes No Figure 4-4. Respondents awar eness of the ecological se rvices provided by LNNP 0 5 10 15 20 25 30 012345678910 (Not at all) (Somewhat) (Very)Number of Respondents Figure 4-5. Concern for environmental issues in Colombia among respondents
86 Figure 4-6. Location of LNNP (in red) and number of visitors in sample by state
87 0.00 5.00 10.00 15.00 20.00 25.00 30.00A m a z onas Antioquia Arauc a Atlntico B ol v ar Boyac C al das C aquet C a sanare C auca Cesar C hoc Crdoba Cundinam ar ca G uain a Guaviare Huila La G uaji r a M agdalena M et a Nario Norte de Sant an d er P ut um ayo Qui nd o Ri s ar al da S an Andr s y P r ovidencia S ant ander S ucr e Tolima Valle del Cauc a Vaups V i c hada Zonal Data Sample Figure 4-7. Distribution of visitation data and surv ey respondents by state 0 2 4 6 8 10 12 14 16 2,000 4,000 6,000 8,000Number of Visitors Yes No Figure 4-8. Dichotomous choice responses by card
88 Figure 4-9. Summary of WTP respons es from the open-ended inquiry $ 0 $ 50,000 $ 100,000 $ 150,000 $ 200,000 $ 250,000 $ 300,000 $ 350,000 $ 400,000 $ 450,000 $ 500,000 0 50010001500200025003000350040004500 VjMean TC Figure 4-10. Scatter plot a nd trend line of mean travel costs and visitation rates
89 $ 0 $ 50,000 $ 100,000 $ 150,000 $ 200,000 $ 250,000 $ 300,000 $ 350,000 $ 400,000 0 10,000 20,000 30,000 40,000 50,000 60,000 VisitsTravel Cost Figure 4-11. Predicted values for travel cost model (in COP) for specification 1 $ 0 $ 50,000 $ 100,000 $ 150,000 $ 200,000 $ 250,000 $ 300,000 $ 350,000 $ 400,000 $ 450,000 $ 500,000 020,00040,00060,00080,000100,000120,000140,000 VisitsTravel Cost Figure 4-12. Predicted values for travel cost model (in COP) for specification 2
90 $ 0 $ 200,000 $ 400,000 $ 600,000 $ 800,000 $ 1,000,000 $ 1,200,000 $ 1,400,000 $ 1,600,000 05,00010,00015,00020,00025,00030,00035,00040,00045,00050,000 VisitsTravel Cost Figure 4-13. Predicted values for travel cost model (in COP) for specification 3 $ 1,000 $ 10,000 $ 100,000 $ 1,000,000 $ 10,000,000 $ 100,000,000 $ 1,000,000,000 $ 10,000,000,000 $ 100,000,000,000 $ 1,000,000,000,000 $ 10,000,000,000,000 0100,000200,000300,000400,000500,000600,000700,000800,000900,000 VisitsTravel Cost Figure 4-14. Predicted values for travel cost model (in COP) for specification 4
91 Table 4-1. Per capita travel cost inform ation and 7-month visitation rates by state Travel Cost Total Population, Visitation Mean Std. Visits, POPj per mil. State (1,000 COP) Dev. USERSj (1,000) (Vj) Antioquia $119.9 $60.2 6,496 5,682 1,143.2 Atlntico $162.1 $0.0 418 2,166 193.0 Boyac $287.2 $0.0 251 1,255 199.9 Caldas $39.0 $27.7 3,906 969 4,032.0 Cesar $81.2 $0.0 100 903 110.7 Cundinamarca $209.2 $125.9 6,585 9,120 722.0 Huila $98.0 $0.0 263 1,011 260.0 Nario $430.9 $0.0 71 1,542 46.0 Quindo $50.6 $11.7 999 535 1,868.8 Risaralda $55.6 $25.7 1,500 898 1,671.3 Santander $227.7 $229.6 682 1,958 348.3 Tolima $20.0 $0.0 705 1,365 516.3 Valle del Cauca $157.3 $115.2 4,270 4,161 1,026.1 Sources: The travel cost information is from the survey, the visitation data are from the selected months between October 2006 and May 2007 from th e park service, and the population data are from the 2005 census. Table 4-2. Travel cost model results Intercept Slope Functional Form Est. of 0 p-value Est. of 1 p-value R2 1) Vj = 0 + 1(TC) 1646.571*** 0.0043 -0.005* 0.0779 0.256 2) Vj = 0 + 1(Ln TC) 8671.510** 0.0417 -666.996* 0.0637 0.279 3) Ln Vj = 0 + 1(TC) 7.269*** <.0001 -7.18E-5** 0.0133 0.441 4) Ln Vj = 0 + 1(Ln TC) 15.989*** 0.0025 -0.844** 0.0361 0.341 Notes: Triple, double and single asterisks indicate statistical significance at the 0.01, 0.05, and 0.10 level, respectively.
