Rural Residents’ Intention to Participate in Pro-Poor Tourism in Southern Xinjiang: A Theory of Planned Behavior Perspective

: It is well recognized that the sustainability of pro-poor tourism (PPT) lies in the continuous support and participation of the poor themselves. Previous studies have attempted to explore the effectiveness and patterns of pro-poor tourism, yet few have empirically explored the determinants of the poor’s intention to participate in PPT. Drawing on the theory of planned behavior (TPB), this study investigates the impacts of the attitude of rural residents towards PPT, the perception of support for PPT, and the participation behavioral control of PPT on their intention to participate in PPT in Southern Xinjiang, China. One thousand and twelve valid questionnaires were collected via a household survey in six National Pilot Villages of PPT in Southern Xinjiang. According to the results, the locals’ attitude towards PPT and their participation behavioral control of PPT both positively affected their intention to participate in PPT. However, there was no signiﬁcant correlation between the perception of support for PPT and their participation intention. This study highlights the importance of endogenous impetuses for residents of remote poverty-stricken areas with harsh environmental conditions, such as Southern Xinjiang, to participate in state-aided projects.


Introduction
In the late 1990s, the concept of pro-poor tourism (PPT) was first proposed by the UK Department for International Development, with the benefit of the poor being its focus [1].Since then, PPT has been proactively promoted as an important means to reduce poverty by governments worldwide.Previous studies have verified the effectiveness of tourism on poverty alleviation in both developed [2] and developing regions [3][4][5][6].However, the studies on PPT have mostly focused on the model [7,8] and macroeffect from the perspective of decisionmakers [9][10][11][12][13], whereas empirical studies from the microperspective of the people in need have been relatively limited [14][15][16].This is problematic, as the poor are the focus of PPT.Notably, the participation of the poor in tourism, and the benefit that they gain, depend on a series of critical factors, such as the category of tourism, market context, training, policy, and external support [17].For example, in China's Southern Xinjiang, although there are unique natural and cultural tourism resources in the region, the locals show insufficient intention to participate in the process of assistance via PPT to seek further development.This can inhibit the continuous poverty-reduction effect of local tourism, as the sustainability of PPT lies in the continuous support and participation of the poor themselves [9].Accordingly, finding the factors that affect the intention of the poor is of importance to facilitate their participation in PPT.
The theory of planned behavior (TPB) holds that human behavior is conscious and active induced behavior that depends on specific behavioral intentions and that is a sub-jective probability judgment [18].Behavioral intention indicates the subjective probability that an individual plans to carry out an action, and it is influenced by attitude, subjective norms, and perceived behavioral control [19].The TPB has been applied to analyze the relationship between the will and behavior of residents in the process of poverty reduction [20,21], but it has rarely been adopted in the context of PPT to understand the poor's participation intention and behavior.Given the lack of microlevel PPT research [14], this study draws on the TPB to investigate the impacts of the attitude of rural residents towards PPT, their perception of support, and the participation behavioral control of PPT on their intention to participate in PPT in Southern Xinjiang, China.It contributes to the tourism and sustainable-development literature on the poor's perspectives on PPT, with empirical evidence from an underdeveloped ethnic-minority area of China's Xinjiang region.
The remaining part of this paper is constructed around four sections.The literaturereview section explains and contextualizes the main components of the TPB in the context of PPT in Southern Xinjiang, and it critically outlines studies that relate to the participation of residents in tourism, thereby establishing the conceptual framework.The methodology section explains, in detail, the research context, survey instrument, and data collection and analysis.The findings section reveals that the attitude towards PPT and the participation behavioral control of PPT both positively affected the locals' participation intention, yet there was an insignificant correlation between the perception of support for PPT and their participation intention.These findings are then discussed in the final section, with a focus on the inconsistency with conventional wisdom (i.e., the poor's participation intention is uncorrelated to their perception of the government support for the program).

