02 Nov 2017
Table of Contents
Willingness to pay for green electricity in Small Island Developing State: Mauritius
Small island developing states (SIDS) depend very heavily on imported fossil fuels for their energy needs. The high freight costs for the imports of fossil fuels, a small national fossil fuel demand and diseconomies of scale for electricity generation all combine to make production of electricity very expensive and financially risky in the long term (Weisser, 2004). Moreover, the use of fossil fuels for electricity generation produces air pollutants that have a detrimental effect on human health and the environment (European Commission, 2003). These are of great concerns to SIDS which, due to their smallness, remoteness and often high population density, suffer from significant economic and environmental vulnerabilities (Weisser, 2004; Elahee, 2011). The heavy dependence on fossil-fuels imports also raises the issue of the security of energy supply.
One of the key global strategies for mitigating the harmful impacts of fossil fuels is to reduce their use by replacing them with renewable energy sources (RES). â€˜Green electricityâ€™ is generated from RES such as wind, hydro, biomass, solar or geothermal power, having little or no impact on human health and the environment (Hansla et al., 2008). Unfortunately, one of the principal barriers to the large-scale deployment of renewable energy technologies is their higher cost of installation compared to conventional fossil-fired power plants (Yoo and Kwak, 2009).
Mauritius, like all SIDS, faces the above constraints. In 2008, the Mauritian Prime Minister launched the Maurice Sustainable Island concept which has, as one of its main objectives, the reduction of the dependency of the country on fossil fuels by promoting RES and better energy efficiency (Government of Mauritius n.d.). In line with this concept, government plans to increase the share of RES in electricity generation from 20% in 2010 to 35% in 2025 (MREPU, 2009). However, as expected, a major hurdle toward the implementation of the MSI vision is the substantive amounts of financial resources required (MESD, 2012). In that context, it is vital that all policy mechanisms that can promote renewable energies be carefully examined.
Wiser et al. (1998) highlight three policies that can support RES: (1) Government/regulator imposing on the electricity retailer that a minimum percentage of electricity come from RES, (2) Obligatory surcharge on consumersâ€™ electricity bills to collect funds, (3) Consumersâ€™ voluntary purchase of green electricity. It is the last mechanism that is the focus of this research. I have decided to study Mauritian consumersâ€™ willingness to pay (WTP) for green electricity because it is unknown whether Mauritians would voluntarily pay for green electricity despite popular demands for more RES-based electricity generation. A large number of WTP studies have been carried out in developed countries (e.g. Japan, Sweden, South Korea and USA) but none in SIDS like Mauritius. Such a study would be particularly timely given the current intense debate on the future national energy strategy. As an engineer in the Corporate Planning department of the national electric utility, I believe that the proposed study would have worthwhile policy implications for the utility and the government.
A preliminary review of the literature reveals that investigations to determine the WTP for green electricity has received quite a lot of attention since the mid 90â€™s. WTP studies can be classified into three categories (Zori
and Hrovatin, 2012). The first category consists of determinations of percentage premiums or absolute amounts that customers are willing to pay for green electricity from unspecified RES (e.g. Nomura and Akai, 2004; Hansla et al., 2008: ZoriÄ‡ and Hrovatin, 2012; Zhang and Wu, 2012). The second category employs choice experiment to investigate either, the preference of customers for different shares and types of RES (Borchers et al., 2007; Kim et al. 2012; Gracia et al. 2012) or, the WTP for different shares of generic green electricity to replace fossil fuels and nuclear power using choice experiments (Grosche and Schroder, 2011). To evaluate the WTP for potential environmental benefits (such as reduced emissions, conservation of fossil fuels, preservation of wildlife) resulting from investment in RES, a third category of study (Ku and Yoo, 2009; Adaman et al, 2011) uses the contingent valuation method (CVM) together with the choice experiment (CE).
The CVM is widely used for the determination of the WTP for green electricity and the factors that affect it (Nomura and Akai, 2004; Yoo and Kwak, 2009: Zagrofakis et al., 2009: Adaman et al., 2011; Zhang and Wu, 2012: Kim et al., 2012). It is a non-market valuation technique generally employed in the areas of environmental cost-benefit analysis and environmental impact assessment (Venkatachalam, 2004).
The WTP can be expressed either as a percentage or an absolute amount and either as an increase in the electricity bill or as an increase per kWh of the electricity supplied (ZoriÄ‡ and Hrovatin, 2012). The amount of WTP for green electricity has been seen to vary widely in different studies. This difference is most probably due to the fact that the surveys were carried out in different countries, at different times and using different methodologies. The main results of some research on WTP for green electricity in different countries are shown below.
