Cross-Group Knowledge Transfer and Capacity-Building for the Poor : A Case Study of Tourism Targeted Poverty Alleviation in Guangdong Province , China

This paper mainly explored the strategies of capacity-building for the poor in the early stage of tourism development in poverty-stricken areas. Given the requirements of tourism poverty alleviation, the capacity-building strategies during the transition period need to focus on non-school-age and impoverished population, and integrate both explicit and tacit knowledge, as well as absorb foreign advanced knowledge adequately, while some limitations still exist in the literatures now. In this paper, we analyzed the cross-group knowledge transfer process of “migrant workers-local poor people” to explore the capacity-building strategies that meet various requirements during the transitional period for the poor.


Introduction
In the 1960s, scholars began to pay attention to the relationship between tourism development and poverty alleviation (Spenceley & Meyer, 2002).The research mainly focused on two issues-whether tourism development is conducive to poverty eradication, and how poverty can be eliminated through tourism development.After nearly half a century, the researchers have reached a consensus on the first issue.They believed that, despite there are still some arguments (Sinclair, 1991;Akama, 1999;Blake et al., 2008), tourism benefits for poverty reduction (Bennett, Roe, & Ashley, 1999;Scheyvens, 2002).In contrast, the debates on the second topic were heated.Studies showed that one of the main obstacles was scarce capacity of poor population (Ashley, Roe, & Goodwin, 2001;Liu & Wall, 2006).Therefore, researchers proposed various strategies to cope with it, such as basic education and tourism vocational education in poor areas (Mayaka, & Akama, 2007), public-welfare vocational training for the poor launched by government or NGO (Hummel & Duim, 2012), creating "learning by doing" opportunities by developing community tourism, alternative tourism and community empowerment (Sofield & Sofield, 2003).However, these approaches were limited in practice (Hausler, 2008;Spenceley, 2008), especially in the early stages of tourism development.
For impoverished region, several challenges faced to capacity-building for the poor in the early stages of tourism development: firstly, the majority of the poor who actually are involved in tourism development are non-school-age, besides many of the school-age children throw away their own chances for compulsory education initiatively (Truong, Hall, & Gary, 2014), which indicates that school education is a long-term strategy to enhance the abilities of poor people, nevertheless it cannot completely promote capacity of the poor at the beginning of tourism development; secondly, most of the skills required by tourism operation are tacit knowledge (Cooper, 2006), which is the foundation for the development of poverty-stricken people's capacity.However, the current vocational training is the lecture mainly lunched by government or non-governmental organizations, which features with low frequency and small-scale (Hummel & Duim, 2012); thirdly, comparing to operational skills, the poor population is deficient in knowledge of management and marketing (Spenceley, 2008), which need to be imported from foreign enterprises (Scheyvens, 2011a(Scheyvens, , 2011b) ) who, however, usually take the initiative to prevent advanced knowledge diffusing to the poor population for maintaining their technological superiority (Chok et al., 2007).In general, the poverty-stricken areas in the early stage of tourism development remain in transitional period of capacity-building for the poor, and starve for a capacity-building strategy for the non-school-age population and all the poor, which takes into account the spread and accumulation of tacit and explicit knowledge, effectively absorbs advanced knowledge from external markets or mainstream enterprises.This article intends to explore it.
We have visited several tourism poverty alleviation villages in Guangdong Province in the past two years and found that under the policy of "targeted poverty alleviation" (TPA), and found out that the capacity-building performance of the impoverished population has significantly improved.The strategy led by government and supported by finance, took impoverished families as the unit to establish a cross-group knowledge transfer mechanism with migrant workers for promoting the poor people's learning abilities, professional abilities and self-development abilities.Therefore, this research adopts a multi-case study approach to extract the general law of capacity building for the impoverished Advances in Applied Sociology population during the transitional period based on three representative villages.
In the next section, this paper summaries related theories to sort out the theoretical gap and identify research task.In the third section, this paper introduces the research design and methods through obtaining and processing data.In the fourth and fifth sections, this paper compares and discusses the cross-group knowledge transfer and the capacity-building process of impoverished population.In the last section, an integrated research framework, contributions, implications, limitations of this paper are summarized with future research direction proposed.