Literature Review and Research Hypotheses
The TPB is a psychological theory that links beliefs to behavior.The theory holds that three interrelated components, namely, attitude, subjective norms, and perceived behavioral control, together shape an individual's behavioral intention, which, in turn, determines his/her actual behavior [22].These three components are not only related to people's subjective perceptions, but they are also subject to the influence of objective factors that can have an impact on subjective perceptions [23].
In tourism studies, the theory is generally used to explore tourist behaviors, such as the adoption of more sustainable lifestyles [24], undertaking environmentally responsible behavior [25], and the use of smart tourism technologies (STTs) [26].Tourism scholars have also adopted the TPB to study the host attitudinal support for tourism and tourism development since the 2010s [27].Shen et.al (2020), for instance, investigated the effects of residents' place attachment (PA), subjective norms (SNs), and perceived behavioral control (PBC) on their attitudes towards behavior (AB) and their behavioral intention to support tourism (BI) [28].While the TPB has been proven to be an effective theoretical perspective from which to explain the attitude towards and support for the tourism of residents [27][28][29], it should be noted that, in these studies, the attitudinal support for tourism and tourism development are often measured at the collective or destination level.In this sense, it is different from the behavioral support intention, as reflected by the individual intention to participate in tourism.As for the latter, the applicability of the TPB has rarely been confirmed.A notable exception is the study of Lu et al. (2017) on the willingness of impoverished Tibetan farmers and herders to participate in tourism.They found that the behavior attitudes of the poor farmers and herdsmen, the support of important others, and personal competence/control all had significant positive impacts on their willingness to participate in tourism poverty alleviation [30].
According to the TPB, attitude is a positive or negative perception that is formed by people's generalization of a certain behavior.It is the result of belief intensity and value estimation.In the context of PPT, whether it can increase employment, increase income, and improve residents' life skills may be the intuitive factors that lead to attitudes towards PPT.Studies have consistently verified the practical benefits that result from tourism as a powerful determinant of their support for tourism [31], and especially in the poverty-stricken areas of China [32].A subjective norm refers to the social pressure that individuals feel about whether to carry out a specific behavior.Hence, the influence of objective factors, such as social systems [33], fairness [34], and the natural environment [35], may be considered to understand the formation of subjective norms.In the context of Southern Xinjiang, which is a remote poverty-stricken area with limited arable land and natural resources, tourism is considered by the state to be the major viable modern sector to lift locals out of poverty, and proactive measures have thus been taken to promote locals' participation in PPT.Residents often receive tourism-poverty-alleviation materials provided by the government, and skills training provided by local tourism enterprises.Due to the nonmandatory nature of participation in PPT, such active support from relevant parties may be considered as social pressure on the poor to participate in PPT.Perceived behavioral control refers to an individual's perceived ease or difficulty in performing a specifical behavior.When individuals think that they have more resources and opportunities and less obstacles, the stronger their perceived behavioral control will be.Studies have noted that the participation ability of poor residents is the most powerful determinant of their willingness to participate in tourism [36,37].In the context of Southern Xinjiang, this has much to do with the richness of the local tourism resources, the availability of participation opportunities, and the stability of the local tourism.
To this end, based on the TPB, as well as on previous similar studies as presented in Table 1, a conceptual framework was thus developed for this study, as shown in Figure 1.The attitude towards PPT, the perception of support for PPT, and the participation behavioral control of PPT were expected to affect the intention of the residents to participate in PPT, which could further stimulate the internal needs of self-development.Consistent with the framework, the following three hypotheses were made: Being rational and emotional: An integrated model of residents' support of ethnic tourism development [40] Residents in Xijiang Miao Village, Leishan County, Guizhou Province, China (N = 294) Residents' perceived benefits positively influenced their support for tourism development.
Research on impoverished Tibetan farmers' and herders' willingness and behavior in participating in pro-poor tourism: Based on a survey of 1320 households in 23 counties (cities) in the Tibetan areas of Sichuan province [30] Farmers and herdsmen in Tibetan areas of Sichuan (N = 1320) The attitude of poor farmers and herders towards PPT had a positive impact on their participation behavior; the support perception of important others had a significant positive impact on their willingness to participate in PPT; perceived behavioral control had a significant positive impact on their willingness to participate in PPT.Residents' perception of place positively affected their perception of tourism impacts, which positively influenced their support for tourism development.
Considering emotional solidarity and the theory of planned behavior in explaining behavioral intentions to support tourism development [45] Residents in Izmir (Turkey) (N = 740) Perceived behavioral control significantly predicted their behavioral intentions to support tourism development.
Local intentions to participate in ecotourism development in Taiwan's Atayal communities [46] Residents in three Atayal communities in Yilan County, Taiwan (N = 301) Perceived behavioral control had a positive influence on the residents' behavioral intentions to participate in ecotourism development.
Sustainability 2022, 14, x FOR PEER REVIEW 3 of 17 and improve residents' life skills may be the intuitive factors that lead to attitudes towards PPT.Studies have consistently verified the practical benefits that result from tourism as a powerful determinant of their support for tourism [31], and especially in the povertystricken areas of China [32].A subjective norm refers to the social pressure that individuals feel about whether to carry out a specific behavior.Hence, the influence of objective factors, such as social systems [33], fairness [34], and the natural environment [35], may be considered to understand the formation of subjective norms.In the context of Southern Xinjiang, which is a remote poverty-stricken area with limited arable land and natural resources, tourism is considered by the state to be the major viable modern sector to lift locals out of poverty, and proactive measures have thus been taken to promote locals' participation in PPT.Residents often receive tourism-poverty-alleviation materials provided by the government, and skills training provided by local tourism enterprises.Due to the nonmandatory nature of participation in PPT, such active support from relevant parties may be considered as social pressure on the poor to participate in PPT.Perceived behavioral control refers to an individual's perceived ease or difficulty in performing a specifical behavior.When individuals think that they have more resources and opportunities and less obstacles, the stronger their perceived behavioral control will be.Studies have noted that the participation ability of poor residents is the most powerful determinant of their willingness to participate in tourism [36,37].In the context of Southern Xinjiang, this has much to do with the richness of the local tourism resources, the availability of participation opportunities, and the stability of the local tourism.
To this end, based on the TPB, as well as on previous similar studies as presented in Table 1, a conceptual framework was thus developed for this study, as shown in Figure 1.The attitude towards PPT, the perception of support for PPT, and the participation behavioral control of PPT were expected to affect the intention of the residents to participate in PPT, which could further stimulate the internal needs of self-development.Consistent with the framework, the following three hypotheses were made:

