0862-A2

Quasi Option Value of Biodiversity of a Tropical Forest: Athirappally Forests in Kerala State, India

T.S. Amjath Babu^[1] and S. Suryaprakash

Abstract

Biodiversity of tropical forests is under threat because of increasing human population and rapid developmental activities. The tropical forests in the Western Ghats of India are one of the world's biodiversity hotspots. The Athirappally forest range in the Western Ghats consists of tropical wet evergreen, semi-evergreen and moist deciduous forest types. A part of the forest is less than 300 m above mean sea level and it is mostly of riparian type. There are hardly any forests in the whole of Western Ghats at such a low altitude.

There is a proposal with the state government to construct a hydroelectric project, which may lead to submergence of a part of this unique piece of forest land. Hence, estimation of the quasi option value of preserving the biodiversity of this forest area is important because environmental impacts and economic implications of developmental activities are often ignored in the decision-making process.

This paper tries to estimate the quasi option value by the visitors to Athirappally forests using a double bounded dichotomous contingent valuation method. To enhance the plausibility of estimates for policy framing, Krinsky-Robb confidence intervals for the mean willingness to pay were also estimated. To check the efficiency of the double bounded logit model, a single bounded model was constructed using the responses to the first bid and mean willingness to pay and confidence intervals were estimated.

Preface

Athirappally forests is situated in Kerala State, India, which is bestowed with magnificent waterfalls. The height of the biggest fall is 42 meters and width is 220 meters. The Athirappally forest range extends over 14850 ha and consists of tropical wet evergreen, semi-evergreen and moist deciduous forest types. One special feature of the forests is that part of the forest is less than 300 Meters above Mean Sea Level (MSL) and it is mostly of riparian type. There are hardly any forests in the whole of Western Ghats^[2] at such a low altitude. There is a proposal with state government to construct a hydroelectric project, which may lead to submergence of a part of this unique piece of forestland.

As is the usual practice, the developmental plans in Kerala State have largely ignored the environmental implications. The biological diversity of this State currently faces heavy threat on its existence, mainly due to high human pressure and swift developmental activities. Most of the developmental activities exert adverse impact on the environment and its economic implications are often ignored in decision-making process. However, an increased interest is being created in the international arena to consider all cost and benefits (direct and indirect) in decision-making process.

Specifically- a quasi- option value is estimated here, which is defined as the economic value of choosing not to take irreversible steps if new information about alternative outcomes will be available in future (Bishop, 1978). There is no market price that can be satisfactorily used as direct measure or proxy of value for estimating the quasi option value of biodiversity. However, it is possible to 'construct' markets in order to try and estimate consumer's WTP for such goods and services. The standard methodology to estimate the value is contingent valuation. Hence, this technique was employed on the user group. Here a double bounded format of contingent valuation was used to estimate the economic surplus associated with the conservation of the tropical forest in question.

Methodological overview

Dichotomous choice contingent valuation

The referendum (dichotomous choice) question starts with a specific monetary amount and the respondent is asked whether he/she is willing to pay that amount for the good in question. Single bounded dichotomous choice CVM method was pioneered by Bishop and Heberlein (1979). Here only one dichotomous question is asked with a threshold amount and the respondent is expected to answer either 'yes' or 'no' to that amount.

The dichotomous choice format presents the respondents a take it or leave it price (bid level) for the good being valued. It is most frequently used and is recommended by NOAA (National Oceanic & Atmospheric Administration, USA). This format got more recognition because it emulates the choice that people confront in real markets. A drawback of Dichotomous choice format is the requirement of large sample, which can be, however, overcome by adopting a double dichotomous choice format (Bann, 1999). The double bounded dichotomous choice (DBDC) format supplements the initial question with a follow-up question. Thus this approach gives more information on underlying WTP than single dichotomous choice question (Hanneman, 1991).

The double bounded dichotomous contingent (DBDC) valuation

DBDC contingent valuation method was used in this study to estimate the willingness to pay for biodiversity conservation of Athirappally forests by the visitors. This method was first suggested by Hanneman (1985). If the response to the first bid is 'yes' then a higher bid amount was presented. If the response is 'No' to the first bid, a lower bid amount is presented. It has been shown by Hanneman et al (1991) that the double bounded procedure is statistically efficient than single bounded procedure.