Logical Foundation for Tourism Poverty Alleviation: "Income
Poverty", "Power Poverty" and "Ability Poverty" The question that how to eradicate poverty through tourism has been one of the most controversial topics in academia all the time (Sofield & Sofield, 2003;Williams, 2004;Meyer, 2010aMeyer, , 2010b;;Scheyvens, 2011aScheyvens, , 2011b;;Mowforth & Munt, 2008).Based on the hypothesis of "income poverty", neoliberals who believe that the poverty is caused by the lack of opportunities to engage in non-agricultural industries and social exchange with the outside market (Spenceley & Meyer, 2012), advocate attracting foreign capital to create employment for poor population through "Trickle Down Effect" of tourism development (Friedman, 1992;Williams, 1998).However, this proposition triggered widespread doubts (Akama, 2004;Scheyvens, 2002Scheyvens, , 2007)).For example, Blake et al. (2008) and Wattanakuljarus & Coxhead (2008), based on the empirical studies in Brazil and Thailand, pointed out that tourism development reliefs, to a certain extent, absolute poverty, but it will lead to more remarkable relative poverty.
Post-constructivists introduce the method of political economy and put forward the hypothesis of power poverty (Zhao & Ritchie, 2007), which implying that "powerless" of poor population is the primary obstacle for tourism to reduce poverty (Pearce, et al., 1996;Sofield & Sofield, 2003).They advocate eliminating inequality and power poverty by structural reforms in tourism development (Britton, 1982;Wilkinson, 1987) and propose some countermeasures such as expanding community involvement (Murphy, 1985), empowering community (Akama, 1996) and developing community tourism (Scheyvens, 2002).Compared to the "income poverty" hypothesis, Spencyley and Meyer (2012) considered that "power poverty" was appropriate for the structure of social relations in tourism pro-poor practices, but existing efforts did not actually solve the problems of power disequilibrium (Macbeth, 1996;Jamal & Detz, 1995).
The plight of the "power poverty" hypothesis makes researchers think further.
McGregor (2005) pointed out that power is not given by others but created by ourselves.Ashley, Roe, & Goodwin (2001), found that insufficient capacity and human capital were the major reason of poverty based on comparative analysis of multiple cases in South Africa.Therefore, capacity-building is supposed to be

Research on the Capacity-Building: From the Perspective of Donor Agencies to Impoverished Population
In the 21 st century, the scholars have paid significant attention to the capacity-building of poor population (Spenceley, 2008;Hausler, 2008;Meyer, 2010aMeyer, , 2010b;;Scheyvens, 2011aScheyvens, , 2011b;;Koot, 2016).While studies centering with capacity-building were still insufficient (Liu & Wall, 2006), which embodies as following: Firstly, the roles and responsibilities of the government in capacity-building were not clear.As the capacity-building required long-term involvement, the duration for poverty alleviation usually exceed the project schedule set by government (Ashley, Roe, & Goodwin, 2001), which leads to related research findings were overlooked by donor agencies.Secondly, the core position of capacity-building has not yet been established.Manwa & Manwa (2014) proposed that related regulations or laws should be formulated to support capacity-building of poverty-stricken population.Thirdly, the research on the strategy of capacity-building mostly focus on the topic of "Creating Opportunity for Impoverished population", which meant that the capacity of the poor would be improved naturally as long as giving a chance.Obviously, it is not in line with practice (Truong, Hall, & Gary, 2014).
In general, the current research on poverty alleviation focuses on the analysis of policy input and output based on the perspectives of the donor agencies (Hummel & Duim, 2012), rather than process of capacity-building from the perspective of the local poor people.Chok et al., (2007) found that, even though providing opportunities, residents who engaged in basic work cannot develop their human capital and may not necessarily favor the higher-level work regarded as desirable by donor agencies (Tomaselli, 2017).It not only revealed the complicated psychology of impoverished population, but also showed that process of capacity-building needs to be explored in order to clear the concrete forms, connotations and technological paths of building capacity.

