determinants for csr in developing countries: the case of

1.　Introduction

　The worldwide production of palm oil has in-creased over the last few decades due to its ex-tensive use in food and biofuel products （Tres et al., 2011）. Oil palm, Elaeis guineensis, is the most popular oil crop in tropical regions due to its higher yield and cheaper cost compared to other crops （Carter et al., 2007 ; Henson, 2011）. The planting area of oil palm in Indonesia has expanded rapidly in recent years, and Indonesia has been the world’s largest oil palm producer since 2006 （FAO, year unknown）. Expansion of oil palm cultivation, however, has caused vari-ous problems. While attention on palm oil pro-duction has primarily focused on the environ-mental consequences of the land conversion caused by deforestation （Gillespie, 2011）, social

problems such as conflicts over land titles among palm oil companies, and local communi-ties and farmers are also commonly observed

（Marti, 2008 ; IRIN, 2010 ; Tiominar, 2011） even under the Nuclear Estate Smallholder （NES） scheme launched by the Indonesian government in 1977. The NES project is a kind of land alloca-tion agreement between small-scale farmers and plantations. See Appendix 1 for more detailed information about the NES scheme.　To reduce the risk of conflict with local com-munities, palm oil companies have increased their interest in Corporate Social Responsibility

（CSR）. Although Kemp （2001） concluded that the cultural, economic and political contexts fol-lowing the Asian Economic Crisis made it diffi-cult to consider the likelihood of CSR becoming popular in Indonesia, recent developments indi-

18

［Jpn. J. Rural Econ. Vol. 17, pp. 18─34, 2015］　

Determinants for CSR in Developing Countries :The Case of Indonesian Palm Oil Companies

Tomohide Sugino,1* Henny Mayrowani2 and Hiroaki Kobayashi3

Corporate social responsibility （CSR） is playing an increasingly important role in developing countries like Indonesia, where the palm oil industry, in particular, has caused public con-cern over its possible impacts on the environment. A land reallocation program called NES has been initiated by the Indonesian government with the aim of harmonizing the benefits of palm oil production between the palm oil processing companies and small-scale farmers. This is of particular interest as the CSR activities of the companies could provide an indica-tor of their attitudes toward a socially acceptable management style. This research aims to find empirical evidence to identify the driving forces behind the CSR activities carried out by palm oil companies in Indonesia. The results of the regression analysis performed on data from a questionnaire survey of large-scale palm oil companies concluded that not only the size of the company but also the performance of the NES scheme and the perception of relationships with local communities are major driving forces that stimulate participation in CSR activities.

Key words : oil palm, Indonesia, corporate social responsibility （CSR）, NES （nuclear estate smallholder） scheme, GAPKI （Indonesian palm oil association）

1 Japan International Research Center for Agricultural Sciences [email protected] Indonesian Center for Agricultural Socio Economic and Policy Studies [email protected] Chiba University [email protected]

cate encouraging signs （Sedyono, 2007）.　In spite of the above problems, palm oil com-panies in Indonesia have exhibited a degree of commitment to a variety of CSR activities. Such activities include not only publishing reports on environmental concerns and the performance of the company but also taking steps to benefit lo-cal communities effectively in a variety of as-pects, such as primary education, health im-

provement, and other social and economic interests. CSR could be defined as actions taken by companies to contribute to establishing envi-ronmentally-friendly, sustainable, transparent and responsible social communities rather than being concerned solely with prompt profit-mak-ing that goes beyond legal limits in some cases.1）　Table 1 shows the disclosure of CSR informa-tion on the websites of eight Indonesian palm oil

19Determinants for CSR in Developing Countries

Table 1. CSR disclosure on websites by palm oil companies

No. Information available on website Activities revealed on website

1Private

company

Disclosing detailed report of 137 pages.

Infrastructure, Education, Environ-ment, Community development

2 The annual report includes a CSR part of 21 pages.

Infrastructure, Culture, Environ-ment, Community development

3

State-owned company

The annual report includes a CSR part of 18 pages.

Infrastructure, Culture, Environ-ment, Community development

4 CSR information page but no contents available N. A.


Education, Culture, Environment, Community development


Environment, Community develop-ment, Health


Education, Culture, Health, Com-munity development

8 Disclosing detailed report of 116 pages.

Education, Environment, Health, Community development

Source : By authors based on information on websites of respective companies.