Hypothesis 1 (H1).
The attitude towards PPT has a positive impact on the intention to participate in PPT.

Hypothesis 2 (H2).
The residents' perception of support for PPT has a positive impact on their intention to participate in PPT.

Hypothesis 3 (H3).
The participation behavioral control of PPT has a positive impact on the residents' intention to participate in PPT.
However, it is noted that the tourism-participation intention is often related to demographic characteristics that can impact the subjective probability judgments of residents, which then affects their behavior [47].Specifically, factors such as age, education, gender, ethnicity, household income, and family structure can cause significant differences in the effects of PPT, which further influence their willingness to participate in PPT [48].Moreover, it cannot be denied that objective factors are also at play, as they can act on residents' subjective perceptions.Studies have found that the natural ecological environment [35] and social public service facilities [49,50] for the survival of poor residents are the objective factors for the formation of tourism-participation intentions.To this end, we treat relevant demographic factors and objective factors as control variables to better delineate the impacts of the attitude towards PPT, perception of support for PPT, and participation behavioral control of PPT on the locals' intention to participate in PPT in China's Southern Xinjiang.

Research Context
Data used in the study were collected via a questionnaire survey in six National Pilot Villages of PPT.The six villages were in the national-level poverty-stricken counties in Southern Xinjiang, including Tashikuergan County, Aketao County, and Shufu County in Kashgar Prefecture, and Hotan County in Hotan Prefecture (Figure 2).Five of the villages in Kashgar Prefecture are located along the most classic tourist routes of Southern Xinjiang, while the village in Hotan is near the only 4A scenic spot (i.e., the most well-known tourist attraction) of Hotan Prefecture.Therefore, the residents of the villages were not unfamiliar with tourism.Moreover, thanks to the lead author's social networks, these villages had good accessibility for data collection.