Knowledge Transfer in Tourism Poverty Alleviation:
"Business Perspectives", Individual Perspectives" and "Group Perspectives" The essence of abilities is a kind of superior knowledge embedded in behavioral and routines (Dutta et al., 2005), as a result, it is logical to analyze capacity-building based on knowledge management.For the poor, incapacity is rooted in insufficient knowledge and skills (Truong, Hall, & Garry, 2014), especially for business operations, product R&D and marketing (Spenceley, 2008), which cannot be generated in poor communities spontaneously rather than being acquired by collaborating with enterprises (Hausler, 2008;Scheyvens, 2011a, Advances in Applied Sociology 2011b).However, the cooperation between poor communities and foreign enterprises may extend their distance of socio-economic.Senior knowledge (such as tourism e-commerce technology) was usually monopolized by enterprises (Chok et al., 2007).As a result, the impoverished communities would be cornered in obtaining advanced knowledge.
From the perspective of knowledge management, the dilemma above was derived from knowledge transfer mechanism.Researchers were perplexed by monopolistic knowledge of enterprises due to concern the knowledge transfer of inter firms overly (Baggio, Cooper, Scott, & Laws, 2010).However, the capacity-building were influenced directly by knowledge transfer among individuals or groups.Individuals such as enterprise employees and tourists are important sources of knowledge, Nonetheless, the topic that how to integrate individual knowledge sharing, featured with transcend business boundaries and spread on different occasions, into cross-group knowledge transfer, characterized with intended, should be paid more attentions (Shawa & Williams, 2009).

Research Reviews and Gaps
Despite scholars began to focus on competency development of poor people, the research on capacity-building strategy in the transitional period remained gaps.
First of all, there was lack of microcosmic and longitudinal analysis of the capacity-building.Although knowledge management can be a suitable tool to explore the development of capabilities, the current research on it in tourism poverty alleviation was almost scarce.Secondly, the institutional and market environment for poverty alleviation were unique in China, which narrowed the applicability of exiting theories.Finally, the traditional stakeholder approach hindered the analysis of social relationships in capacity building and knowledge transfer for the poor.

Case Selection
Since April 2016, under the assistance of the Poverty Alleviation Office of Guangdong Province and governments of all localities, we have carried out the preliminary survey of 11 villages.After discussing, three typically villages-Likeng Village, Banling Village and Dongsheng Fishing Village-were selected as the research objects.The reasons were as follows: Firstly, the cases were typicalness.The three villages were all provincial impoverished villages.The income of villagers was much lower than that of other neighboring regions.Besides the tourism development was less than 8 years which means the tourism development was still in the early stage.Secondly, the cases were representativeness.they represented various modes that cross-group knowledge transfer and the capacity-building of the poor.According to the literature, the modes of knowledge transfer in tourism mainly included: foreign enterprise-led mode, SMEs clusters or learning regions mode, and communities-based mode (Shawa & Williams, 2009).Thirdly, the findings and analysis of complex social interactions were validity for we interviewed a number of stakeholders, and tracked them for a long time.

Data Collection
For triangulation of data (Yin, 1994), a variety of data collection methods were employed, containing semi-structured interviews, informal interviews, on-site observations, tourists survey and secondary data collection.
Interview was the main data collection methods of this study.We have conducted repeatedly field surveys on three villages, mainly including semi-structured interviews and informal interviews.All interviews were made on-site notes and sorted it out after interviewing.The forms of semi-structured interviews includes: 1) Villagers survey we investigated; 2) Focus group interview; 3) individual interview, including employees and managers in the foreign enterprise, local cadres and so on.Informal interviews conducted randomly at the village or enterprise took about 10 -15 minutes to discover the details of cross-group knowledge transfer activities, as well as the perceptions of villagers on their own ability.
In order to understand the interaction between migrant workers and the local in work, and to evaluate the capacity-building performance of the poor, we experienced various tourism facilities and services in the cases, as well as attended several internal work meetings.In addition, this article investigated the assessment of tourists to the ability of villagers.All the above data was recorded by the form of observation notes.
The second-hand information in this paper was composed of: 1) Government documents and reports including official documents, reports, implementation plans, management documents and work summary reports of all levels of government on (tourism) poverty alleviation.2) Enterprise records and graphic materials containing staff training plans, personnel management system, meeting minutes and bulletin board photos.3) Villagers graphic and entity information covering villagers study notes, articles of social media, such as Microblog and the Moment in Wechat.