1）　Various definitions of CSR have been proposed in previous literature. The World Business Council for Sustain-able Development （WBCSD）（1999） defined CSR as “the continuing commitment by business to behave ethical-ly and contribute to economic development while improving the quality of life of the workforce and their fami-lies as well as of the local community and society at large.” McWilliams and Siegel （2001） defined CSR as “actions that appear to further some social good, beyond the interests of the firm and that which is required by law.” According to the new definition by the European Commission （2011）, CSR is “the responsibility of enter-prises for their impacts on society.” In the USA, CSR has traditionally been defined in terms of a philanthropic model, while the European model focuses more on operating the core business in a socially responsible way

（Baker, 2004）. Our definition is clearly based on the latter. CSR is very often seen as philanthropic acts in Indo-nesia, while some companies are trying to embed the societal issue in their core management process （Sedyono, 2007）. This indicates that CSR in Indonesia is most closely aligned with the North American definition although the Indonesian companies are gradually shifting to CSR more befitting the European definition. Another aspect of CSR is found in the United Nations Global Compact which places emphasis on human rights, labor, environ-ment and anti-corruption （United Nations, 2014）. However, the movement, launched in 2000, involves only a lim-ited number of public and private sector enterprises, amongst which are only 114 members from Indonesia as of 2014. While four of the member companies from Indonesia belong to the plantation sector, they are all paper manufacturing companies. If the number of Indonesian members shows a steady increase in the future, it would reveal that Indonesian companies implement CSR not only as philanthropic acts but also as a part of their ef-forts to conduct their overall business activities in a socially responsible way. This assumption can be made be-cause the movement can be recognized as an indicator of the willingness of private Indonesian companies to commit to sustainability and corporate citizenship.

companies. These companies are listed in the DJP （2011b）, which summarizes the CSR activi-ties implemented by major palm oil companies in Indonesia. Although all eight companies dis-close their CSR activities, there are some differ-ences in the extent and quality of the disclosed content. The table also indicates various CSR activities implemented by the palm oil compa-nies, covering infrastructure, education, culture, environment, health, and community develop-ment. However, those companies seem to be ex-ceptions since many companies do not disclose their CSR activities, making it impossible for us to determine which factors are important for motivating the palm oil companies to conduct CSR. See Appendix 2 for more information about CSR activities in the Indonesian palm oil industry.　The primary purpose of this study is to iden-tify the determinants of CSR activities in the In-donesian palm oil industry. In spite of the in-creasing interest in CSR, it is still difficult to access detailed information such as the number of companies implementing CSR and the compo-nents of their CSR activities. To address this shortcoming, we conducted a questionnaire sur-vey of the members of the Association of Palm Oil Companies in Indonesia, known locally as the Gabungan Pengusaha Kelapa Sawit

（GAPKI）. We then analyzed the results of the above survey and found empirical evidence to identify which of the companies’ determinants or characteristics enhance their CSR activities in this sector. Considering the importance of CSR as a practical tool to mitigate conflicts be-tween palm oil companies and local communi-ties, it is useful to know which factors motivate the companies to implement CSR.　In the following parts of this paper, we first review relevant literature and then present our questionnaire survey and its results. Some hy-potheses concerning the determinants of CSR activities and our analytical framework are then discussed, and the empirical evidence used to identify the determinants of CSR activities is re-vealed. We summarize this paper in the final part.

2.　Literature Review

　The CSR movement, particularly in develop-ing countries, is still in its early stages. While the volume of literature on the subject has in-creased recently, a lot of information, knowledge

and analytical work still need to be developed and accumulated.　Considerations from economic theories char-acterize CSR as the provision of public good by the private sector （Kitzmueller and Shimshak, 2012 ; Blackman, 2010 ; Besley and Ghatak, 2007）, and in this sense, CSR could be analogous with environmental damage abatement. In one of the pioneering studies which analyzed the factors determining companies’ behaviors from this viewpoint, Pargal and Wheeler （1996） evaluated how communities’ informal regulation enhanced the voluntary provision of public goods by pri-vate companies. The Indonesian case study con-ducted by Pargal and Wheeler （1996） estimated the demand function of environmental services by firms and identified the importance of the companies’ scale, type of ownership, income and population density in the region, among other factors.　After extensively reviewing 186 empirical studies on the determinants of CSR reporting, Fifka （2013） pointed out that most of the stud-ies originated from Anglo-Saxon and Western European countries. Aside from twenty-one studies with transcontinental perspectives, there were only fifteen articles that reviewed the Southeast Asian context. Shauki （2011） con-ducted a survey of various stakeholders, includ-ing the Indonesian companies directly involved in CSR, and identified some of the determinants of CSR disclosure. The studies of both Fifka

（2013） and Shauki （2011） focused only on the publication or disclosure of reports, whereas CSR actually includes a wide range of activities intended to cope with the requirements of ev-ery part of society as described above.　Fifka （2013） also suggests that the variables which are harder to quantify play important roles in the investigation of CSR. One of the lim-itations of empirical work in this field may be the lack of context-relevant data, which means that specific surveys or interviews will be re-quired. In their assessment of the social and en-vironmental aspects of CSR in 15 Asian coun-tries, including Indonesia, based on a survey of the chief executives and senior management of relevant companies, Baughn et al. （2007） re-vealed strong relationships between CSR and economic, political and social contexts such as per capita GDP, political freedom and corruption in the respective countries. According to Kitz-mueller and Shimshack （2012）, only limited evi-

