The study site was in the NNT NP, Bolikhamxay Province, Laos. This region represents the largest and most diverse natural and protected forest area in Laos. It has been suggested that the NNT NP contains one of the two largest elephant populations in the country (Duckworth & Hedges, 1999). In and around the NNT NP, elephants occur on the Nakai Plateau (1250 km²) and range into the NNT-Phou Hin Poun (NNTPHP) Corridor (738 km²) that links to the NNT NP and the Phou Hin Poun National Protected Area. This study site included seven villages located in two districts in the NNT NP.

The NNT NP was ranked among the highest priority of Lao protected areas for national and global biodiversity importance. It is one of the largest remaining contiguous forest blocks in Indochina. This status was well established through numerous field studies, including Timmins and Evans (1996), Kottelat (1998), Timmins and Duckworth (1999). Many of the elephants in Laos were killed by humans for their body parts for sale and as products for trade, such as tusks, trunks, feet, tails and skins, or in retaliation for crop raiding.

Over the past few decades, there are a number of economic researchers employing economic valuation as a management tool in conservation and management of endangered species (Bandara & Tisdell, 2002; Imamura et al., 2020; Bhattarai et al., 2021; Kassahun & Taw, 2022). This paper does not attempt to develop new theories or methodologies, but the key contribution of this technique is an innovative application of current economic evaluation given the environmental, social and political context in Laos. Several economic evaluations of elephants has been undertaken typically focused on economic issues involved in the conservation of the African elephant, while those for the Asian elephant have been surprisingly limited.

The data for this paper were collected using a self-administered questionnaire survey with closed and open-ended questions. The target population sample of this study was local people who live in, or adjacent to, the NNT NP in seven villages in two districts in Khammouane Province (Table 1). These seven villages feature rich natural forest resources, rivers and wildlife habitats, especially for elephants.

In order to determine appropriate sample size, the Cochran formula (Cochran, 1963) was used to estimate an ideal sample size with a level of precision of 5%, confidence level of 95%, and the estimated proportion of the attribute present in the population (p = 0.5, referring a 50/50 split of users and non-users). By using the Cochran formula, the minimum size for the current analysis was estimated to be 384. A sample of 490 respondents after excluding a missing value of 10 respondents was interviewed in order to maximize the generalization of the findings.

All interviewees were willing to pay for Asian elephant conservation. Before the conduct of the interview, each respondent was briefed about the intent of the study and questionnaire. The general background information of the research was firstly introduced, followed by questions about on basic information about current status and issues in Asian elephant conservation in Laos and the study area. Respondents were assured that all information and data collected were confidential and would be utilized only for the study. The head of the family was chosen and considered suitable for making decisions on valuation of the WTP of Asian elephant conservation in the study area subject to their family’s budget constraints.

The survey was conducted from October to December 2017 by visiting respondents in their homes. Based on the lists of households from the headmen of each village, a simple random sampling method (Abbott & McKinney, 2013; Sekaran & Bougie, 2016) was used to select the respondents, with a total of 490 respondents from a total seven villages picked (Table 1). Surveys were carried out in the evenings, when respondents had returned from their farmlands.

This paper employed a descriptive statistical analysis in order to understand the characteristics of local people in the conservation of Asian elephants by estimating frequencies and percentages. Logit regression was used to model the relationship between the dichotomous dependent variables to respond to the given amount for WTP and the independent variables (respondents’ characteristics,

Source: Field investigation 2018.

such as age, income, education).

The dependent variable was established from the respondents for the main questions regarding payment. In the analysis, the “yes” response, representing the WTP for conservation of elephants, was coded as 1 and the “no” response, representing unwillingness to pay, as 0. This is expressed by a dependent variable, yi for the i-th respondent.

The currency of WTP in Lao Kip was converted into United States Dollar (USD) based on the exchange rate 1 USD = 11,416 Kip (26 December 2017). The personal characteristic (PC) is an independent variable for the preliminary logit analysis. The variables in the analysis consisted of location, gender, occupation, education level, status, ethnicity and living period in the national park (Table 2). Household characteristics HC) is another factor influencing the WTP or unwillingness to pay to support the conservation of the elephant, namely, number of

Source: Becker and Watslad, 1987.

household members, monthly household income, price willing to pay. Environmental characteristics (EC) determine the WTP for conservation of the elephants in the study area.

Model 1: Logit model based on Becker and Watslad (1987)

L o g i t ( c h o s e n ) = α + β P C + δ H C + θ E C + ε i

L o g i t ( c h o s e n ) = ln ( p r o b a b i l i t y ( c h o s e n = 1 ) p r o b a b i l i t y ( c h o s e n = 0 ) ) = α + β 1 G e n + β 2 A g e + β 3 O c c + β 4 E d u + β 5 S t a + β 6 E t h + β 7 H S       + β 8 T l + β 9 M c e o + β 10 L n ( I n c ) + β 11 p r i c e + ε i (1)

Model 2: OLS based on (Miccheal E. Porter, 1996)

W T P = α + β P C + δ H C + θ E C + ε i (2)

W T P = α + β 1 G e n + β 2 A g e + β 3 O c c + β 4 E d u + β 5 S t a + β 6 E t h + β 7 H S     + β 8 T l + β 9 M c e o + β 10 L n ( I n c ) + β 11 Price + ε i

Note that ε_i is an error term.

Of the 495 respondents, 52.65% were male, 34.49% were between 36 and 45 years-old, 87.96% were married, and 64.49% had either primary education or never attended education (Table 3).