Data Coding and Theoretical Analysis
In this paper, grounded theory method was applied for qualitative data analysis, the specific analysis are as follows: Advances in Applied Sociology Firstly, the initial coding analysis.Coding analysis consists of, referred to Strauss and Corbin (1990) three-level coding methods, open coding, axis coding and selective coding, along with writing theoretical memorandum simultaneously; Secondly, identified the attributes, dimensions and inter-relations of generics derived from the concepts and categories of the initial coding, and proposed initial theoretical assumptions; Thirdly, Extended scope of data collecting for completing the generic and attributes, and incorporating new concepts and categories emerging from the new data, so as to perfect the theoretical hypothetical; Lastly, compared qualitative data with theoretical assumptions continuously to reach saturation.After that, advanced a conceptual theoretical model by integrated coding results, theoretical memoranda and logical diagrams.At the beginning, the villagers knew nothing about tourism (big knowledge gap) and were hard to communicate with migrant workers (big cultural gap) so as to take some simple groundworks only and earned a fraction of their incomes (division of labor and wage gap significantly).They (migrant workers) were not friendly to the villagers (lack of trust).As a result many of villagers felt too stressful to quit.

Case Analysis
The government quickly recognized the above problems.
It is not enough to provide job opportunities but to train them to serve and manage.This is the only way for them to get rid of poverty (necessity for knowledge transfer and capacity-building).
At the same time, the governments were undergoing a "top-down" reform of the poverty alleviation work that converted village-based mode into family-based mode, aiming to ensure that each Poor family would benefit from poverty alleviation in tourism.On the one hand, the village cadres built information cards for all poor families.
Efforts are supposed to be made to ensure that each poor family will involve Advances in Applied Sociology In order to stimulate enterprises to cultivate villagers better, the government has secured special funds and favorable policies for the enterprises.The county leadership along with affiliations held meetings with enterprises to solve various problems such as land acquisition and tax relief.
Under the pressure of government, Yuntianhai Resorts gradually created an atmosphere that migrant workers are willing to help local workers with profession skills.One the one hand, various departments have formulated special training plans for villagers; on the other hand, the enterprise identified a list of key local staffs who would be promoted in case of good performance.These new native elites made other villagers to believe that promising future are expectable (demonstration effect of career), and they became the intermediary to smoothen the communication between villagers and managers (knowledge transfer intermediaries).
In addition, the enhancement of local staffs' capacity reinforced enterprises' confidence and willingness to foster the locals, leading to a virtuous circle of knowledge transfer efficiency.
Now we have benefited from local employment.In fact, as long as we can create chances for them to learn and make mistakes (to create a knowledge transfer environment), they are competent for many positions, even some managerial and technical jobs (results of capacity building).

Case B: Banling Village, Relying on SMEs Cluster
Banling Village is located at the junction of Shaoguan and Huizhou, Guangdong After we (the villagers) came to the Resorts, the managers were so busy to train us (lack of training programs).We mainly engaged in sanitation work In view of this, the government issued a series of measures successively: firstly, hire foreign experts to set standards of service quality and procedure for all agritainments, and carry out rounds of inspection.
We (the government) forced them to carry out trainings, and to improve the quality as well.If they don't meet the standards, we will turn them off.
(Urge of training mechanism establishment).
Secondly, fund impoverished villagers who possess related work experience to start small-scale agritainments.
If managers do not take the initiative to guide us, we need to imitate or consult them proactively (urge of active learning).After we are competent, the government would help us start our own business (the government incentive to active learning).
Thirdly, reward the enterprises that established integral training system or the enterprises in which local staffs perform well.
For star-rated agritainments, we (government) will pay cash bonuses and accord priority in promotion (government incentives), but the appraisal standard includes not only hardware, but also the quality of service and training, as well as the contributions to capacity-building and further to alleviate poverty for the impoverished population (incentive principle and guidance).
After nearly two years of TPA program launched, the capacity building work in Banling Village has gained progress gradually.Firstly, the opportunities for poor villagers to learn from migrant workers significantly increased.And it's more efficient to communicate with migrant workers.Advances in Applied Sociology Now they are very passionate and patient to guide us (enthusiastic guidance of migrant workers).We might become partners under the policy of subsidizing our own farm work (incentives for poverty alleviation); Secondly, the self-confidence and professional ability of poor villagers were improved significantly.
Many people in our village depend on agritainments for their livelihood, which is not actually difficult for us (low threshold of learning).The villagers who are dauntless and insightful would set up their own business (start their own businesses), while some others who are conservative would contract to supply food ingredients for the Resorts (Business outsourcing).
Even the weakest ones would work for the Resorts and master cooking skill or obtain a management position after a few years.
With active participation into the rural tourism, the income of villagers kept increasing as well.Despite fishermen were enthusiastic in participating into tourism, the facts were that they learned little and their income improved slightly.