92 Table 4-3. Consumer surplus estimates by state, in thousands COP and USD State Model 1 Model 2 Model 3 Model 4 Columbian Pesos: Antioquia $687,045 $632,052 $481,219 $1,738,630 Atlntico $171,332 $169,611 $134,707 $625,650 Boyac $10,060 $26,423 $32,039 $320,053 Caldas $216,696 $200,466 $146,783 $351,178 Cesar $150,378 $135,585 $101,225 $295,225 Cundinamarca $402,645 $469,938 $407,589 $2,501,745 Huila $147,386 $133,400 $100,316 $320,498 Nario $0 $207 $14,092 $352,915 Quindo $111,130 $101,554 $74,844 $187,532 Risaralda $180,310 $164,093 $121,227 $310,729 Santander $65,054 $84,164 $76,885 $527,349 Tolima $344,530 $330,020 $237,122 $535,160 Valle del Cauca $351,220 $342,846 $270,182 $1,210,861 Total (COP) $2,837,787 $2,790,356 $2,198,230 $9,277,525 U.S. Dollars: Antioquia $344 $316 $241 $869 Atlntico $86 $85 $67 $313 Boyac $5 $13 $16 $160 Caldas $108 $100 $73 $176 Cesar $75 $68 $51 $148 Cundinamarca $201 $235 $204 $1,251 Huila $74 $67 $50 $160 Nario $0 $0.10 $7 $176 Quindo $56 $51 $37 $94 Risaralda $90 $82 $61 $155 Santander $33 $42 $38 $264 Tolima $172 $165 $119 $268 Valle del Cauca $176 $171 $135 $605 Total (USD) $1,419 $1,395 $1,099 $4,639 Note: The estimation was conducted using COP then values were converted to USD assuming 1 USD = 2,000 COP. The exchange rate ranged from 1,950 2,080 COP for 1 USD during the time of the study but 2,000 was used to simplify the conversions.
93 Table 4-4. Summary statistics of variables in the con tingent valuation analysis Variable N Mean Std. Dev. Min. Max. WTP_YES 64 0.813 0.393 0 1 MAX_WTP (in 1,000) 64 6.742 5.080 0 30.0 Ln (MAX_WTP) 57 8.739 0.677 6.2 10.3 VERSION 64 0.516 0.504 0 1 FIRSTVIS 64 0.703 0.460 0 1 CHILD 64 0.469 0.503 0 1 PRIMPURP 64 0.813 0.393 0 1 FIREAWARE 64 0.234 0.427 0 1 PRICE (in 1,000) 64 0.503 0.225 2.0 8.0 Ln (PRICE) 64 8.402 0.523 7.6 9.0 ECOAWARE 64 0.531 0.503 0 1 MEMBER 64 0.141 0.350 0 1 RVISIT 64 0.594 0.495 0 1 AGE 64 0.375 0.488 0 1 INCOME 64 3.578 1.166 1 6 GENDER 64 0.625 0.488 0 1 TOURBUS 64 0.281 0.453 0 1 Table 4-5. Results of the dichotom ous choice CV experiment, model 1 Parameter Marginal Effect Variable Estimate p-value (dy/dx) p-value x INTERCEPT 0.290 0.810 VERSION 0.469 0.225 0.0719 0.283 0.516 FIRSTVIS 0.581 0.321 0.1039 0.403 0.703 CHILD -0.373 0.413 -0.0579 0.413 0.469 PRIMPURP 0.999* 0.062 0.223 0.122 0.813 FIREAWARE 0.693 0.259 0.081 0.153 0.234 PRICE (1,000) -2.397** 0.048 -0.036** 0.030 5.031 ECOAWARE -1.211** 0.028 -0.183** 0.029 0.531 MEMBER 0.711 0.331 0.075 0.179 0.141 RVISIT 1.225** 0.021 0.222** 0.019 0.594 AGE 0.752 0.281 0.101 0.167 0.375 INCOME 0.218 0.213 0.033 0.295 3.578 GENDER -0.972 0.104 -0.128* 0.083 0.625 TOURBUS 0.804* 0.082 0.097 0.098 0.281 Notes: Double and single asterisk s indicate statistical significa nce at the 0.05 and 0.10 level, respectively.
