Please note: This blog post is from Week 4. I began it a couple of weeks ago when I was quite busy with work, but am finally able to finish it now that I have a little down time.
This week I’ve been compiling a literature review of all of the existing studies that have used various forms of economic valuation to estimate the value of sea angling. I have come across studies that have used contingent valuation (CV), the travel-cost model (TCM), and the random utility model (RUM). Though CV and TCM are the primary models of choice, RUM has also been used in association with sea angling/ recreational fishing valuation. Since I have already described TCM in great detail, I will attempt to explain the other two methods. I will do my best to describe how each method is used, which model is best suited for providing specific types of data, as well as the pros & cons of each. This blog post, however, will focus exclusively on the CV method.
From my understanding, contingent valuation is commonly used for ecosystem and recreational valuation. CV provides a foundation for estimating the value of a resource and/or recreation as well as estimating peoples’ willingness to pay (WTP) to protect or maintain the current status/ quality of the given resource/ recreation. It can also estimate peoples’ marginal willingness to pay for resources. CV is oftentimes the first method used to estimate the value an ecosystem/ recreation/ ecosystem service (before TCM or RUM studies). The name “contingent” valuation is derived from how CV surveys are framed; people are asked to state their willingness to pay, which is contingent on a hypothetical scenario and/or description of the environmental service provided. For example, Inland Fisheries Ireland’s CV study was based on the following question,”How much would you be willing to pay each year, through general taxation, for the next 10 years, to preserve the current fish stocks and current quality of recreational fishing in Ireland?”
CV is referred to as a “stated preference” method since it obtains peoples’ stated values/ preferences as opposed to inferring their values, which is what the “revealed preference” methods do (i.e. the RUM model). Since CV relies on how much people say they would pay to protect an environmental resource instead of how much they would actually pay when push came to shove, it can be susceptible to bias. It can also be susceptible to other forms of bias including: strategic bias, embedding affects, and hypothetical bias. One of the studies I included in my literature review completed a CV study of recreational fishing, but took measures to reduce the risk of bias and to negate the existence of bias.
The study is entitled Valuing New Zealand recreational fishing an an assessment of the contingent valuation estimates and was completed by Wheeler & Damania (2011). In the study, they looked at the marginal WTP of each additional fish caught. The ultimate purpose of the study, in part was to determine the value of various fish species in order to compare the value of recreational fish to commercial fish for a proper allocation of the fishing resource. This study is much unlike IFI’s study that uses CV to determine the WTP to protect Ireland’s fish stocks and the quality of the current angling experience; however, it does illustrate the usefulness and and effectiveness of the CV method to value recreational fishing. Once IFI’s CV study is published I will be sure to offer my critiques of their study and provide a summary of how they went about disproving the presence of bias. Until then, here is the only successful example of CV in my literature review and how they approached and disproved the presence of various forms of bias.
Wheeler & Damania (2001) argued that CV was an effective way to estimate the marginal WTP of recreational fishing in New Zealand because it allowed them to determine values for various species of fish (including fish that are caught primarily for sport, for consumption, and fish that are caught both for sport and consumption) and the added value of each additional fish caught. They also argued that contrary to popular belief, CV is a legitimate form of economic valuation, and illustrated this by addressing common forms of bias and how they both mitigated the potential for bias and proved that bias did not compromise the study. The biases they addressed are below:
Strategic bias occurs when people understate their true WTP because they believe they will have to pay a tax/ fee at some point in the future. To test for strategic bias, Wheeler & Damania (2001) asked a follow-up question at the end of the survey: “Do you believe that the government will impose a recreational fishing tax in the next year or so?” They then ran two separate models (one for those who thought a tax would be implemented [17.5% of the total sample] and one for those would did not think a tax would be implemented [82.5% of the total sample]) and then compared the statistical significance of each model. Wheeler & Damania  ultimately found that both models were found to be highly statistically significant and furthermore, the models were comparable in terms of statistical significance. As such, they were able to combine the two samples and infer that their survey did not suffer from strategic bias.
This CV study was also susceptible to embedding effects. Embedding effects can occur CV studies like this one where respondents are asked to estimate how much they would be willing to pay for each additional fish caught. Embedding effects, also referred to as “perfect embedding” occurs when respondents state that they are willing to pay the same amount for goods that differ in quality or different amounts for the same good (Carson & Mitchell 1995). Wheeler & Damania  were able to negate the presence of embedding effects . First, they tested consumers’ WTP for 5 different fish species and discovered substantial differences between the marginal WTP of the various species. Second, they tested different models (one for anglers who caught and kept 4 or more snapper and one for anglers who caught or kept 3 or fewer snapper). They discovered that the marginal WTP was $10.25 for those who caught and kept 3 or fewer fish and $2.94 (Wheeler & Damania, 2001: 611) for those who caught and kept 4 or more fish, indicating that the value fishermen attach to fish are greatly influenced by the number of fish they are able to catch. The difference in attached value for various fish species and number of fish caught disproves the presence of embedding effects.
A third type of bias that can affect any type of CV study is hypothetical bias. Hypothetical bias occurs when respondents are unfamiliar with the product/ resource being valued. Hypothetical bias was invalidated in this case since surveys were conducted at boat ramps and at the end of fishing trips, so the “hypothetical” situation is non-existent.
In sum, Wheeler & Damania (2001) managed to use contingent valuation to estimate the value of recreational fishing in New Zealand in a fairly sound manner. They were able to test and disprove various forms of bias and came up with figures comparable to what is found in the RUM studies that also seek to value the value of each additional fish caught/ reveal preferences for fish species. Reading Wheeler & Damania’s study (2001) gave me a better understand of the CV and its validity.
Willingness to Pay: the maximum amount an individual is willing to pay to protect a resource/ recreation. In the case of sea angling, protecting the resource/ recreation involves: maintaining fish stocks and water quality, proper allocation of fish resources (between anglers and commercial interests), limiting pollution, as well as providing sufficient monitoring and enforcement for all of the aforementioned criteria.
Marginal Willingness to Pay: the amount someone is willing to pay for each addition ‘good’, which in the above case is each additional fish caught and consumed.