Questionnaire Design and Data Collection
The scale of the attitude towards PPT was adapted from Lu (2017) [30], which includes three items: "PPT helps to increase my employment opportunities", "PPT helps to

Questionnaire Design and Data Collection
The scale of the attitude towards PPT was adapted from Lu (2017) [30], which includes three items: "PPT helps to increase my employment opportunities", "PPT helps to increase my income", and "PPT helps to improve my life skills".For the measurement of the perception of support for PPT, the items were adapted from the scale used by Sirivongs and Tsuchiya (2012) [42].They are: "I am satisfied with the tourism service materials provided by the government", "I am satisfied with the number of tourism training provided by government and tourism enterprises", and "I am satisfied with the incentive of pioneer of PPT".For the measurement of the residents' participation behavioral control of PPT, the items were adapted from the scale used by Lu (2017) [30] and Truong (2014) [51], but they were reworded with regard to behavioral control [52,53].This was measured by "I have confidence in local tourism resources, "I have opportunity to participate in PPT", and "I think the local tourism industry is fragile".The item "I think the local tourism industry is fragile" is a negative statement, which was reversely coded for later analysis.
The questionnaire design was based on the scale and comprises five parts: the attitude towards PPT, the perception of support for PPT, the participation behavioral control of PPT, combined with the residents' individual characteristics and tourism resource conditions.The first three parts are as described above, and the fourth part included gender, generation (age), education level, annual household income, Mandarin level, and satisfaction with personal health status.A resource endowment is a prerequisite for residents to participate in tourism poverty reduction, and the fifth part included four such indicators: the annual household income, housing structure, residential location, and proximity of residence to a scenic spot.All the items were measured on a five-point scale, ranging from 1 (strongly disagree) to 5 (strongly agree).
The questionnaire was first pilot tested with 50 residents in the old city of Kashgar in August 2019.Minor alterations were made with regard to the rewording of some items.The total number of households in the six National Pilot Villages of PPT was 2313 (on-site survey data), and the questionnaire survey was conducted with a convenient sampling approach in the six villages from October to December in 2019.Notably, the questionnaire survey was carried out with the strong support of county-level tourism-management departments, village committees, and the students of Kashgar University and the Hotan education college.With the assistance of the village committee, 200 questionnaires were distributed to each village.A total of 1200 questionnaires were collected, of which 1012 were valid.

Data Analysis
STATA 16.0 was employed to conduct the descriptive and statistical analyses of the data.The ordered logit model (OLOGIT) is the standard model for ordered dependent variables [54].As the intention to participate in PPT is treated as an ordered-category variable, the OLOGIT was selected for verification to analyze the impacts of the attitude towards PPT, the perception of support for PPT, and the participation behavior control of PPT on the intention to participate in PPT (Model 1).
When the dependent variable is a continuous-category variable, ordinary least squares (OLS) regression is often appropriate, and the predicted values from the OLS and OLOGIT were highly correlated [55].If the results of the OLS and OLOGIT are similar, then the adoption of OLS may be simpler and easier to explain [56].Thus, the OLS was used to test the robustness (Model 2), and the results are almost consistent with the results of the OLOGIT, except for "proximity of residence to scenic spot".