94 Table 4-6. Results of the dichotom ous choice CV experiment, model 2 Parameter Marginal Effect Variable Estimate p-value (dy/dx) p-value x INTERCEPT 7.497 0.081 VERSION 0.4807 0.210 0.075 0.265 0.516 FIRSTVIS 0.560 0.329 0.100 0.408 0.703 CHILD -0.305 0.501 -0.047 0.499 0.469 PRIMPURP 0.973* 0.068 0.218 0.132 0.813 FIREAWARE 0.687 0.261 0.083 0.16 0.234 LN(PRICE) -0.994* 0.057 -0.152** 0.042 8.402 ECOAWARE -1.165** 0.032 -0.1785** 0.031 0.531 MEMBER 0.628 0.388 0.070 0.240 0.141 RVISIT 1.213** 0.016 0.223** 0.021 0.594 AGE 0.687 0.302 0.095 0.208 0.375 INCOME 0.208 0.246 0.032 0.312 3.578 GENDER -1.008* 0.083 -0.135* 0.074 0.625 TOURBUS 0.825* 0.086 0.100* 0.086 0.281 Notes: Double and single asterisk s indicate statistical significa nce at the 0.05 and 0.10 level, respectively.
95 Table 4-7. Results of the open-ended willingness to pay models Model 1 Model 2 Model 3 Model 4 Variable Parameter Est. (p value) Parameter Est. (p value) Parameter Est. (p value) Parameter Est. (p value) INTERCEPT -8,748.12** 5.872*** -30,849.00*** 1.262 (0.021) (<.0001) (0.021) (0.272) VERSION 2,708.89** 0.460*** 2,627.55** 0.445*** (0.019) (0.001) (0.003) (0.001) WTP_YES 4,101.87** 1.168*** 4,060.66* 1.163*** (0.047) (<.0001) (0.025) (<.0001) FIRSTVIS 2,545.53* 0.250* 2,637.31** 0.271* (0.052) (0.085) (0.054) (0.075) CHILD 2,202.21* 0.227* 2,058.95* 0.200 (0.056) (0.076) (0.077) (0.129) PRIMPURP 737.32 0.072 808.40 0.088 (0.64) (0.068) (0.615) (0.630) FIREAWARE -650.04 -0.044 -541.91 -0.022 (0.605) (0.755) (0.672) (0.878) PRICE 0.79*** 0.0002*** 3,072.39*** 0.637*** (0.003) (<.0001) (0.007) (<.0001) ECOAWARE 709.31 0.029 558.75 0.0005 (0.588) (0.840) (0.675) (0.997) MEMBER 5,203.03*** 0.494** 5,341.50*** 0.528*** (0.003) (0.011) (0.003) (0.010) RVISIT -659.05 -0.233* -683.23 -0.235* (0.581) (0.085) (0.576) (0.096) AGE -4,295.87*** -0.513*** -4,101.63*** -0.478*** (0.002) (0.001) (0.004) (0.003) INCOME 1,344.82** 0.182*** 1,386.17*** 0.189*** (0.010) (0.002) (0.009) (0.002) GENDER 58.67 0.070 211.81 0.101 (0.960) (0.589) (0.858) (0.454) TOURBUS -1.56 -0.055 -21.63 -0.061 (0.999) (0.687) (0.986) (0.670) Model Statistics: R2 0.52 0.71 0.50 0.68 Adj. R2 0.36 0.61 0.33 0.57 Notes: Triple, double and single asterisks indicate statistical significance at the 0.01, 0.05, and 0.10 level, respectively. Model 3 uses LN(PRICE).