20

dence is found to support the various hypothe-ses that labor markets and shareholder preferences drive CSR, but there is consistent empirical evidence that consumer markets, pri-vate politics, and public politics affect CSR. The OECD （2005） indicates that there are striking differences between countries even within the same region as CSR has become increasingly common in emerging markets.　Environmental management is one of the ma-jor concerns in CSR activities, and many studies focus on this issue in developing regions. Kathu-ria （2007） reported from the Indian setting that informal regulation represented by press re-ports about pollution works effectively as a reg-ulator if there is sustained interest in news about pollution. Hettige et al. （1996） found that local income is a predictor of the effectiveness of informal regulation and that state ownership of a company positively affects pollution control in developing countries. Arli and Lasmono

（2010） found that consumers in developing countries are often unaware of and unsupport-ive toward CSR, unlike in developed countries.　From the literature review, we have found that the lack of context-relevant data is a limita-tion of empirical work on CSR in developing countries. This finding consequently justifies the approach of our study, which focuses on the palm oil industry in Indonesia and depends on primary data collected through a questionnaire survey.

3.　Methods of the Survey

　The participants of the survey were members of the Indonesian Palm Oil Association

（GAPKI）, which is the sole association of Indo-nesian palm oil producers. As of the time of the survey, GAPKI consisted of 569 mostly large-scale companies, while there was a total of 1,151 large-scale estate companies in Indonesia in 2009 （BPS, 2010 : p. 241）. These companies culti-vate estate crops on state-owned land under a concession, named Hak Guna Usaha （Cultivation Rights Title）, granted by the government （BPS, 2008 : p. 173）.　The questionnaire comprises four sections : 1） profile of the company and respondent person-nel, 2） NES scheme, 3） CSR activities, and 4） relationships with local communities around the plantation. The sections relating to the NES scheme and CSR activities comprise nine and sixteen components, respectively, as shown in

the following section. The NES scheme and CSR activity literature of Indonesian palm oil compa-nies （DJP, 1992, 2011b ; GAPKI, 2009 ; Jelsma et al., 2009 ; Zen et al., 2005） provided the basis for the initial selection of the components, which were then finalized through discussions with the Indonesian experts who collaborated in the survey. Some of the components are common to both the NES and CSR, such as technical sup-port and purchasing products. To avoid confu-sion in differentiating NES from CSR, we in-formed the respondents that any components which focused on the plasma farmers, i.e., small-land holders, should be classified as NES compo-nents, while any which focused on other farm-ers outside of the plantation should be classified as CSR components.　As for the relationships with the local com-munities, the respondents were requested to evaluate the current situation on a five-level Likert scale, i.e., “very good”, “good”, “neither good nor poor”, “slightly poor” and “very poor”, corresponding to 5, 4, 3, 2 and 1, respectively. The respondents were also requested to evalu-ate the importance of ten kinds of activities and concepts with which companies would contrib-ute to establishing better relationships with lo-cal communities on a five-level Likert scale, i.e., “very important”, “somewhat important”, “nei-ther important nor unimportant”, “not very im-portant” and “not important at all”, correspond-ing to 5, 4, 3, 2 and 1, respectively. These activities and concepts were selected based on the principles and criteria indicated in various guidelines for sustainable palm oil production certification, such as those of the RSPO （2007） and DJP （2011a）. For example, principal 1 of the RSPO （2007） is oil palm growers’ commit-ment to transparency. Therefore, we selected “Transparency of plantation management” as one of the concepts to be addressed in the ques-tionnaire. If a company has more than one es-tate, the respondent was requested to answer about the entire company, not about a specific estate.　The survey was carried out between July 2012 and March 2013. First, the authors commu-nicated with all of the GAPKI member compa-nies by phone to identify the relevant respon-dents in each of the companies. Most of the selected respondents were members of the CSR, legal, and management sections. Next, we sent questionnaires to the 366 companies which gave


us their consent to participate in the study. The questionnaires were sent by post mail, email or facsimile, and the companies were requested to return the answer sheets after they completed the questionnaires. By the end of March 2013, a total of 132 companies had replied.

4.　Results of the Survey

1）　Farmland and FFB production　The size of the landholdings of the companies participating in this study is shown in Table 2. On average, the total area of each plantation is 20.2 thousand hectares, of which 14.6 thousand hectares belong to the nuclear estate and 5.8 thousand hectares to the plasma farms.2） This is much larger than the average land size of the country’s total of 1,151 large-scale estate compa-nies, which is 3.9 thousand hectares （BPS, 2010）. The sample, therefore, clearly has an up-ward bias in the companies’ scale as measured by farmland area.