As no previous study was conducted in this forests, there was no apriori base to determine the bid structure. Hence, bid structure was designed by pre-testing, using the payment card format and the bids of DBDC format were designed to accommodate values from Rs.0 to Rs.1200^[3].

A double bounded Logit was employed to analyze the data. For double bounded model, we observe two dichotomous variables, i.e. the answers to the first question and to its follow up question. This method produces four possible outcomes i.e. 'YES-YES' (YY), 'YES- NO' (YN), 'NO-YES' (NY) and 'NO-NO' (NN).

Following Hanemann et al (1991), the following response probabilities were obtained for the Logit model.

P_i ^YY= 1/ (1+ e ^{-(a +b HIGH BID)})
P_i ^NN= 1 - 1/ ( 1+ e ^{-(a +b LOW BID} )
P_i^YN = 1/ (1+ e ^{-(a +b HIGH BID)}) - 1/ (1+ e ^{-(a +b FIRST BID)})
P_i^NY =1/ (1+ e ^{-(a +b FIRST BID)}) - 1/ (1+ e ^{-(a +b LOW BID)})

Where

FIRST BID - Starting Bid value
LOW BID -Low value in Follow-up bid
HIGHBID - high value in Follow-up bid

The double bounded log-likelihood function is of the form

i = 1,2............. 102.
Where I_i indicated the response category of i^th respondent.

Hanemann et al (1991) showed that the mean willingness to pay could be estimated using the formula

WTP* = a / | b |

The truncated or restricted mean was given by

WTP** = ln (1+e^a)/ | b |

Where,

| b | - Absolute value of bid coefficient.

Referendum CVM programs (GAUSS) written by Cooper (1999) was used to estimate the double bounded Logit regression.

Bockstael and Strand (1987) have emphasized that the parameter estimates used to calculate welfare measures (a and b) are themselves random variables. So it is essential to determine a confidence interval for the welfare estimates. Krinsky Robb confidence intervals estimation procedure as suggested by Park et al (1991) was employed in the study. This technique can be implemented using the information given by the estimated logit model i.e. the estimated parameter vector b^ˆ and the estimated variance-covariance matrix, denoted by V^{^}. Multiple random drawings to create a new parameter vector b^{^^} were made from multivariate normal distribution with variance covariance vector V^{^} and mean b^{^}. From each drawing of b^{^^}, WTP was calculated. An empirical distribution for WTP is then obtained for the Logit model using the complete set of replications. A (1-a) confidence interval was obtained by ranking the vector of calculated WTP values and dropping the a/2 values from each tail of the ranked vector. Krinsky Robb confidence interval was obtained using the Referendum CVM programs (GAUSS) written by Cooper (1999).

Data: The required data for the study were collected from 102 visitors to the forest/waterfall using the personal interview method. The sample was chosen randomly and data were collected using a pre-tested interview schedule developed for the purpose. The double bounded format of CVM is used to increase the accuracy of the estimates while reducing the sample size. The population is the visitors to the waterfall /forest.

Results and Discussion

The results of double bounded Logit regression are presented in table 1. The coefficient of bid (b) was significant and negative, which was as per expectations about the respondent's behavior. Coefficient of education was positive and significant, implying that educated would be more aware of the importance of environment and its conservation.. The value of Log likelihood function was high at -118.99. The point estimate of WTP worked out to Rs.206.17 from the DB Logit model (Table 2).

Table 1: Double bounded Logit model results

**Significant at 1 percent level
* Significant at 5 percent level

Since the consumer surplus estimates are derived from maximum likelihood method, the estimates of a (intercept) and b (bid coefficient) are random variables. So construction of confidence interval is essential for the estimates being plausible. In order to get the confidence interval, Krinsky and Robb Monte Carlo simulation technique as adapted by Park et al (1991) was used. The referendum CVM program (GAUSS) written by Cooper was used for this also. The results are presented in the table 2. The 95 percent confidence interval worked out as Rs.111.1 to Rs.277.37, the average and median values being Rs.202.82 and Rs.204.70.

Table 2: Mean WTP and Confidence Intervals (Double bounded logit model)

To check the efficiency of the double bounded logit model, a single bounded model was constructed using the responses to the first bid and mean WTP and confidence intervals were estimated (table 3). Here also bid was significant at 1 percent level and it was negative. The log likelihood was -51.4.

Table 3: Single Bounded Logit Results

The WTP point estimate was Rs.176.0 with a 95 percent confidence interval of Rs.4.58 to Rs.277.65 (table 4). The confidence interval was narrowed in the case of DBDC method compared to single bounded model, which indicates its statistical efficiency.