96 CHAPTER 5 CONCLUSION Implications The results of this study show that conservation is not only s ocially valuable but that it has the potential to pay for itself. Our estimates of consumer surplus from the recreational use of LNNP by Colombian nationals are high relative to the operational costs of the entire network of PAs financed by the central government. The exis tence of LNNP as a vi able land use is then fully justified as its recreational benefits outweigh the costs. The high ratio of recreational benefits to operational costs also suggests the possibility that expansi on of the protected area network in general and LNNP in particular can be a sound social investment. The evidence also suggests that recreational us ers are willing to pay for the restoration of areas affected by the 2006 wildfires in LNNP. While the annual sums raised may not cover a one-time investment, some form of ongoing suggested donation mechanism may fund a continuing effort to maintain and improve the eco logical integrity of LNNP. Heavy involvement of local communities in such a program may also bring about improvements in other problematic areas, such as livestock grazing and pastoralist fire s within the borders of the park. Expansion of the parks existing outreach pr ograms, involvement of local schools, and co-management of grasslands and grazing areas are some possible ways to involve local communities in park management in a potentially beneficial manner. Furthermore, the data and analysis compile d in this study provide information to a country where little or no information regarding the economic value of recreation in PAs exists. Most problems of PAs in genera l are in one way or the other th e result of undervaluation of parks and reserves as national and social assets which in turn results in alternate land uses being valued more highly and opening the way for environmen tally destructive practi ces. However, these
97 results suggest that conservation and recreational use of PAs are a high-value land use, but that the conservation value of land is not always as ev ident as it can be for more intensive uses of land. New and creative mechanisms that allo w the capturing of the values of PAs will undoubtedly result in decreas ed park-people conflicts. The statutes that created LNNP and the UAESPNN mandate, among other objectives, the conservation and administration of the natural values of Colombia as well as the investigation of the values of the natural renewable resources of the nation. The information gap filled in part by this study is not only an important academic exercise; it is the first study to my knowledge that fulfills two of the objectives delineated by the Co lombian legislature at the time of the creation of the nations parks and reserves. There are a total of 53 PAs under the UAESPNNs authority, most of which are larger than LNNP. If the consumer surplus of each of Colombias protected areas was at least half the lowest estimate in this analysis, the unpaid recr eational benefits accruing to Colombians from the existence of the nations protected area system w ould be about 57 billion COP, more than twice the amount currently invested to protect the countrys natural heritage in the form of parks and reserves. There is, therefore, evidence to sugge st that Colombias protected area system is under-funded and that Colombians would benefit by increases in public expenditures on parks and reserves toward the improvement of infrastr ucture and services targeted at recreational visitors, as well as the creation of new protected. Policy Prescription The economic relevance of PAs in general and of LNNP in pa rticular was highlighted through the use of well-accepted economic valuati on methods, and it seems that environmental conservation in Colombia is a socially desirable outcome. The high ratio of recreational benefits to costs for LNNP suggest that budgetary increases for the UAESPNN toward improved
98 management are justified. Expansion of the park and improvements in park quality may also be desirable social investments. In regards to restoration of the areas affected by the 2006 wi ldfires, the existence of a positive and significant WTP of users indicates that a decentralization of UAESPNN finances may help park managers raise their own money for restoration. Increases in the entrance fee or some other self-funding mechanism would allow the establishment of an ongoing restoration or improved management program that should also heavily involve local communities. Caveats and Limitations of the Analysis The analysis conducted in this study is entirely reliant on the integrity of the sam ple as representative of the population. While great efforts were made to maintain such integrity, budgetary and time constraints prohi bited a truly randomized sampli ng of visitors to LNNP and the implementation of a thorough pre-testing prot ocol. The seasonality inherent in tourism activities also implies that the visiting population may be complete ly different at periods other than July and August, which was when the fiel dwork for this study was conducted. Most notable was the exclusion of international tourists, which are the traditiona l focus of non-market valuation estimation of protected areas in developing countries. The secondary visitation data used to construc t the visitation rates th at were integral to the TCM analysis were also limited in that th ey excluded several peak periods of visitation, namely during the summer months. And, of course the exclusion of the opportunity cost of time results in consumer surplus estimates that are biased downward. Uncertainties regarding model specification and estimation may also be a limiting factor in the interpretation of this study s analysis. As is shown by the dramatically different estimates of consumer surplus under different functional sp ecifications, our estimate s should be considered ball-park estimates and be used as such. While it can be argued that there is an important and