Case
We know nothing about tourism business but driving the boat for travel agencies (less learning opportunities), while the remuneration is determined by them (price negotiation disadvantage).
The government attached great attentions to poverty alleviation and carried out investigation into every household and introduced various effective measures with the assistance of the village committee.First of all, technical officials and experts were organized to conduct home-based instructions and trainings.
We cannot rely solely on those travel agencies, we (government officials) can also teach them.For example the Tourism Bureau once guided villagers to hold local festival (festival guidance) and it turned out to be good.
Young villagers learned substantial marketing knowledge from government training as well.
We (villagers) print QR code now (sales skills knowledge) and hand them out to tourists.Besides, we update the moment of WeChat every day to promote ourselves.Tourists will contact us by WeChat (independent marketing capabilities), where we discuss how to plan the journey directly (Product planning ability).As the result, we earned more out of this mode than cooperating with travel agencies.
Secondly, the government provided law consultancy and market regulation to establish a stable relationship between villagers and reliable travel agencies.
The government introduced several travel agencies that are friendly to us (long-term cooperation).They not only hired us for sailing but also guided us to interact with visitors (knowledge transfer), and allocated some business to us for instance arranging for touring routes and preparing small gifts for tourists (division level to enhance).We obtained some skills that can be utilized for our own business (Enhancement of professional ability).
With the government's help and efforts, the fishermen have fostered a sense of tourism services and learned certain advanced knowledge for travel services and marketing.
Every family has a boat (or two) and we take QR code along always (tourism marketing awareness), at the same time, we learned how to serve tourists (tourism service capacity improvement).
However, generally speaking, knowledge transfer in neither government training mode nor travel agencies guide modes, which can be classified to community-based mode, is as efficient as of that developer-led and SMEs clusters modes.
After visiting fishing villages in other places, we (the government) found that our villagers were less knowledgeable and skillful than others (knowledge and skills gap), it was not enough to foster skills by means of government training since it features low frequency (less training time), or travel agencies for usually agencies contacted villagers too less (weak link) to guide theminconveniently (lack of opportunities for imitation).