Table Main findings of some research on WTP for green electricity
ZoriÄ‡ and Hrovatin (2012)
77% of customers were willing to pay an average of â‚¬ 4.18 per household per month
Nomura and Akai (2004)
Median willingness to pay was US $ 17 per household per month
Hansla et al. (2008)
81.5% and 66% were willing to pay at least SEK 0.01 and 0.02 per kWh of green electricity, respectively (1 SEK = US $ 0.15)
Zhang and Wu (2012)
Mean WTP ranges from US $ 1.15 to 1.51 per month for urban residents in Jiangsu province
Kim et al. (2012)
Average WTP was KRW 1562.7 (US $ 1.35) per month per household.
The WTP studies generally show that larger WTP are reported by those with higher income, are younger and are more educated. (Adaman et al, 2011; Zhang and Wu, 2012; ZoriÄ‡ and Hrovatin, 2012). WTP is also related to environmental awareness (Zagrofakis, 2009: Adaman et al, 2011; ZoriÄ‡ and Hrovatin, 2012). Hence, an important policy implication is that education of the public on the positive impacts of renewable energy sources can increase the WTP for green electricity.
In addition to explaining customersâ€™ stated WTP through demographic factors (age, income, education etc.), some researchers (Bjornstad and Kahn 1996; Getzner 2005 cited in Adaman et al. 2011: 690) have suggested that incorporating the social and institutional contexts will help to describe more accurately what drive customers to participate in green electricity programme. The possible effects of institutional and social context can be categorised under three headings (Adaman et al, 2011). The first area of discussion deals with the question of what is the change to the WTP figure when payment is made under compulsion, or voluntarily. The second area of discussion is concerned with the impact the customerâ€™s expectation of others making lesser or firmer support to the renewable electricity project. The third point concerns the effect the identity of the institution in charge of the project has on the WTP. Concerning the first point, Wiser (2007) sees evidence that customers state a higher WTP when confronted with a mandatory surcharge than with voluntary payment whereas the opposite is concluded for the least preferred RES by Borchers et al (2007). The respondentâ€™s WTP amount is also strongly correlated with his expectation of otherâ€™s WTP (Wiser, 2007; Adaman et al, 2011). Concerning the institutional context, the private supply of green electricity prompts a higher WTP than does government supply (Wiser, 2007) whereas the degree of trust the respondent has in the institution responsible for the project had a very strong effect on the WTP (Adaman et al., 2011).
To statistically model the factors that drive customers to report WTP for green electricity, studies generally use a number of discrete choice models such as the probit model (Ku and Yu, 2010), spike model (Yoo and Kwak, 2009; Kim et al, 2012), tobit (ZoriÄ‡ and Hrovatin, 2012) and Mlogit (Zhang and Wu, 2012),.
One of the main criticisms levelled against the WTP studies that use CVM (such as those referenced previously) is that they are generally based on stated preferences rather than revealed preferences. With stated preferences, it is thought that there is the risk that respondents state a WTP figure that is higher than they would actually contribute in reality (ZoriÄ‡ and Hrovatin, 2012). This is illustrated by the contrast between the high levels of stated willingness to support green electricity programmes and the actual very low levels of participation, which generally range between 1 to 3 per cent and do not exceed 10 per cent (ZoriÄ‡ and Hrovatin, 2012). Several studies (Roe et al., 2001; Borchers et al., 2007) point out that the higher price of green electricity is the key factor behind the low actual participation in green electricity programmes. While this appeared economically plausible; Litvine and Wustenhagen (2011) investigated the psychological aspect of the problem to help "light green customers walk the talk". The result of their behavioural intervention survey showed that the purchase of green electricity is not only hampered by the price but that information about the benefits of buying green electricity and targeted information to convince retail electricity consumers are also very important.
Finally, the adequacy of estimated WTP to cater for national renewable energy targets has been tackled by some studies. For the UK (Batley et al. 2001), Italy (Bollino, 2009) and Korea (Kim et al., 2012), the WTP are not enough to meet the national renewable energy targets. In Mauritius, it has been estimated that MUR 5 billion would be required for development of renewable energy projects up to year 2015 (CEB, 2013). Therefore, the results of the proposed study can provide valuable policy information to local decision makers.
Based on my preliminary literature review, I have developed the following research questions and objectives.
How do demographic characteristics such as age, income, education, occupation and geographical location affect WTP for green electricity in Mauritius?
What is the level of awareness concerning environmental impact and economic implications of the dependence on fossil fuel and how does this awareness impact the WTP?
How does the institutional context affect Mauritians willingness to participate? That is does response differ based on whether project is run by a public or private organisation?
What is the average WTP for Mauritius and is this sufficient to fund projects to meet the target of 35 % renewables in the energy mix by 2025?
To design a Contingent Valuation survey to elicit the response of Mauritians on the purchase of green electricity.
To build a statistical model to explain the WTP through demographic factors (age, gender, income, education, location), institutional context and attitude towards the environment.
To evaluate the mean WTP and compare it with the funds required to achieve targeted levels of RES in energy mix.
To propose policy recommendations based on the results of the research.
The research plan will consist of the critical literature review and the data collection.