2）　NES scheme　Representing 69％ of the total sample, 91 par-ticipating companies stated that they are cur-rently implementing the NES scheme. Among

these 91 companies, the share of land allocated to plasma farms is only 32％ on average and less than 40％ in most cases （Table 3）. This re-sult indicates a continuation of the trend toward a reduction in the area of land allocated to plas-ma farmers （see Appendix 1）.　With the aim of evaluating the efforts of the respondent companies to cope with the require-

22

Table 2. Characteristics of the respondent companies （n＝132）

Nuclear estate

Plasma farm Total

Farmland area （hectares）（Number of respondents）＜1,000 12 62 91,000─10,000 79 46 6910,000＝＜ 39 22 54

（hectares）Average 14,617 5,777 20,154 Minimum 0 0 0Maximum 114,434 63,021 150,376

Annual FFB production （metric tons）（Number of respondents）＜10,000 37 82 3310,000─100,000 52 23 44100,000＝＜ 38 20 50

（metric tons）Average 170,513 66,750 234,752 Minimum 0 0 0Maximum 2,227,347 935,948 2,742,158

2）　Under the NES scheme, the plantation consists of two parts : one area named nuclear estates, which are di-rectly managed by the company, and a separate area planted by smallholders （plasma farmers）, which are named plasma farms. Usually a 2-hectare plot of plasma farm is allocated to each plasma farmer （DJP, 1992）.

Table 3. Number of companies by the percentage of land allocated to plasma farmers （n＝132）

Share （％） No. of companies

0 410＜, ＜20 4120＝＜, ＜40 2040＝＜, ＜60 1460＝＜, ＜80 1180＝＜, ＝＜100 5Total 132

Note : 41 companies not implementing the NES scheme are counted as hav-ing a plasma land share of 0％ .

ments of the basic procedures of the NES scheme, we asked whether or not they are cur-rently implementing each of the nine programs under the NES scheme. “Technical support for crop production”, “Purchasing products at a guar-anteed price” and “Infrastructure development” are the most popular （Table 4）. A total of 56％out of the 91 companies which are implement-ing the NES scheme answered “yes” to at least

these three programs. This is an unexpected re-sult, because it is essential that these three pro-grams be implemented whenever the NES scheme is launched （see DJP, 1992）. However, nearly half of the respondents whose companies are implementing the NES scheme do not imple-ment at least one of these essential programs.

3）　CSR activities　Similarly to the case of the NES scheme, we


Table 4. Implementation of nine programs in the NES scheme （n＝91）

Components of the NES scheme No. of companies （％）

NES1 Technical support for crop production 68 （74.7）NES2 Purchasing products from plasma farmers at a guaranteed price 67 （73.6）NES3 Infrastructure development （roads, irrigation facilities, etc.） 65 （71.4）NES4 Lending money for crop production 47 （51.6）NES5 Providing seeds, fertilizers and other materials as market price 43 （47.3）NES6 Use of facilities in the company free of charge or at discounted price 16 （17.6）NES7 Providing seeds, fertilizers and other materials free of charge or at discounted price 16 （17.6）NES8 Purchasing products from plasma farmers at market price 12 （13.2）NES9 Land lease to plasma farmers 0 （0.0）

Note : Percentages of the 91 respondents implementing the NES scheme.

Table 5. Number of companies devoted to CSR activities （n＝132）

CSR activity No. of companies （％）

CSR1 Infrastructure development which can be used by communities （schools, hospi-tals, etc.） 113 （85.6）

CSR2 Education support for local communities （training, scholarship, etc.） 89 （67.4）CSR3 Existence of a unit or special staff for CSR in the company 86 （65.2）CSR4 Environmental conservation （planting trees, cleaning, etc.） 76 （57.6）CSR5 Meetings with local communities to exchange opinions about the company’s ac-

tivities 74 （56.1）

CSR6 Provision of health improvement programas for local communities 68 （51.5）CSR7 Publication of CSR report 51 （38.6）CSR8 Purchase of products from nearby farmers （except plasma farmers） at a guar-

anteed price 50 （37.9）

CSR9 Technical support for crop production for nearby farmers （other than plasma farmers） 40 （30.3）

CSR10 Promotion of local traditional cultures 37 （28.0）CSR11 Financial support to underprivileged households 31 （23.5）CSR12 Monetary loans for crop production for nearby farmers （other than plasma

farmers） 19 （14.4）

CSR13 Provision of seeds, fertilizers and other materials free of charge or at discount-ed prices 19 （14.4）

CSR14 Use of the facilities in the company free of charge or at discounted prices for nearby farmers （other than plasma farmers） 18 （13.6）

CSR15 Purchase of products from nearby farmers （other than plasma farmers） at the market price 13 （9.8）