Comparisons and Characteristics of Cross-Group Knowledge Transfer Modes
Based on case analyses above, this paper compares the characteristics of three cross-group knowledge transfer modes under different scenarios, which are shown in Table 1.
Table 1 illustrates the characteristics of three modes, while some common construction properties are concluded.Under the targeted poverty alleviation policies, the distribution of migrant workers and local impoverished population, as well as their demands and characteristics are identified systematically to establish a cross-group knowledge transfer system led by government, centered with two major groups, and supported by other stakeholders such as village committees, enterprises and local elites.On the one hand, the government defines and profiles impoverished population, figures out the classification of the distribution and features of migrant workers, and specifies knowledge recipients and absorber in knowledge transfer process; the other hand, it redefines the roles of other stakeholders from the perspective of knowledge transfer, so that individual knowledge sharing activities can be integrated into cross-group knowledge transfer.
Taking the impoverished families as a unit, each family combined with migrant workers to maintain the stability of the knowledge transfer, and to upgrade contents of knowledge transfer.As in Case A, the government offered opportunities for each family to work in Resorts on account of characteristics of the impoverished population, such as age, character and willingness.The jobs created by government and enterprises are beneficial for the villagers to get in touch with migrant workers frequently.As a result, it forms the mentoring pairs to eliminate discrimination (Beddee, 2004) and advance connection chances (Hansen, 1999).
The government, on the one hand, establishes the formal coordination networks from top-to down relying on village committees and enterprises; on the other hand, establishes the informal coordination networks from bottom to up depending on local elites.The two networks co-ordinate to narrow the knowledge and cultural gap between impoverished population and migrant workers (Cummings & Teng, 2003), which practically guarantees the effective operation of the knowledge transfer system.In consideration of social relationship between knowledge transfer participants is too complicated to be coordinated by single firm (Ashley, Roe, & Goodwin, 2001).Therefore the coordination mechanism that effectively integrates work relationship and subnetwork (Shawa & Williams, 2009) is necessary, just as the "government-enterprises-local elites" model in Cases A and B.
The government affords the transaction costs for the operation of cross-group knowledge transfer system through manpower, material, financial and social resources, and takes on most of the altruistic tasks to drive externally the knowledge transferring.The issues would arise if neglecting transaction cost of knowledge transfer (Caniels, 2000), such as migrant workers are reluctant to teach the villagers (Chok et al., 2007).Similarly, the impoverished population aren't willing to learn (Truong, Hall, & Garry, 2008).The practices in case A and case B,  (Spenceley, 2008), for this reason, the opportunities for trial and error learning and autonomous learning are crucial to capacity-building (Argyris, 1992).Under the circumstance of adequate opportunities, the self-development ability of the poor has improved significantly, which contributed to reduce the cost of knowledge transfer and the dependence on migrant workers.Eventually it converts knowledge transfer into a cycle of cumulative enhancement.In Case B, with the support of the government, the poor accessed to low-risk entrepreneurial opportunities which breeds a large number of small agritainments.The locals are more active than ever in learning from foreign managers.Besides, the content of learning was shifted from the operational knowledge to senior knowledge such as enterprise strategy and brand building.

Promotion of Cross-Group Knowledge Transfer to Capacity-Building of Impoverished Population
According to the case analyses, this paper elaborates the process of capacity-developing of impoverished population based on the upgrade of knowledge transfer, shown in Table 2: The cross-group knowledge transfer featured with large-scale, full-coverage and long-term, improves initially the learning ability of impoverished population.The frequent interaction between the two groups in workplace enhances directly the abilities of expression and communication of impoverished population, which is the origin to develop abstract thinking, eventually to make for absorbing new knowledge by trial and error learning (Darr & Kurtzberg, 2000).In Case A, the formal training contributed to acquire basic skills.However with the mentorship maturing gradually, the poor began to go beyond the allowed content of knowledge transfer to constantly explore new ways.The transition from "single-loop learning" to "double-loop learning" (Argyris, 1992) enabled impoverished population to develop self-learning ability.
In the wake of learning ability enhancing, the vocational competency of the impoverished population has been significantly improved.The enhancement of learning ability, on the one hand, helped the impoverished population to convert the knowledge directly into professional competency; on the other hand, it fostered abstract thinking ability conducing to handle problems through trial and error (Scheyvens, 1999).In Case B, the career of nearly one-third of impoverished population followed such a growth path, which means that the poor would like to challenge senior practices after reaching a goal with the continuous improvement of learning abilities.
The enhancement of vocational competency significantly improves the incomes of impoverished population and the game power of communities At the individual level, the improvement of vocational competency narrowed the division of labor and remuneration gap between the locals and migrant workers; at the community level, with the opportunities to cooperate decision-making increasing, community's game power has risen as well (Chok et al., 2007).In Case A, as numerous locals got promoted to be managers, enterprise trusted local employees and respected local communities more than before (Hansen, 1999).
When problems came up, for example, the contradiction of land acquisition, enterprise solved it by negotiating with the community rather than appealing to government intervention.
The improvement of the socio-economic conditions lifts the efficiency of cross-group knowledge transfer, which forms a virtuous cumulative cycle that is conducive to the self-development of impoverished population.With increase of income and narrowness of the social distance, the poor got the chances to participate in tourism economic activities in depth, where the knowledge transfer contents and deepness have been continuously boosted.At the same time, communities were more attractive for foreign workers to join in, so that the structure of knowledge transfer can be further optimized.The constant circulation between the knowledge transfer system and the capabilities-building Advances in Applied Sociology made it possible for impoverished population to acquire sufficient ability for self-development.