99 significant economic value of recreation in L NNP with some certainty, the magnitudes of consumer surplus would be very likely to change if data collected in diffe rent periods were used. Future Directions Non-Market Valuation in Lo s Nevados National Park This study m ay serve as a stepping stone towards a more comprehensive valuation study in LNNP. The datasets used contain the na mes and addresses of over 27,000 visitors, which could be used to conduct a comprehensive mail survey. Given that both the UAESPNN and the Eco-tourism Concession continually collect data on visitors, collabo ration with them could prove to be extremely productive in procuring visi tation information acro ss a continuous 12-month period to enable the derivati on of annual totals directly. A more comprehensive duplicate of this study where data are collected at several times throughout the year and a larger sample is used may also provide a better overview of recreational demand at LNNP. But short of a ne w and improved study, the further investigation of average income by stratum and by state could prove fruitful in including the opportunity cost of time into the travel cost estimates. While not directly related to non-market valuat ion, LNNP is a potential research site for a wide variety of projects. Since it is located close to several ma jor urban areas, field research can be conducted easily. A combination of high tour ism, unique biodiversity, valuable hydrologic and other ecological functions, and classic common pool resour ce problems provide for an excellent site for applied research in economics and other disciplines. Non-Market Valuation in Developing Country Protected Areas Econom ic valuation of PAs in general and of parks and reserves in the developing world in particular is an area where lit tle research exists but much more is needed. As is the case with LNNP, most parks in the developed and devel oping world alike are highly valuable but are
100 extremely undervalued under the existing institu tional framework, opening the way for other land uses. Research on the total economic value of these parks and reserves can fill an important gap and allow decision-makers and the general public to make mo re informed choices regarding these important social assets. It is also important to note that several hypotheses regarding the effect on willingness to pay estimates related to questionnaire design were supported or disputed (e.g., the anchoring effect and information bias, respectively), which rais es several interesting theoretical questions regarding the behavior of residents in de veloping countries.
101 APPENDIX Interview Script Valuation of Los Nevados National Park Version 1 Card A / Services After Informed Consent My name is Sergio Alvarez and I am a graduate student at the University of Florida. The purpose of this study is to evaluate the benefits cr eated by recreation in th e park. The answers you provide will help us better understand the econom ic benefits created by the park, and will be shared with the managers of the park so that they can better se rve you in the future. The study is being carried out by the Tropical Conservation and Development Program, the Department of Food and Resource Economics of the University of Florida in the United States, and the Colombian National Parks Special Management Unit. You will be asked about your current and past visits to the park. Th e duration of the interview is about 10 minutes. The survey is anonymous in that you will not be asked to provide your name or any other type of identification. There are no risks for pa rticipating in this study. Your partic ipation in this study is completely voluntary. There is no penalty for not participatin g. You have the right to withdraw from the study at anytime without consequence. For questions about your rights as a research participant, contact the IRB office at 352-392-0433 or irb2@uf l.edu. For questions regarding this study, you can contact me at 352-870-7604 or firstname.lastname@example.org, or my supervisor (Dr. Larkin) at 352-3921845 ext. 431 or email@example.com. Will you participate in the study? Yes No Introduction Thank you for participating in the Valuation of Los Nevados National Park survey. Date:_________________, Time: _____________________ Visitante de: Salida Entrada
102 Part 1. Visitation to Los Nevados National Park 1) Is this your first visit to Los Nevados National Park? Yes No 2) If No, how many times did you visit the Park in the past 12 months? _____ trips 3) And, about how many days total did you spe nd in the park during those trips? ____ days 4) When will you be leaving this Park? _____________________________ (date) 5) Is this Park the primary purpose of your trip? Yes No Part 2 Travel Expenses to Visit the Park The following questions pertain to this spec ific visit to Los Nevados National Park: 6) How many other people, if any, are you currently trav eling with? ______ people 7) If you are traveling with others, how ma ny including you are employed full time? ______ 8) How many are under 18 years of age? ________ children 9) Where are you traveling from? ___________________________________ 10) During this trip, how many nights wi ll you spend away from home? _______ nights 11) What transportation method(s) have you used to come to the park? Explain thoroughly. ______________________________________________________________________________ ______________________________________________________________________________ ______________________________________________________________________________ The next four questions ask about expenses fo r your group on this trip for transportation, lodging, equipment rental and guide fees Approximately, how much was spent on. 12) Transportation? ____________ Columbian pe sos per: person group 13) Lodging? ____________ Columbian pesos per: person group 14) Equipment rental? ____________ Columbian pe sos per: person group 15) Guide fees? ____________ Columbian pesos per: person group