CSR16 Land lease to nearby farmers （other than plasma farmers） 3 （2.3）

asked the respondent companies whether they are familiar with the phrase “CSR”, whether they are implementing any CSR activities, and whether or not they are implementing 16 specif-ic CSR activities. Only one of the respondents had not heard of the expression “CSR”, and all but two respondents indicated that their compa-nies are implementing CSR activities. Although Law No. 40/2007 on Limited Liability Companies does not define the scope of the natural re-source sectors in which companies are obliged to practice CSR, our results confirm that CSR has already become a common practice among the companies participating in this study. Out of the 16 kinds of CSR activities that we specified, “Infrastructure development” is the most popular, being implemented by 113 companies （Table 5）.　The other activities implemented very often are “Education support” （89 companies）, “Exis-tence of CSR staff” （86 companies） and “Envi-ronmental conservation” （76 companies）. Exam-ples of education support include the establish-ment of schools entirely or partly managed by the companies as well as the issuing of scholar-ships （PT Astra Agro Lestari, 2012）. Nearly two thirds of the respondents have CSR staff in their organizations. However, the number of re-spondents who published CSR reports is only 51, or 39％ of the respondents.3）

　Out of the 78 respondents who provided an-swers as to when their companies commenced CSR activities, more than half （41 respondents） began within the previous 10 years （2004─2013）, while 22 respondents began in 1994─2003, and 15 respondents began before 1993 （Table 6）. Among the 41 companies which started CSR ac-tivities after 2004, only two companies were newly established after 2004. This means that not only newly established companies but also old companies started CSR activities within the last decade. Our survey clearly shows that CSR activities have expanded significantly among the respondent companies over the last decade.

4）　Relationships with local communities　Most respondents evaluate their current rela-tionships with local communities as 5 （very good） or 4 （good）, although one respondent an-swered with 2 （slightly poor）（Table 7）. These

results might contradict the general perception that Indonesian palm oil companies are facing various conflicts with local communities. The Sawit Watch, one of the Indonesian NGOs moni-toring large-scale plantations, has recognized 632 cases of such conflicts between local com-munities4） and palm oil companies since 2006

（IRIN, 2010）. The Sawit Watch reported that the communities have serious conflicts with the companies mostly because of a lack of consulta-tion between the companies and the communi-ties.　It is very possible that the answers to this question have some bias because companies fac-ing serious conflicts with local communities might have tended not to indicate such prob-lems when participating in our survey. More-over, companies with serious conflicts might have chosen not to respond to the survey. The Sawit Watch did not disclose the names of the companies which were involved in the reported conflicts. However, one company, which was re-ferred to in IRIN （2010） as having a conflict with local farmers, chose not to respond to our survey even though the company is a member of GAPKI.　Among the ten kinds of activities and con-cepts for establishing better relationships with local communities, “Communication with local communities” and “Obeying laws and regula-tions” earned the highest scores （4.9）, followed by “Fairness of plantation management” and “Re-spect for local land rights” （4.7）（Table 8）. The results indicate that most of the companies place importance on many different aspects. It is very encouraging that the companies give high priority to compliance and land right enti-tlement. Law enforcement in Indonesia is some-times still weak and compliance is an important issue in corporate management. For example, an environmental impact assessment （EIA） must be carried out before the land entitlement can be given to plantation companies. However, the EIA process is rarely observed （Marti, 2008 : p. 8─9）. Land rights are the most serious source of conflict in the plantation business in Indonesia （Marti, 2008 : p. 10）.

24

3）　Chambers et al. （2003） also found that 72％ of the CSR reports made public in Indonesia consisted of only one or two pages.

4）　Local communities in this context mean farmers or residents who live or cultivate land in the areas surround-ing oil palm estates and who suffer negative impacts from the estates.

5.　Determinants of CSR Activities

1）　 Hypothesis and model for empirical analysis

　Fifka （2013） reviewed 186 empirical studies which examined the determinants of CSR re-porting and concluded that 87％ of the studies found a positive correlation between company size and the tendency toward CSR reporting. The company size could be a proxy for several factors enhancing CSR. For instance, larger companies are, generally speaking, more capable of managing various activities ; they tend to face more criticism and monitoring ; and they have greater concern for their reputations.5） Although

the studies reviewed by Fifka （2013） focused only on CSR reporting, we can reasonably as-sume that company size will also influence other CSR activities.　However, the discussions shown in the previ-ous sections of this paper suggest several other aspects concerning CSR activities and their de-terminants in the Indonesian palm oil industry. As mentioned in Section 3, NES programs and CSR activities have very similar components. Some of the CSR activities, such as purchasing products at a guaranteed price, are common practice under the NES scheme. It is relatively easy for the companies to expand the scope of these practices to farmers other than the plas-ma farmers. We can, therefore, assume that the NES scheme will positively influence CSR activi-ties.6）

　In addition, our survey evaluated companies’ attitudes toward relationships with local com-munities by asking what the respondents think of ten kinds of activities and concepts which contribute to establishing better relationships with local communities. If palm oil companies have better awareness of their relationships


5）　When considering the external reputations, we should be very careful about the problem of endogeneity, be-cause such reputations could be the outcomes of the past activities including the CSR of the company.