The Roles of Government in Cross-Group Knowledge Transfer and the Capacity-Building of the Poor
The question that what roles the government plays as in tourism poverty alleviation has been a hot shot in academic (Williams, 2004;Spenceley & Meyer, 2012).
The case study shows that most of altruistic works need to be organized, driven and promoted by the government for cross-group knowledge transfer in the early stage of tourism development.With the lifting progressively of knowledge transfer efficiency ability of impoverished population, government intervention began to weaken and market mechanism function kept rising, concluded in following statements: Responsible subject -the government takes the primary responsibility for promoting knowledge transfer and the capacity-building of impoverished population.The case indicated that the cross-group knowledge transfer system in impoverished communities contained a large number of contradictions, which were hard to resolve by other stakeholders due to lacks of competent or motivations, and government was the only subject who has responsibility, capacity and motive to maintain the development of the knowledge transfer system (Sofield & Sofield, 2003).
Coordinator The government acts as the organizer who facilitates the establishment of combinations between poor family and migrant workers, meanwhile provides institutional support for each bond.In Case A, the government offered the chances, which were stabilized by cooperation with the enterprises, for locals to obtain knowledge in work by means of "sand-mixed", so that they could be contact with migrant workers frequently (Darr, Argote, & Epple, 1995).Over time, multiple social networks, providing a platform for knowledge transfer, formed gradually between impoverished population and migrant workers (Baggio, Cooper, Scott, & Laws, 2010).In the process, the effective coordination of government was crucial to the evolution of the relationship.
Opportunity Providers -The government provides opportunities for the The knowledge sources should be compensated (Osterloh & Frey, 2000) by government because other stakeholders lacked capability or willingness to afford it.
In addition, the integration of diversified knowledge transfer activities was based on numerous public-welfare works rather than profitable ones.In the cases of this paper, altruistic works, such as funding for tourism poverty alleviation and importing human resources, were vital to promote knowledge transfer systems and build capacity of impoverished population.
Drivers -The government has driven a virtuous cycle of "cross-group knowledge transfer -capacity-building -improvement of socio-economic conditions", which makes communities develop independently.Government supports were regarded as an external driving force for the development of communities.In contrast the capacity-building of impoverished population actually need to be realized through their own efforts.In our cases, the social and economic conditions have been improved following the professionalization of the locals, which upgraded local tourism industry and knowledge transfer system.For example, during the construction of the resorts in case A, it renewed the migrant workers by introducing professional managers, as well as expanded knowledge transfer content.

Conclusion
In order to discover the mechanism of cross-group knowledge transfer and its impact on capacity-building of impoverished population, this paper proposes an integrated framework, as shown in Figure 1  Thirdly, the external driving force for the development of system mainly comes from the government's altruistic supports while the intrinsic impetus derives from the self-development ability of impoverished population.
Fourthly, the government plays the roles of the responsible subject of establishing and maintaining cross-group knowledge transfer system, as well as coordinator, opportunity creator, cost undertaker and driver.Government interventions would gradually fade away with the abilities of impoverished population increasing, and eventually be replaced by market mechanism.

