Problem statement

The forest is a supplier of several products and services. These services range from the provision of timber and non-timber products to local communities, to water regulation, carbon sequestration, nutrient cycling, and microclimatic regulation (KFMP, 1994; Glenday, 2006). Farmers around Kakamega forest deciding whether to encroach on the forest in order to clear it for agriculture focus on the potential crop yields they may obtain. They pay little heed to the ecological services that would cease to exist if they were to do so. Many of the environmental benefits lost are not only lost to them but also lost to a much wider set of people from the local to the regional to the global level. These are negative external effects (Mishan, 1967) as direct negative effects on other economic agents are not appropriately considered in decision-making. On the other hand, the positive environmental services of the forest are, currently, enjoyed by stakeholders from the local to the global level without

having to pay for these services. Coupled with high conversion pressure because of a high local population density, this situation is likely to result in high deforestation rates - unless forest usage is not effectively regulated well by formal or informal institutions (Ostrom, 1990).

Some of those ecological services transcend local and even national boundaries, benefiting the global community at large. The global benefits derived from a rain forest include direct-use values from recreation (eco-tourism) and from the provision of genetic material for scientific research; indirect-use value in the form of a carbon storage service mitigating global warming; option values in the form of unknown genetic material which may be used for medical purposes in the future; and an existence value derived from the mere satisfaction of knowing that a place exists where a vast number of fauna and flora species live in their natural environment (Andersen, 1997). Thus, there is a need to incorporate as many of those benefits to whomever they may accrue, in order to demonstrate the true value of the forest.

Moreover, given that a good deal of forest benefits and costs fall upon the local communities (Bawa et al, 2004) and considering the fact that continued forest existence depends on the support of these communities, it is deemed essential that valuing their domestic dimension is given priority. This is not to say that forest conservation can be economically viable solely from a local point of view. Strictly local benefits, especially indirect and non-use values, constitute by default just a fraction of the Total Economic Value of the forest since a very large share of benefits accrues regionally and/or globally (Balmford & Whitten, 2003). Thus, local benefits hardly ever suffice to counterbalance the disproportionate magnitude of costs associated with conservation.

Nevertheless, a few studies have supported the conclusion that local benefits exceed costs of conservation, thus justifying conservation from a local perspective (Andersen, 1997; Bann, 1997). This conclusion has been reached through the use of a range of valuation techniques, each aiming at valuing the different benefits accruing to the local communities. The use of different techniques is necessary because different forest values lend themselves to valuation by different kinds of valuation

techniques.Therefore, one should be wary of any attempts to value the multitude of local forest benefits using only one method.

This is especially the case for Contingent Valuation Method (CV) which has been described as a technique that can be employed to measure virtually all types of environmental values (Turner, 1999). The use of CV, however, is often accompanied by doubts about its reliability because various potential biases (strategic, hypothetical, and information bias, as well as embedding) have been identified (Hausman, 1993). Those biases aside, CV may not be the ideal technique for valuing forests since it does not take fully into account the multi-attribute nature of forest values and the presence of complementary and substitution effects (Rolfe et al., 2000). Furthermore, CV cannot easily distinguish between use and non-use values when applied in a context where respondents interact with the environmental resource, in this case Kakamega forest.

A study by Gregory et al. (1993) examines the use of CV methods for estimating the economic value of environmental changes and argues that a principal constraint on the validity of CV is the imposition of unrealistic cognitive demands on respondents.

The authors propose a new CV approach, based on multi-attribute utility theory and decision analysis to better accommodate multidimensionality of value, minimize response refusals, and exclude irrelevancies.

The limitations of CV raise the need to test a different technique for measuring local benefits, namely Choice Modelling (CM). Unlike CV, Choice Modeling does not require survey respondents to place a direct monetary value on a contingently proposed environmental change. Rather, individuals are asked to make comparisons among environmental alternatives, with the environmental good described in terms of its attributes, or characteristics, and the levels that these take. It is the attributes that are important and it is marginal changes in the attributes that are eventually assigned a monetary value. In order to do so, one of the attributes must constitute a monetary amount (Hanley et al., 2001)

The use of both techniques in Kakamega forest constitutes the novelty of this thesis,

both techniques in the context of Kakamega forest enhances our understanding of the values local communities place on the conservation of the forest, especially indirect and non-use values. It can also help test for convergence validity between the values obtained from the two methods and highlight their respective potential for valuing tropical forests in similar contexts.