6）　Here we face the problem of endogeneity because some specific characteristics or conditions of the company will enhance both CSR and implementation of the NES scheme. In our case, however, the government program of the NES was implemented in 1977. The exact year of the specific NES launch of each individual company is not available in our survey. However, we use the year of each company’s establishment as a proxy, because NES implementation has been a condition for permitting oil palm plantation businesses to operate since the gov-ernment launched the NES program in 1977 （Zen et al., 2005）. Among the 68 companies implementing NES and for which we have information about the year of the companies’ establishment, 44 companies were established during 1980─1999. Therefore, we can recognize that most projects under the NES scheme were launched during the 1980s and the 1990s. We could assume the above causality because most CSR activities emerged after the 2000s. The problem of endogeneity of this kind is extensively investigated by Blackman （2010） in cases of pollu-tion control.

Table 6. Starting year of CSR activities and year of company’s establishment （n＝78）

Starting year of CSR

Year of company’s establishment

Number of companies

2004─ 2004─ 21994─2003 15─1993 14Unknown 10

Sub total 41

1994─2003 1994─2003 7─1993 7Unknown 8Sub total 22

─1993 ─1993 6Unknown 9Sub total 15

Total 78

Table 7. Evaluation by companies regarding their relationships with local communities （n＝128）

Evaluation No. of companies

Very good 38Good 89Neither good nor poor 0Slightly poor 1Very poor 0Total 128

with local communities, we can assume that they will engage in more CSR activities. The causality of the relationship with local communi-ties and CSR may not be one way. Greeting and Turban （2000） reported that CSR can increase a company’s reputation thus enabling it to at-tract a better quality workforce, and that such personnel tend to have a better perception of the company’s relationship with local communi-ties. However, since most respondents only started their CSR activities within the previous decade, as shown in Table 6, it may be too early for CSR activities to influence employee charac-ters.　In conclusion we propose the following three hypotheses in order to identify the determinants of the CSR activities of palm oil companies based on our survey. We can expect that :　H1 : The size of the palm oil companies will

have a positive impact on CSR activities,　H2 : NES performance will have a positive im-

pact on CSR activities, and　H3 : The companies which think a lot of their

relationships with local communities tend to be more devoted to CSR activities.

　Taking the above hypotheses into consider-ation and assuming other factors suggested in previous literature, we can use the following equation to determine the CSR activities of palm oil companies.7）

Y＝f（X1, X2, X3, PS, RF, PD, RG）　The first three explanatory variables corre-spond to the above hypotheses H1─H3. We gen-erate composite variables for X2 and X3 by prin-cipal component analysis （PCA）. The last four items act as controls to relieve possible bias of

the estimation and verify the above hypotheses. The last two of these represent some external pressures. Several other factors could be consid-ered as explanatory variables, including mem-bership of the Round Table on Sustainable Palm Oil （RSPO）, which, although important, is not included in our regression model. While RSPO membership implies a company’s stronger per-ception of the importance of CSR activities as well as pollution control, this variable must be endogenous. As of now, we have also not found any proper data for variables representing some aspects of the supply chain. Definitions and data sources of all variables are as follows :　Y: The degree of CSR activities evaluated as

the number of ‘yes’ answers to the compo-nents as shown in Table 5. We excluded “infrastructure development” because this component is difficult to distinguish from the NES scheme.8）

　X1 : The size of the company evaluated by FFB production in metric tons.

　X2 : A composite variable representing the company’s efforts in the NES scheme.

　X3 : A composite variable representing the ex-tent to which the company thinks each activity/concept will be effective in estab-lishing better relationships with local com-munities.

　PS : The share of plasma farmers in terms of farmland area in the company.9） Higher shares imply that the NES scheme has a longer history, because of the reason ex-plained above.

　RF: Annual rainfall （mm） from BPS （year unknown） in the province where the es-tate is located. If the companies own es-

（1）

26

7）　The specification of the equation （1） will naturally lead to some regression analysis. We are conditioned by strong assumptions in terms of causalities among the variables generated from the questionnaire survey, which are qualitative in many cases. A more flexible method, such as structural equation modeling （SEM）, might be an alternative framework to consider the entire picture of mutual relationships among all variables and to avoid the problem of endogeneity pointed out in other parts of this paper. Such a study would be a valuable project for the future after the collection of more information, while our interest at this time focuses on CSR activities in various aspects and their determinants.