Contribution and Implications
The purpose of the cases study is to develop new theories from empirical data to make up for existing theoretical gap, as well as further strengthen theory guidance on practices.The main contributions and implications of this paper are as follows: Firstly, in response to the debate on neo-liberalism or post-constructionism for relieving poverty, specifically whether the government should intervene or not (Williams, 2004;Sofield & Sofield, 2003), this paper puts forward that the government needs to be a leader in the transition period of capacity building from perspective of knowledge transfer.Only by resorting to government's al- Secondly, in response to the controversy over the "PPT" (Pro-poor tourism) theory proposed by DFID (Chok et al., 2007), the case study shows that the key of PPT theory lies in putting capacity-building for impoverished population on central position.The strategy of cross-group knowledge transfer proposed in this paper not only verifies the validity of PPT theory, but also provides the exercisable approaches for it.
Thirdly, the discussion on the poverty reduction effects of the "developer-leading" and "community-based" modes lasts for years.This research sug-

in
Yuntianhai by means of employment or cooperating with it.On the other hand, the government repeatedly negotiated with enterprises and urged them to keep stable employment proportion of impoverished population, employ them properly (enterprises duties of knowledge transfer) by target training and mentoring (approaches of knowledge transfer), offer them opportunities of management and marketing attempts (encouragement of learning and trying), and cultivate villagers as examples who are willing to guide the others and reward these examples (incentive for migrant workers).
province.The government built the "Feng Jiang Food Corridor" which gathered nearly one hundred rural folk restaurants along the Xinlong Road.Most of these restaurants are shared with foreigners and employ locals with full-time or part-time jobs.The scales of agritainments are various, however, in which managers, chefs and salesmen are foreigners in general (ways of foreign workers involvement).The enterprise is not as bureaucratic as high star-level hotel (low hierarchy) and locals get along with foreigners well, who usually rent villagers' house to live (living space intersection).Advances in Applied SociologyThe main goal of a SMEs clusters led by government is to alleviate poverty through local elites and foreign investors.Under the support of the County Poverty Alleviation Office and County Tourism Administration, most of the agritainments employed the poverty-stricken villagers recommended by government to counter labor shortage caused by employee turnover.But accordingly, some problems arose.
The government required the bosses to focus on the quality, training, and star rating.Then the boss required the managers (migrant workers) to give us lectures (formal training), and to transfer us skills or knowledge such as how to a la carte, how to serve the guests (on-the-job training), how to calculate gross profit, or how to check the inventory (senior knowledge transferring).
C: Dongsheng Fishing Village, Relying on Community-Based Tourism Dongsheng Village, located in Huizhou City, Guangdong Province, is an impoverished fishing village.Local fishermen participate in tourism by forms of sailing trip, fishing services, seafood sales.Most of these businesses are family-based since there is no tourism enterprise actually at present.Migrant workers on the community are mainly travel agencies and tour guides who cooperate with indigenous families.The tourism in Dongsheng village developed， spontaneously depending on families (community tourism), but the villagers cooperated with migrant workers such as bosses and tour guides who mainly in charge of marketing and teaching them related skills (spontaneous knowledge transfer).
where the government rewards the enterprises where the local perform well, and in case C, where the government provides technical support for villagers to develop new products and expand market, are examples of paying for transaction costs and taking on altruism work.Advances in Applied Sociology Government stimulates the enterprises to empower the local to participate in management, and encourages the local to update the content of knowledge transfer by engaging in challenging works to transform capacity-building into the internal driven force of knowledge transfer system.The low-level vocational training hinders the local to obtain senior skills

"
powerless" residents to acquire, absorbs and applies new knowledge and creates synergetic development mechanism for all stakeholders.Sharing knowledge means essentially exchanging power(Baggio, Cooper, Scott, & Laws, 2010), nevertheless impoverished locals were unlikely to develop capacity in virtue of knowledge transfer due to lack of equivalence.In case A and B, the government, on the one hand, built platforms for impoverished population with administrative practices; on the other hand, it established the mechanism of joint development by policy regulation, aiming to facilitate capacity-building to be the ground of incentives for all stakeholders.Cost undertaker -The government pays for the transaction cost and undertakes the most of altruistic works in organizing, driving, coordinating and controlling the capacity-building process of impoverished population.
Figure 1.Across-group knowledge transfer and capacity-building for impoverished population.
truistic works and coordination networks, could impoverished population get the chances to enhance capabilities.The government interventions in tourism development would reduce gradually after impoverished population are equipped with self-development ability, and ultimately give way to marketing advocated by neoliberalism.

Table 1 .
Comparison of cross-group knowledge transfer mode of tourism poverty alleviation.

Table 2 .
Knowledge transfer and development of impoverished population capacity.