Descriptive Analysis
Most of the respondents (94.57%) were minority villagers, including Uighur, Kazak, Tajik, and Kirgiz, which reflects the characteristics of the multiethnic settlement in Southern Xinjiang.The sample was balanced in terms of gender (51.09% male, 48.91% female).The education level of the villagers in this area is generally low (the education level of junior middle school and below accounted for 68.58%), and only 10.57% of the residents had a college degree or above.The proportion of households whose per capita cultivated land area of villagers was less than China's poverty criteria (0.6/0.8 mu, 1 mu ≈ 666.67 m 2 ) was 37.85%.In Southern Xinjiang, the desertification of local saline alkali land is serious, and the area of cultivated land is seriously insufficient.To this end, the sample characteristics are consistent with the general conditions of the rural residents in Southern Xinjiang.
The government sent cadres stationed in the village to organize villagers to transform the courtyard, create a "courtyard economy", and realize self-sufficiency in growing vegetables.However, 49.01% of households still rely on government subsidies as their main source of income.The households whose tourism income accounted for more than 50% of their household income was only 5.93%.Nearly half of the respondents were from families whose tourism income accounted for less than 10% of their family income.The villagers are more dependent on agriculture and animal husbandry (38.53%), forest and fruit planting (25.20%), and the traditional handicraft industry (27.60%).Although they have the desire to participate in the tourism industry, in terms of the way to participate, they are most willing to independently operate catering, farmhouse entertainment, or provide raw materials (31.42%).They do not know much about the stable work provided by tourism enterprises, and they relatively exclude the fixed work provided by hotels.They think that the work of waiters is too tired, which also affects their dignity.Driven by the poverty alleviation of the local tourism industry, 76.19% of the respondents had received tourism-related training in the past year (see Table 2).As shown in Table 3, in terms of the attitude towards PPT, the relative mean score of the construct (M = 3.22, SD = 1.42) suggests that, overall, compared with the government's active policy of PPT, the residents' attitude towards PPT was not that positive.In terms of the perception of support for PPT, the relatively low score of the construct (M = 3.12, SD = 0.66) suggests that supports from government and other social organizations are not well acknowledged by the poor.Although tourism service materials, such as facilities to operate B&Bs or vegetable seeds, were provided by the government to the poor, they were unwilling to work with these materials.Notably, the residents were not satisfied with the incentives of PPT pioneers (M = 2.80, SD = 1.23).This means that the exemplary role of PPT pioneers is not realized.In terms of the participation behavior control of PPT, to some extent, the residents were willing to believe the value of local tourism resources (M = 3.27, SD = 1.53), and they may have the opportunity to participate in tourism and to benefit from it with the promotion of PPT by the government (M = 3.23, SD = 1.53).However, they realize, to some extent, the vulnerability of local tourism (M = 3.27, SD = 1.51).Overall, the intention to participate in PPT was moderately strong (M = 3.26, SD = 1.06).

Regression Analysis
The OLOGIT (Model 1) and OLS (Model 2) were adopted to investigate the relationship among the different constructs.As shown in Table 4, two regression analyses were conducted.It is worth pointing out that the proportion of poverty caused by disease is relatively high in Southern Xinjiang, and so personal health status is regarded as one of the control variables.Moreover, the local ethnic minorities account for more than 95%.The Mandarin level of the residents directly affects their communication and understanding of the government's tourism-poverty-reduction policy.Therefore, their Mandarin level is also regarded as one of the special control variables.As Southern Xinjiang has a vast area, the geographical proximity of the residences to scenic spots is also considered as one of the control variables.In Model 1 and Model 2, the attitude towards PPT, the participation behavior control of PPT, and the Mandarin level had significant impacts on the intention to participate in PPT.The pseudo R 2 of Model 1 is between 0.16 and 0.17, which meets the fitting standard.The Adj-R 2 of Model 2 is 0.791, and the fitting effect is more ideal (see Table 4).Firstly, the attitude towards PPT had a significant positive impact on the intention to participate in PPT in both models (Model 1: 0.489; Model 2: 0.220), which indicates that the higher the local residents' recognition of tourism poverty reduction, the stronger their participation intention.Secondly, the participation behavior control of PPT also had a significant positive impact on the residents' intention to participate in PPT in both models (Model 1: 0.509; Model 2: 0.224), which indicates that the higher the residents' understanding of the local tourism resources and tourism market, the more confident they are in the development of tourism, and the stronger is their intention to participate.Thirdly, the proximity of the residence to a scenic spot was also a positive predictor in the model (Model 1: 0.065).However, the perception of support for PPT had no significant impact on the intention to participate in PPT in both models, and, thus, Hypothesis 2 is not supported.In this sense, the supporting policies by the government to promote the participation of locals in pro-poor tourism are not that effective.