8）　NES and CSR have several similar components. We advised the respondents in the questionnaire that if the activities focused on plasma farmers, these activities should be classified as NES. We believe that whether these activities belong to NES or CSR is clearly distinguished based on the information regarding targets or beneficia-ries. As for the infrastructure development, we mentioned in the questionnaire that roads and irrigation facili-ties are examples of NES infrastructure development while schools and hospitals are examples of CSR infra-structure development. However, some infrastructure developments might be difficult to categorize under NES or CSR. We excluded the responses about infrastructural development from the further analysis in order to avoid the issue of endogeneity.

9）　We cannot single out the shares of independent farmers from our survey results.

tates in more than one province, we use the data of the province which consti-tutes the majority of the estate. The same procedure is applied to PD and RG below.

　PD: Population density in the province from BPS （year unknown）. It is anticipated that population density positively affects the degree of CSR activities because higher population density increases the possibilities of conflict with local commu-nities so companies consequently have more motivation to implement CSR activ-ities.

　RG: Regional Gross Domestic Product per capita in the province from BPS （year unknown）. It is anticipated that regional GDP positively affects the degree and level of CSR activities because companies should pay higher costs for their poor performance and bad reputation in richer areas. Pargal and Wheeler （1996） inves-tigated the effects of informal regulations on industrial pollution and found that community income has a negative rela-tionship with emissions from the plants. This variable is often applied in similar empirical studies, among which Welford

（2005） identified the importance of eco-nomic development and the supply chain aspects of CSR. Demand from community members for CSR and other socially re-sponsible actions by the company will become more significant in accordance with their increase in income.

　We omitted the samples which do not have

FFB production data and also the samples which did not answer all the questions about the importance of the ten components for better relationships with local communities. The sam-ple size of the analysis below was subsequently reduced to 115. The RF, PD and RG data of 115 respondents includes 13 provinces with the means and （standard variations） at 2,663 mm

（Standard variation＝304 mm）, 89.5 person/km2 （Standard variation＝124.4 person/km2） and 47.4 Million Rupiah/capita （Standard variation＝38.1 Million Rupiah/capita）, respectively.　Regarding X2, we first applied PCA for seven variables of the programs in the NES scheme as shown in Table 4 （NES1, 2, 4─8）, and then generat-ed a composite variable as the score of each sample on the first principal component. No re-spondent replied ‘yes’ to NES9, ‘Land lease to plasma farmers’. NES3, “infrastructure develop-ment”, was excluded because of the difficulty in distinguishing it from similar CSR activities. Be-cause, in the estimation by PCA, we excluded the samples which did not implement the NES scheme, we have zero for X2 in cases where the companies have not implemented the NES scheme at all.　As in the case of X3, we applied PCA for ten variables of component （COM1─10 in Table 8） re-garding the activities/concepts of the companies for contributing to establishing better relation-ships with local communities, and then generat-ed a composite variable X3 as the score of each sample on the first primary component. We used the SPSS 17.0 program in our statistical analysis.


Table 8. Scores on importance of activity/concept in establishing better relationships with local community （n＝132）

Activity/concept Average score

Number of respondents

COM1 Communication with local communities 4.9 132COM2 Compliance with laws and regulations 4.9 132COM3 Fairness of plantation management 4.7 131COM4 Respect for local land rights 4.7 132COM5 Economic benefit to local communities 4.6 132COM6 Environmental conservation in communities 4.5 131COM7 Infrastructure development for local communities 4.4 132COM8 Transparency of plantation management 4.3 122COM9 Ensuring of food availability in communities 3.9 129COM10 Prioritization of benefit to communities ahead of benefit to the company 3.5 126

Note : Average scores are calculated using a five-level Likert-scale as explained in the text.

2）　Results of the estimation　The estimation results of PCAs regarding X2, the NES scheme, and X3, perceptions of the re-lationships with local communities, are summa-rized in Tables 9 and 10, respectively. In both cases, three principal components have eigen-values exceeding unity, and the first principal components have positive correlation, or energy contents, on all of the variables involved. We can recognize that the first principal component in each case of estimation represents the overall efforts in the NES scheme and the overall evalu-ation of the relationships with local communities, respectively.10, 11） Here we have the data for X2 and X3 generated from these estimations by the abovementioned procedures.　The results of the regression are shown in Table 11. First we estimated the full model of the equation （1）, and then tried variations by omitting some of the explanatory variables.

Variance inflation factors （VIF） do not suggest a problem of multicollinearity.12）

　The full model of equation （1）, Model 1 in Ta-ble 11, is effective for determining the level and degree of CSR activities by the companies at a 1％ level of statistical significance.　First, X1, company size, is important in influ-encing the degree of CSR activities at a 1％ sig-nificance level. We can, therefore, conclude that hypothesis H1 is supported overall in line with most of the literature reviewed in Fifka’s study

（2013）. Second, the partial regression coefficient on X2 is positive and significant at a 1％ signifi-cance level. Consequently, we can conclude that the NES scheme has positive impacts on CSR activities. Therefore, hypothesis H2 is strongly supported.　Third, X3, perceptions of the relationships with local communities, is important in influenc-ing the degree of CSR activities at a 1％ signifi-

28

10）　To determine the internal consistency of the relationships between individual components, Cronbach’s alpha was calculated to confirm the reliability among individual components in each generated variable. The reliability of the constructs was acceptable, with Cronbach’s alpha results higher than the 0.7 threshold （Nunnally, 1978）.