Table 4: Mean WTP and Confidence Intervals (Single bounded logit model)

The total quasi option value (WTP) of the Athirappally forest was estimated from the unrestricted point estimate of WTP derived from double bounded logit model. Considering the average annual visitation rate (two visits per person) and the payment vehicle (one time payment to the biodiversity conservation fund), the total quasi option value was assessed by multiplying the average WTP with half the total visitors recorded. This was worked out to Rs.65.7 Million .The estimate would not be mixed up with the consumer surplus of the visitors because the proposed dam has a provision to let out water during day time so that water flow to waterfall wont be affected. As the recreation opportunities are not affected the value elicited by CVM is purely the quasi option value.

Conclusion and policy recommendation

Many of the products and services provided by forests do not have a directly observable market or social value. Complete and accurate valuation of forest values is vital for appraising projects and policies that will affect the use of forests. Under-valuation of forest ecosystems distort land use policies leading to conflict with the goal of maximizing economic welfare and this is found to be the key constraint in sustainable forest management. By defining all the values of the forest, forest policy and management decisions can be made more acceptable. The multiple outputs of forest make their economic valuation challenging, but the evolution of appropriate valuation techniques makes this task possible.

Here the quasi option value of the user group (visitors of the waterfall) underlines the importance of considering the non-marketed economic values in cost benefit analyses of projects, which may change the land use from the forests. Expanding the scope pf economic valuation is the need of the hour because of the institutional tendency to integrate the development activity based on its costs and benefits. The addition of considering environmental values to the built nature of institution is absolute necessity to avoid irreversible development that may affect the welfare present or the future society.

Limitations of the study

The people who are visiting the Athirappally waterfall or forest are assumed to have genuine interest in protecting the forest from irreversible developments. Hence the population considered is the visitors, who are from near by and far away Districts of Kerala. Because of time and resource constraints, the people who do not visit and still having genuine interest in protecting the diversity of this forest could not be included in the purview of the study. But still the WTP estimates provides an estimate of quasi option value as the visitors were from all except two out of fourteen districts of Kerala were present in the sample taken and literacy and income levels were comparable.

References

Adamowicz, W., Boxall, P., Willians, M. and Louviere, J., 1998, Stated preference approaches for measuring passive use values: Choice experiments and contingent valuation. American Journal of Agricultural Economics, 80: 64-75.

Bishop, J.T. (Ed.), 1999, Valuing forests: A Review of Methods and Applications in Developing countries. International Institute for Environment and Development, London.

Bishop, R. C., 1978. Endangered species and uncertainty: the economics of a safe minimum standard. American Journal of Agricultural Economics, 60 (1):10-18.

Bishop, R.C., 1982. Option Value: An Exposition and Extension. Land Economics, 58: 1-15.

Bishop, R.C. and Heberlien, T.A., 1979, Measuring values of extra market goods: are indirect measures biased? American Journal of Agricultural Economics, 61: 926-30.

Bishop, R.C. and Heberlien, T.A., 1982, Option Value: An exposition and extension., Land Economics, 58: 1-15

Bowker, J.M. and Stoll, J.R., 1988, Use of dichotomous choice non market methods to value the whooping crane source. American Journal of Agricultural Economics, 70(2): 372-381.

Cooper, J., 1994, A Comparison of Approaches to Calculating Confidence Intervals for Benefit Measures from Dichotomous Choice Contingent Valuation Surveys. Land Economics, 70(1): 111-122.

Cooper, J.C., 1999, Referendum CVM programs. Economic Research Service, USDA, Washington, DC.

Cory,D.C. and Saliba, B.C., 1987, Requiem for option value. Land Economics, 63(1): 1-10.

Dixon, J.A., Scura, L.F., Carpenter,R.A. and Sherman,P.B., 1998, Economic analysis of environmental impacts. Earthscan Publications Ltd, London, U.K.

Edwards, P.J. and Abivardi, C., 1998, The value of biodiversity: Where ecology and economy blend. Biological Conservation, 83(3): 239-246.

Hadker, N., Sharma. S., David, A., and Muraleedharan, T.R., 1997,Willingness to pay for Borivili National Park: evidence from a contingent valuation, Ecological Economics, 21: 105-12.