Analysis of Influencing Factors (1). Analysis of attitude towards PPT
The residents' attitudes were affected by their perceptions towards the implementation of local tourism poverty alleviation and the advantages of tourism resources.The impact of the attitude towards PPT on the participation intention was significant at the level of 1% in both models.This means that, when residents perceive that pro-poor tourism can bring more practical benefits, such as employment opportunities, their intention to participate in PPT is likely to increase.This result responds to existing studies [31,32,57].The level of social and economic development in Southern Xinjiang is relatively backward, but, in recent years, the basic livelihood of the poor residents has been greatly improved thanks to the government's poverty-alleviation policy.According to the World Bank's poverty standard of USD 1.9 per person per day, over the past 40 years, the incidence of poverty in China has decreased from 88.1% in 1981 to 0.3% in 2018, and the number of people living in poverty has decreased by nearly 800 million.During the eight years of poverty alleviation, the special financial poverty-alleviation funds invested in by different levels of governments amounted to nearly CNY 1.6 trillion, of which the central government invested CNY 660.1 billion.At the macrolevel, the government implements many measures at the same time, including paying attention to education, implementing export-oriented development, paying attention to infrastructure investment, maintaining macroeconomic stability, and supporting market competition.At the microlevel, China's poverty-reduction experience includes "using agricultural technology extension as a driving force for the improvement of agricultural productivity", "use the power of e-commerce and strong domestic logistics network to integrate rural areas into urban supply chain", "using rural tourism industry to drive rural revitalization", etc.These beneficial measures and policies direct the attention of the residents of Southern Xinjiang to the improvement in living standards brought about by poverty reduction.Thus, the attitude towards PPT matters when it comes to participation in tourism poverty reduction.
(2). Analysis of perception of support for PPT The perception of the support for PPT had no significant impact on the intention to participate in PPT in both models; that is, the thrust from the government does not play a direct role in the residents' intention to participate in PPT.This contradicts previous findings regarding a positive correlation between policy satisfaction and the intention to participate [4].The local tourism-management department has provided residents with skills training in farmhouse/B&B management and has provided basic service facilities.For example, the government helps the residents to transform their houses into buildings with folk characteristics, to set up farmhouses/B&Bs, and even to prepare basic household appliances for the operation of farmhouses/B&Bs.Poor villagers can obtain the support of the poverty-alleviation working team sent by the government, the local tourism administration, or state-owned travel enterprises.However, there was no significant correlation with their intention to participate in PPT, which means that their enthusiasm has not been mobilized.In fact, less than 10% of the residents rely on tourism for family income.
While tourism fits well into neoliberal interpretations of poverty alleviation, it may lead to inequalities if it is allowed to operate in a free-market environment [58].The residents do not have sufficient funds or skills to join the tourism industry.Moreover, the residents have a low satisfaction level with the "PPT pioneers" in Southern Xinjiang.In the field investigation, it was found that the elected leaders practice nepotism with village cadres, which means that fairness needs to be improved.Finally, the personality traits of the poor residents also seem to be at play [59].Nearly one in five residents who obtain poverty-relief materials issued by the government, such as seeds and breeding sheep, would bury the whole bag of seeds in the ground instead of farming them, and would eat the breeding sheep instead of breeding them.Because of the serious salinization of the local land, the villagers have only planted highland barley for a long time, and they are unwilling to try to plant new crops.This indicates that the locals' internal development needs have yet to be motivated.