This research proposal can be classified as a case study design where the Mauritian case is analysed extensively and in detail.
The critical literature review will be carried out so that I learn and understand the work of other researchers who have tackled the same research question as me. More specifically, the critical review will help me to gain knowledge on how to design an effective questionnaire that will accurately elicit the response of people and that will also capture demographic factors, attitudinal preferences. I will also need to perform literature review concerning the statistical models currently used for making statistical inferences in studies like mine. This will enable me to build the relevant model to analyse the survey data obtained. The literature will be in the form of peer-reviewed journal papers that reflect current thinking and government publications.
The secondary data I will need will come from Statistics Mauritius, local researchers, consultancies, websites etc.
Adoption of the CV method for eliciting the WTP for green electricity will require the use of a survey to collect data from the Mauritian population. The survey will be designed as per guidelines from the National Oceanic and Atmospheric Administration (NOAA, 1993). Thus the contingent valuation survey will include an accurate description of the green electricity programme being proposed. I will also include the latest cost per unit of electricity from difference sources of energy such as solar photovoltaic, wind, coal, heavy fuel oil. This will give respondents an idea of the relative costs of green versus conventional electricity. I will develop the survey questions after an extensive review of previous studies conducted in other countries.
Given that I will be required to determine statistical characteristics of the Mauritian population in regards to WTP; I will obligatorily need to use probability sampling technique as recommended by Saunders, Lewis and Thornhill (2009). The first step in the process of probability sampling is identifying the sampling frame. For my research project, the sampling frame is the complete set of residential electricity customers of the island of Mauritius and they number around 347,757 as per latest figures (CEB, 2009). The next step is to decide on a suitable sample size. For my research work, taking into consideration time and money constraints, I have decided that a confidence level of 90% and a margin of error of 7.5% would be appropriate. Based on the equation from Saunders et al reproduced below.
n is the minimum sample size required
p% is the proportion belonging to the specified category
q% is the proportion not belonging to the specified category
z is the value corresponding to the level of confidence required
To cater for the worst scenario, p% and q% have been set to 50. For a level of confidence of 90%, the z value is 1.65. Substituting all the numbers in the equation gives a minimum sample size of 121.
The sampling technique I will use is the stratified sampling technique which can be used to select a probability sample. The stratification will probably be made on the basis of gender, education and age. In fact, I plan to adopt the data collection methodology described by Gracia, Barreiro-HurlÃ©, and PÃ©rez y PÃ©rez (2012). My data collection will proceed as follows: the questionnaire will be administered face-to-face by an interviewer who will randomly select respondents and first ask them a screening question: are you responsible for the payment of electricity in your household? In case of a no, the interviewer will randomly select another individual of a given stratified group until he obtains someone who actually pays the electricity bill.
The validity of the research depends on whether the right variables are being measured to answer the research questions (Jewel and Hardie, 2009). In my case, I will review the numerous peer-reviewed studies carried out in the same area and determine the variables to be selected. Hence, I have much confidence that the validity of the research will be upheld.
The reliability of a research refers to the possibility of the research being replicated by other researchers and obtaining the same results. In my case, replicability is not guaranteed because influencing factors change with time. For example, the electricity tariff might be increased and this will certainly reduce the number of people who will be willing to contribute as well as their individual contributions themselves.
The question of generalizability is certainly interesting since with my results I will be able to confirm whether or not the same factors are at play in determining the WTP in Mauritius compared to developed and transition economies.
Quantitative analysis will be carried out and both descriptive and inferential statistics will be used. Descriptive statistics will be provide information such as mean, standard deviation, minimum and maximum of all variables used such as WTP, age, education etc. The inferential statistics will be carried out using a statistical model which will allow to statistically describe how the WTP is influenced by age, income, education, attitude etc. Since probability sampling will be used, the results obtained can be generalised to the whole population, hence realistic conclusions can be drawn.
I will endeavour to carry out my research work in the most ethical manner possible by adhering to the Coventry University Ethical Procedures. When carrying out interviews for data collection, I will respect the rights and integrity of respondents. I will fill out the low risk ethical approval form before I start my research. In addition, I will use a participation information leaflet and a consent form for my interviews.
Table Gantt chart showing tentative progress of works
Draft literature review
Submit draft to tutor - feedback
Draft research plan
Review secondary data
Make appointment for interviews
Draft interview questions
Draft survey questionnaire
Enter data into software
Draft Chapter 3
Submit draft to tutor - feedback
Modification to draft
Complete remaining chapters
Submit draft to tutor - feedback
Revise draft/format for submission
Print, bind and submit
If you are the real writer of this essay and no longer want to have the essay published on the our website then please click on the link below to send us request removal:Request the removal of this essay
Get in touch with our dedicated team to discuss about your requirements in detail. We are here to help you our best in any way. If you are unsure about what you exactly need, please complete the short enquiry form below and we will get back to you with quote as soon as possible.