103 Part 3 Value of Park Restoration 16) How familiar are you, if at all, of there be ing recent wildfires at Los Nevados National Park? not at all familiar, somewhat familiar, or very familiar In July 2006 several wildfires sw ept through this Park and burned about 7,000 hectares of forest and paramo and affected the ecosystems. Park au thorities are consideri ng increasing the entrance fee to finance a full restoration of the area affected, which would completely restore the integrity of the ecosystems. I am going to show you a card with the proposed fee increase. Please consider how much you spend on recreation each year and for this trip. Would you have been willing to pay this extra amount per person during this visit? 17) Card A: 2,000 pesos yes no dont know refuse 18) If your answer is NO, what is the high est amount you would be willing to pay as an increase of the existing park fee on this visit? _________________pesos 19) If your answer is YES, what is the highe st amount you would be willing to pay as an increase of the existing park fee on this visit? _________________pesos Part 4 Environmental Services The next three questions ask about your awareness of the environm ental services provided by the park. 20) Are you aware that this Park is the source of several rivers in the states of Caldas, Risaralda, Quindio, and Tolima, and supplie s drinking water for more than 2 million people? Yes No 21) Are you aware that this Park offers refuge for several rare and endangered species of plants and animals, including Andean bears, wax palms, and condors, among many others? Yes No 22) Are you aware that the trees in this Park he lp to offset the gree nhouse effect and thereby improve the environment by absorbing carbon dioxide? Yes No
104 Part 5 Demographics 23) In a scale from 0 to 10 where 0 is not at all, 5 is somewhat and 10 is very, how concerned are you about environm ental issues in Colombia? 0 1 2 3 4 5 6 7 8 9 10 24) Are you a member of, or do you donate money to, any environmental or nature-related organizations? Yes No 25) Do you visit National Parks or other natural areas regularly? Yes No 26) What year were you born in? ___________ 27) What is your marital status? single married divorced/separated widowed 28) What is the highest level of education you have completed? _____ Less than elementary school _____ Elementary School _____ Secondary School _____ College Degree _____ Graduate or Professional Degree 29) From the card that Im showing you, please select your annual household income? A B C D E F That is all. Thank you for your time. Have a good stay. 30) Gender of respondent: male female Time completed: ________________
105 LIST OF REFERENCES Barzetti, V., 1993. Parks and Progress: Protect ed Areas and Econom ic Development in Latin America and the Caribbean. IUCN/Int er-American Development Bank, Gland, Switzerland and Washington, D.C., MD. Bowes, M.D., Loomis, J.B., 1980. A note on the use of travel cost models with unequal zonal populations. Land Economics 56, 465-470. Carey, C., Dudley, N., Solton, S., 2000. Squandering Paradise? The Importance and Vulnerability of the Worlds Protecte d Areas. World Wide Fund for Nature International, Gland, Switzerland. Carranza, N., 2005. Colombia, otra vez fumigan lo s Parques Nacionales con Glifosato. Online. Available at http://www.ecopor tal.net/content/view/full/53254 Clawson, M., Knetsch, J.L., 1966. Economics of Ou tdoor Recreation. Johns Hopkins University Press, Baltimore, MD. CORPOCALDAS, undated. Proyecto Conservacion y Proteccion del Condor Andino en el Parque Nacional Natural Los Nevados. Online. Available at http://www.corpocaldas.gov.co/ad min/files/anterior/condor.htm Costanza, R., dArge, R., de Groot, R., Farber S., Grasso, M., Hannon, B., Naeem, S., Limburg, K., Paruelo, J., ONeill, R.V., Raskin, R., Su tton, P., van den Belt, M., 1997. The value of the worlds ecosystem services a nd natural capital. Nature 387, 253-260. Eagles, P.F., McCool, S.F., Haynes, C.D., 2002. Sustainable Tourism in Protected Areas: Guidelines for Planning and Management. IUCN Best Practice Protected Areas Guidelines, Gland, Switzerland, and Cambridge, UK. Farber, S.C., Costanza, R., Wilson, M.A., 2002. Economic and ecological concepts for valuing ecosystem services. Ecological Economics 41, 375-392. Fleming, C.M., Cook, A., 2007. The recreational value of Lake McKenzie: an application of the travel cost method. Paper presented at the 51st Annual Conference of the Australian Agricultural and Resource Economic s Society, Queenstown, New Zealand. Franco, P., Saavedra-Rodriguez, C.A., Kattan, G.H., 2007. Bird species diversity captured by protected areas in the Andes of Colomb ia: a gap analysis. Oryx 41 (1), 57-63. Fundacion Gabriel Piedrahita Uribe, undated. C onservacion in Situ. On line. Available at http://www.eduteka.org/pdf dir/Biodiversidad07E.pdf Gujarati, D.N., 2004. Basic Econometrics. Tata McGraw-Hill, New Delhi, India.