(3). Analysis of participation behavioral control of PPT
The participation behavioral control of PPT had a significant positive impact on the intention to participate in PPT.The results show that, for each additional unit of the residents' participation behavioral control of PPT, the intention to participate in PPT increased by 0.509 units (M1) and by 0.224 (M2).It is thus the strongest predictor of the participation intention in both models.The participation behavioral control of PPT could be the internal factor that determines the residents' participation in action.The residents in Southern Xinjiang have confidence in the local tourism resources, and they believe that the unique local natural resources can bring them opportunities to participate in tourism development, although they have not yet realized that their own lifestyles are a special tourist attraction.The authenticity of natural and cultural resources is well protected here.Moreover, the Chinese government has continuously improved the tourism-development environment and has publicized local tourism resources in recent years to support the development of the tourism industry.The local tourism-management department also continuously helps residents to improve their tourism-participation skills.Locals would be aware of more adequate development opportunities if they were willing to participate in tourism poverty reduction.
However, the external natural environment can profoundly affect the ideology of the residents living in this environment for a long time, which thus affects their decisions.The ecological environment and natural climate conditions in Southern Xinjiang are extremely harsh.The season that is suitable for tourism is only four months a year, from June to October.For farmers and herdsmen, there is more security in obtaining sufficient food through the agriculture and animal husbandry that they depend on than in obtaining income by participating in tourism, with its weak stability.The locals believe that tourism income can be used as a supplement rather than as a backbone of family income.In addition, economic barriers should not be ignored.The development of tourism in remote and underdeveloped areas, such as Southern Xinjiang, is also faced with limited financing channels, a lack of infrastructure construction, low-level agricultural industrialization, single industrial structures, and limited tourism enterprises [60].Such factors are far beyond the control of the locals and can also partially explain their weak intention to participate in PPT.

(4). Analysis of personal characteristics
Firstly, the Mandarin level had a significant negative impact on the intention to participate in PPT (Model 1: −0.087; Model 2: −0.043), which means that the higher the Mandarin level, the less the residents wanted to participate in PPT.Ethnic minorities account for more than 95% in Southern Xinjiang, and the residents have limited Mandarin communication skills.Thus, it is not surprising that the residents' Mandarin communication abilities can affect their intention to participate in PPT in Southern Xinjiang.Generally, the better their language communication abilities are, the more likely it is that the residents will participate in PPT [41].
However, this was not the case in this study, as the residents with good Mandarin in Southern Xinjiang were more willing to take jobs with fixed wages.They are generally more willing to work for local government or tourism-management departments.The Mandarin language is a bridge between the local people and the outside world, and it can determine their risk-prevention awareness and PPT decision making.They can more carefully weigh the advantages and disadvantages of government support and participation in PPT because their Mandarin level is higher, and they can make choices that are more beneficial to themselves.
Secondly, the geographical proximity of the residence to a scenic spot had a positive impact on the intention to participate in PPT in Model 1 (0.065), which indicates that the closer the residents are to the scenic spots, the more likely they are to participate in PPT.Residents close to the scenic spots (within 5 km) showed higher enthusiasm when talking about tourism, and they put forward clear development needs.Moreover, the field survey found that even the residents who lived far away from scenic spots (more than 20 km) hoped to completely eradicate poverty through government assistance and participation in tourism.However, they rely more on agriculture and animal husbandry, and they are not sure that tourism can help them to maintain a stable livelihood.