Hammit, J.K., Liu, J.T. and Liu, J.L., 2001, Contingent valuation of a Taiwanese wetland. Environment and Development economics, 6: 259 - 268.

Hannemann, W. M., 1984, Welfare evaluations in contingent valuation experiments with discrete response. American Journal of Agricultural Economics, 66: 332-41.

Hannemann, W. M., 1985, Some issues in Continuous and Discrete Response Contingent valuation studies. North East Journal Agricultural Economics, 66: 332-41.

Hanemann, W. M., Loomis, J. B. and Kanninen, B. J., 1991, Statistical Efficiency of Double-Bounded Dichotomous Choice Contingent Valuation. American Journal of Agricultural Economics 73:1255-1263

Jordan, L.J. and Elnagheeb, A.H., 1994, Consequences of using different question formats in Contingent Valuation: A Monte Carlo Study. Land Economics, 70(1): 97-110.

Krinski, I and Robb, A., 1986, Approximating the statistical properties of elasticities, Review of Economics and Statistics, 68 : 715 - 719.

McFadden, D., 1994, Contingent valuation and social choice, American Journal of Agricultural Economics, 76: 689-708.

McLeod, D.M. and Bergland, O., 1999, Willingness to pay Estimates Using the Double-Bounded-Dichotomous-Choice Contingent Valuation Format: A test for validity and precision in a Bayesian Framework, Land Economics, 75(1): 115-125.

Mitchel, R.C., Carson, R.T., 1989, Using surveys to value public goods: the method, Resource for the Future, Washington

O'Neill, J., 1997, Managing without prices: The monetary valuation of biodiversity. Ambio, 26(8):546-550.

Park, T., Loomis, J.B and Creel, M., 1991, Confidence intervals for evaluating benefit estimates from dichotomous choice contingent valuation studies. Land Economics, 67 (1): 64 -73.

Pearce, D. and Moran, D., 1994, The economic value of Biodiversity, Earthscan Publications, London, UK.

Pearce, D., 2002, Economic Valuation with Stated Preference Techniques-Summary Guide. Department for Transport, Local Government and the Regions: London, UK.

Peters, C..M., Gentry, A.H. and Mendelsohn, R., 1989, Valuation of an Amazonian Rain Forest. Nature, 339: 655-656.

Shyamsundar, P. and Kramer, R., 1997, Biodiversity conservation-at what cost? A study of households in the vicinity of Madagascar's Mantadia national park. Ambio, 26(3): 180-184.

Smith.K., 1993, Non Market Valuation of Environmental Resources: An Interpretive Appraisal . Land Economics, 69 (1): 1-26.

Stevens, T.H., Echeverria, J., Glass, R.J. , Hager,T and More,T.A., 1991, Measuring existence value of wild life. What do CVM estimates really show? Land Economics, 67(4): 390 - 400.

Appendix

Contingent valuation questions:
SECTION D: Willingness to Pay for Biodiversity Protection

Background information on Athirappally forests

The Athiarappally forests in Thrissur district of Kerala is famous for its waterfalls and virgin forests. Approximately 3.2 lakh people visit this site every year. The Athirappally forest range extends over 14850 ha and consists of tropical wet evergreen, semi-evergreen and moist deciduous forest types. One special feature of the forests is that part of the forest is less than 300 Meters above Mean Sea Level (MSL) and it is mostly of riparian type. There are hardly any forests in the whole of Western Ghats at such a low altitude. There is a proposal with state government to construct a hydroelectric project, which may lead to submergence of a part of this unique piece of forestland.

According to a study by Tropical Botanical Garden and Research Institute (TBGRI), Kerala; the submergible area hosts 329 flowering plants, which includes 24 endemic and 10 rare and endangered species of the Western Ghats. The occurance of Pothos crassipedunculatus is a distributional record, which is hitherto known only from Agasthyamala region. The same report mentions about 215 species on animals which includes threatened species like Asiatic elephant, the wild dog, the mouse deer, the Nilgiri langur and endangered species like great Indian horn Bill and Monitor Lizard. Butterflies like southern birdwing and Budha peacock are the threatened species of butterflies in this area. Broad-billed roller, a rare bird was also recorded in the TBGRI study conducted in this area.

18. Assume that a Non Governmental Organization creates a special fund to raise enough money to protect the Athirappally forest and all the donations to this fund go directly towards preserving the biodiversity of this forest and preventing irreversible developments.

(REMEMBER THIS IS A ONE TIME PAYMENT)