106 Horton, B., Colarullo, G., Bateman, I.J., Peres, C.A., 2003. Evaluating non-user willingness to pay for a large scale conservation programme in Amazonia: a UK/Italian contingent valuation study. Environmental Conservation 30 (2), 139-146. Jabarin, A.S., Damhoureyeh, S.A., 2006. Estimating the recreational benefits of Dibeen national park in Jordan using contingent valuation and travel cost methods. Pakistan Journal of Biological Sciences 9 (12), 2198-2206. Kiker, C.F., Lynne, G.D., 1997. Wetland Values and Valuing Wetlands. In: Coultas, C.L., Hsieh, Y.P. (Eds.), Ecology and Management of Tidal Marshes: A Model From the Gulf of Mexico. St. Lucie Press, Delray Beach, FL. King, D., Mazzota, M., 2000. Ecosystem Valua tion. USDA Natural Resources Conservation Service and National Oceanogr aphic and Atmospheric Administration. Online. Available at http://www.ecosystemvaluation.org. Kling, C.L., 1997. The gains from combining travel cost and contingent valuation data to value nonmarket goods. Land Economics 73 (3), 428-439. Laurance, S.G.W., Stouffer, P.C., Laurance, W.F ., 2004. Effects of road clearings on movement patterns of understory rainfore st birds in central Amazonia. Conservation Biology 18 (4), 1099-1109. Loomis, J., Kent, P., Strange, L., Fausch, K., Covich, A., 2000. Measur ing the total economic value of restoring ecosystem services in an im paired river basin: results from a contingent valuation survey. Ecological Economics 33, 103-117. MacMillan, D.C., Duff, E.I., 1998. Estimating th e non-market costs and benefits of native woodland restoration using the contingent valuation method. Forestry 71, 247-259. Maharana, I., Rai, S.C., Sharma, E., 2000. Valuing ecotourism in a sacred lake of the Sikkim Himalaya, India. Environmental Conservation 27 (3), 269-277. Mathieu, L.F., Langford, I.H., Kenyon, W., 2003. Valuing marine parks in a developing country: a case study of the Seychelles. Environment and Development Economics 8, 373-390. McNeely, J.A., 1995. Expanding Part nerships in Conservation. Island Press, Washington, D.C., MD. Mittermeier, R.A., Da Fonseca, G.A.B., Rylands A.B., Brandon, K., 2005. A brief history of biodiversity conservation in Brazil. Conservation Biology 19 (3), 601-607. Moeltner, K., 2003. Addressing aggregation bias in zonal recreation models. Journal of Environmental Economics and Management 45 (1), 128-144. Mulder, M.B., Coppolillo, P., 2005. Conservati on: Linking Ecology, Economics, and Culture. Princeton University Press, Princeton, NJ.