Discussion
This study investigates the linkage between the intention to participate in PPT and the attitude towards PPT, the perception of support for PPT, and the participation behavioral control of PPT.The attitude towards PPT and the participation behavior control of PPT influenced the residents' intention to participate in PPT in Southern Xinjiang, while the perception of support for PPT had no impact on the intention to participate in PPT, although the local government has given great support to pro-poor tourism.Moreover, the control variables of the sociodemographic characteristics, such as the Mandarin level and the proximity of the residence to a scenic spot, were significantly correlated with the residents' intention to participate in PPT.
Firstly, the study finds that, contrary to the conventional wisdom, the poor's participation intention was uncorrelated to their perception of government support for the program.This corresponds to Banerjee and Duflo's [61] observation that the poor suffer from a lack of incentives to make the best of publicly provided opportunities.To some extent, the poor residents in Southern Xinjiang perceive the support from the government's PPT policies and the assistance of state-owned tourism enterprises.They can judge the positive prospects and possible dividends of PPT, and they show a positive attitude towards eradicating relative poverty.However, they also have sufficient anticipation regarding the harsh local natural environment and the vulnerability of tourism development.Therefore, they prefer conservative choices, they follow the psychological law of "easy refuge", and they seek more stable livelihoods that match their own abilities, such as the original agriculture and animal husbandry, and the work of border guards or forest guards that is provided by the government.They are more willing to accept the help of the government than develop in a challenging tourism sector.Thus, the external aid for PPT from the government and society only stays at the residents' attitude of acceptance, and is not powerful enough to activate the locals' intention to participate in PPT.
Secondly, in previous studies, the intention to participate in PPT has been mainly attributed to the abilities of the poor [17], the capital that they have, social exclusion [62], individual characteristics [59], and education [63,64].The control variables in this paper consider the important influence of communication skills in remote ethnic areas, such as language, on the participation intention, which is consistent with Wasudawan's (2018) research in Sarawak.In remote and poor areas of China, such as Southern Xinjiang, the level of Mandarin can significantly affect the willingness of poor residents to participate in PPT [65].Notably, in areas with wide geographical areas, such as Southern Xinjiang, the geographical proximity of residences to scenic spots can have a significant impact on poor residents' intention to participate in PPT.
Finally, the impact of natural, humanistic, and other objective conditions on the residents' intention and behavior towards participating in PPT in deep poverty-stricken areas with backward productivity is worthy of further discussion.The richness and particularity of tourism resources have brought opportunities for poverty reduction in poor areas [66].However, it is undeniable that the economic and sociocultural contexts may limit the choices of poor residents [67], and they may even lead to their lack of awareness of selfdevelopment.This, in turn, limits the effect of PPT in underdeveloped areas, as is shown in our study, to some extent.Affected by the natural and geographical environment, Southern Xinjiang has formed a unique regional cultural atmosphere, which affects the formation of the subjective values of the locals.The impact of the group intention characteristics formed under such a natural environment, or on the effectiveness of PPT, can be further studied.

Conclusions
A plethora of studies have concentrated on the effects of PPT, yet few have empirically explored the determinants of the poor's intention to participate in PPT.Therefore, the focus of this article is to explain the poor's intention to participate in PPT based on the TPB.The study highlights the importance of endogenous impetuses for residents of remote poverty-stricken areas with harsh environmental conditions, such as Southern Xinjiang, to participate in state-aided projects.
Some practical implications arise from this study.First, governmental investment in tourism infrastructure is highly recommended for PPT development in remote povertystricken areas [68].This should not only help to improve the accessibility of such areas, but it also helps locals to form a positive perception towards local tourism and, hence, to boost their confidence in it.Second, training related to tourism participation should be provided to the poor.As is shown in this study, behavioral control is the most significant predictor of the poor's intention to participate in tourism.Once their participation capability is improved, they are more likely to become involved in PPT.Finally, as endogenous development needs to matter the most to the poor [61], the key question then is how to activate their awareness of self-development.As noted, the lack of self-development needs may be attributed to the remoteness and isolation of the unique environment, and the communication of the locals with the outside world may be enhanced.Local authorities, for instance, may organize locals to visit these pilot tourism villages in the developed parts of China to demonstrate how tourism can help to improve the socioeconomic lives of rural residents.
Finally, this study is not without limitations.This study only investigated the impact of three independent variables on the poor's intention to participate in PPT.Future studies may combine TPB with other psychological theories, such as field theory, to identify other important predictors of the participation intention [69].Studies may also explore the restrictive factors of the poor's participation intention, as the facilitators and inhibitors might be different.The survey was conducted through a convenience sampling approach.Potential bias might still exist, although the sample is a relatively large one and it generally represents the nature of the rural population in Southern Xinjiang.Moreover, given the uniqueness of Southern Xinjiang, caution should be applied when generalizing the findings to other poverty-stricken areas.

Figure 1 .
Figure 1.Conceptual framework (PPT: pro-poor tourism).Hypothesis 1 (H1).The attitude towards PPT has a positive impact on the intention to participate in PPT.Hypothesis 2 (H2).The residents' perception of support for PPT has a positive impact on their intention to participate in PPT.

Figure 2 .
Figure 2. Study Area 1A-5A: the classification standard of Chinese scenic spots; 5A is the highest level.

Figure 2 .
Figure 2. Study Area 1A-5A: the classification standard of Chinese scenic spots; 5A is the highest level.

Table 1 .
A summary of previous similar studies.

Table 4 .
Regression analysis of influencing factors of participation intention and robustness test.