Forty-five point nine-two percent (45.92%) of the respondents mentioned that they used to see the elephant population, 69.80% considered that the population had decreased and almost all respondents insisted that elephant habitat encroachment had increased. It was noticeable that 87.76% of the interviewees mentioned that it was very important to conserve the elephants in the study area and it was surprising that only 23.43% or 71 households said that they were willing to pay on average of USD 1.80 per year to conserve the elephants.

Table 4 presents the descriptive statistics of the main variables used (mean, sd, min, max and number of observations) as well as Chi-square for socio-economic variables. The analysis reveals that the average bid price is USD5.35 with the standard deviation of USD2.51, the maximum value (USD8.76), and the minimum value (USD1.75).

The data gathered from the household survey in NNT NP were analysed to determine factors affecting local residents’ WTP behaviour. Results of the Logit analysis showed that 60% of independent variables were found to have statistically significant effect over respondents’ WTP behaviour (Table 4).

These outcomes were significant, taking account of the socio-economic characteristics of respondents and their preferences for Asian elephant conservation

Source: Estimated from field data, 2018.

Source: Estimated from field data, 2018. Number of obs: 490; Chi² = 85.31; Prob > Chi² =0.0000. “***, **, *” Confident level 99%, 95% and 90% respectively. ^“ns” non-significant 99%, 95% and 90% respectively.

in NNT NP. It is noticeable that six independent variables—viz. Age, Job, Edu, MCEO, Income, and Price were found to be statistically significant. Adds ratio of Age was 1.0245, indicating a positive relationship with their preferences for Asian elephant conservation and implying when residents’ age increased one more year that they were likely to pay 1.0245 times more than the younger group, at the confidence level of 99%. This is because elder respondents were more concerned in conserving the Asian elephants. Similarly, Adds ratio of Ethnic was 1.4147, having a positive relationship with their preferences for Asian elephant conservation. This implies that when surveying households that were Laoloum, a household had a likelihood of paying 1.4147 times more than other ethnic groups. Furthermore, analysis revealed that Income and Edu were found to be positively related with residents’ WTP responses, implying that if monthly household income and education of the respondent increases, the likelihood of contributing more to the conservation of the elephants would also increase.

These outcomes collaborate with well-known the Contingent Valuation Method (CVM) theories and similar types of outcomes stated by others in CVM research (e.g. Bhatt et al., 2014; Zambrano-Monserrate, 2020; Bhat et al., 2021). As expected, the coefficient of household monthly income was found positive, indicating that the likelihood of WTP for conservation of elephant attributes increases with increase in income.

Table 5 shows the calculation of the WTP for the conservation of the Asian elephant in NNT NP from the Logit model. The analysis showed that the variable ‘bid value’ has an expected negative sign, implying an inverse relationship between respondents’ WTP and bid value.

The results of the gender factor estimation showed that the male sample was more willing to pay for the conservation of wild elephants in NNT NP than the female sample at USD 1.20, in accordance with the initial assumption, with a confidence level of 90%. This is similar to findings by White et al. (1997) and Manyama & Roskaft (2014) in that in most rural life it is largely traditional culture handed down from ancestors, with the males who, as heads of families, have the sole authority to decide family issues. Male groups were more likely to be willing to pay for the conservation of wild elephants in the study area.

Source: Estimated from field data, 2018.

The results of the estimation of the age factor of the sample group also showed that the older sample group was more willing to pay for the conservation of wild elephants in the NNT NP than the younger sample group at USD 6.70, in accordance with the initial assumption, with 99% confidence level. This finding aligns with research findings by Kamri (2013) and Manyama & Roskaft (2014). Because older people are more inclined to be more environmentally friendly.

The sample group with jobs as civil servants were more willing to pay for the conservation of wild elephants in the NNT NP than the sample group working in business, trade etc., at USD 0.87. This is similar to Manyama and Roskaft (2014), in that the sample were employed as government employees who worked in various state institutions and had a better understanding of the natural conservation of wildlife.

The results showed that the higher level of education sample group was willing to pay for the conservation of elephants in the NNT NP at USD 12.60 more than the lower level of education sample group. This is a similar finding to Kyophilavong and Bennett (2011) and Manyama and Roskaft (2014). Because most highly educated people are more knowledgeable about the laws of wildlife conservation, these sample groups are more environmentally conscious, have higher incomes.

The results of the estimation of the status factor (Status) of the sample group found that the group who were married and had a family were willing to pay for the conservation of wild elephants in the NNT NP at USD 3.40 more than compared to other groups with the same status as the original 90%. Kamri (2013) and Manyama and Roskaft (2014) reported a similar finding in that families with a married status were more capable of paying more than those with other statuses.

The results of the estimation of the ethnic factors (Ethnic) of the sample found that the Lao ethnic group was willing to pay for the conservation of wild elephants in the NNT NP at USD 1.0 more compared to other ethnic groups, which is in accordance with other studies by Kamri (2013) and Manyama and Roskaft (2014). Because this ethnic group accounts for a large proportion of the population, with permanent settlements and occupations, and less hunting activity than other ethnic groups, they are more inclined to pay for elephant conservation.

It is not surprising that the high-income sample was willing to pay more for elephant conservation in the NNT NP at USD 1.60 compared to the lower-income sample, based on the 99% confidence level. This finding is similar to Truong Dang Thuy (2007), Kamri (2013) and Manyama and Roskaft (2014) in that the high-income sample is more likely to be willing to pay for the conservation of wild elephants in the NNT NP.