107 Mulongoy, K.J., Chape, S., 2004. Protected Area s and Biodiversity: An Overview of Key Issues. CBD Secretariat and UNEP-WCMC Montreal, Canada, and Cambridge, UK. Murphree, M.W., 1998. Enhancing sustainable use: incentives, pol itics and science. Rudy Grah Lecture on Forestry and Sustainable Developm ent, University of California, Berkeley. Murphree, M.W., 2004. Co mmunal approaches to natural res ource management in Africa: from whence and to where. The 2004 Breslauer Graduate Student Sy mposium, Berkeley, California. Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conserva tion priorities. Nature 403, 853-858. Navrud, S., Mungatana, E.D., 1994. Environmenta l valuation in developing countries: the recreational value of wildlife view ing. Ecological Economics 11, 135-151. Ojeda, M.I., Mayer, A.S., Solomon, B.D., 2008. Economic valuation of environmental services sustained by water flows in the Yaqui river delta. Ecological Economics 65, 155-166. Pearce, D.W., Turner, R.K., 1990. Economics of Natural Resources and the Environment. Harvester Wheatsheaf, Hetfordshire, UK. Peres, C.A., 2001. Synergistic effects of s ubsistence hunting and habitat fragmentation on Amazonian forest vertebrates. Conservation Biology 15 (6), 1490-1505. Reed, T., 2002. The function of protected area au thorities: considerations for financial and organizational management. World Bank Summer Internship Program 1999. Runte, A., 1987. National Parks: The American E xperience. University of Nebraska Press, Lincoln, NE. Schwartzman, S., Moreira, A., Nepstad, D., 2000. Re thinking tropical forest conservation: perils in parks. Conservation Biology 14 (5), 1351-1357. Shah, A., 1995. The Economics of Third Worl d National Parks: Issues of Tourism and Environmental Management. Edward El gar Publishing, Aldershot, Hants, UK. Sharp, B., Kerr, G., 2005. Option and Existenc e Values for the Waitaki Catchment. New Zealand Ministry for the Envi ronment, Wellington, New Zealand. Shrestha, R.K., Seidl, A.F., Moraes, A.S., 2002. Value of recreational fishing in the Brazilian Pantanal: a travel cost analysis using c ount data models. Ecological Economics 42, 289299. Smith V.K., 1996. Estimating Economic Values for Nature: Methods for Non-Market Valuation. Elgar Publishing, Aldershot, Hants, UK.
108 Stoll, J.R., 1983. Recreational act ivities and nonmarket valuation: the conceptualization issue. Southern Journal of Agricultural Economics December, 119-125. Tangley, L., 1988. Beyond national parks. Bioscience 38 (3), 146-147. Terborgh, J., 2000. The fate of tropical forests: a matter of stewardship. Conservation Biology 14 (5), 1358-1361. Van Der Hammen, T., Perez Preciado, A., Pinto, P. (Eds.), 1983. Estudios de Ecosistemas Tropandinos: La Cordillera Central Colombiana, Transecto Parque Los Nevados. J. Cramer, Germany. Van Kooten, G.C., Bulte, E., 2000. The Economic s of Nature: Managing Biological Assets. Blackwell, Malden, MA. Waller, D., 2001. Some shaky figures on ANWR drill ing. Time Magazine Online. Available at http://www.time.com/time/column ist/waller/article/0,9565,170983,00.html Whittington, D., 1998. Administering contingent valuation surveys in developing countries. World Development 26 (1), 21-30. Wilshusen, P.R., Brechin, S.R., Fortwangler, C. L., West P.C., 2002. Reinventing the square wheel: critique of a resurgent protection paradigm in international biodiversity conservation. Society and Natural Resources, 15 17-40. Wyngaarden, W.V., Fandino-Lozano, M., 2002. Pa rque Nacional Los Nevados: Un Caso de Seleccion y Zonificacion de Areas de Cons ervacion Biologica. IDEADE-DET. Bogota, Colombia.
109 BIOGRAPHICAL SKETCH Sergio Alvarez was born and raised in Maniza les, Colom bia. In 2001 he moved to South Florida to pursue studies in Environmental Sc ience at Broward Community College. Sergio arrived at Gainesville in 2004 to enroll in the Environmental Sc ience program at the School of Natural Resources and Environment at the University of Florida. In 2006 he joined the Food and Resource Economics Department at UF as a Ma sters Student. Sergio will continue his academic career in Food and Resource Economics at the University of Florida, where he will pursue a doctoral degree.