11）　We might be able to interpret other principal components as follows. For X2 on NES effort, “Business as usu-al” to the second principal components was mainly suggested by the loadings on NES5 and NES8, and “Tendency to support farmers” to the third principal component was suggested by the loadings on NES6 and NES7. For X3 on relationships with communities, “Societal harmonization” to the second principal component was suggested by the loadings on COM2 and COM3, and “Harmonization with local society” was suggested by the loadings on COM1, COM5 and COM9.

12）　Since province-level variables （RF, PD and RG） may be highly correlated, we also added these variables one by one into the equation （1）. No significant difference with the regression results in Table 11 was found re-garding partial correlation coefficients for RF, PD and RG. We also used 13 province dummies instead of those province-level variables （RF, PD and RG）. No significant difference with the regression results in Table 11 was found regarding partial correlation coefficients for X1, X2 and X3.

Table 9. Principal component loadings on seven NES variables

Principal components

First（Overall NES efforts）

Second（Business as usual）

Third（Tendency to

support farmers）

NES1 Technical support 0.8058 －0.0303 －0.0737 NES2 Purchase of products （guaranteed price） 0.8016 －0.2120 －0.2113 NES4 Monetary loans 0.7795 0.1225 －0.1000 NES5 Provision of inputs （market price） 0.5699 0.2986 －0.3304 NES6 Use of the facilities 0.2242 －0.6954 0.4263 NES7 Provision of inputs （discounted price） 0.4703 0.0021 0.7510 NES8 Purchase of products （market price） 0.1038 0.8052 0.4456

Note : We don’t use samples missing FFB production data and samples of imperfect response about relationships with local communities, either. The sample size of the analysis is subsequently reduced to 115 samples. Three proncipal components have eigenvalues exceeding unity. Cumulative energy contents up to the three principal components are 36％, 54％ and 70％ .

cance level. Hypothesis H3 is, therefore, also supported. The more companies think about es-tablishing better relationships with local com-munities, the more diversified are the CSR ac-tivities they implement, which subsequently results in greater benefits in various aspects for the local communities. As is soon clear, we should be very careful regarding the problem of endogeneity in this context of regression analy-sis. According to Hsu and Cheng （2012）, the characteristics of the company or the manager, such as personal values, morality and corporate culture, play a key role in promoting CSR. While the observations of Hsu and Cheng （2012） are limited to small and medium enterprises in Tai-wan, it would be reasonable to assume similar causality in our case of Indonesian palm oil com-panies.　To deal with this problem, we presented three more estimations in Table 11 in which we excluded X3 （Model 2）, X2 （Model 3）, and both

X2 and X3 （Model 4）, respectively. We did not find any significant difference between Model 1 and the other models on the partial correlation coefficients. In terms of Akaike’s Information Criteria （AIC）, Model 1 best fits the data.13）

　Fourth, the partial regression coefficients of PS, the share of plasma farmers ; RF, rainfall ; PD, population density ; and RG, Regional Gross Domestic Product per capita, are insignificant. The partial regression coefficient of PS is posi-tive, which shows a tendency of positively af-fecting CSR. Unlike those in the research con-ducted by Pargal and Wheeler （1996）, PD and RG in the present study do not have a tendency to positively affect CSR or even negatively af-fect CSR, although the estimates are not statisti-cally significant. A possible reason would be the geographical condition of the palm oil industry. Compared with the general manufacturing in-dustries which were the focus of Pargal and Wheeler’s study （1996）, palm oil factories are


13）　We tried several other specifications to confirm the robustness of the regression results. One included “Esti-mated number of years since the companies started the NES scheme （NY）” in equation （1）. Among the 115 samples used for regression analysis by equation （1）, we have only 56 samples where NY data is available. Al-though the sample sizes are different, no significant difference was found between equations with and without NY regarding partial correlation coefficients for X1, X2 and X3. Another model included a dummy for “establish-ment after 1997 （D1997）” that will capture the impact of the Asian economic crisis in 1997. Since Gillespie

（2011） and Zen et al. （2005） reported that the amount of plasma land share decreased rapidly after the crisis, we expect it to have some influence. However, among our sample companies, there are only 17 that were estab-lished after 1997, and the regression results including D1997 were not fundamentally different from those of the equation （1）.

Table 10. Principal component loadings on ten activities/concepts to relationships with communities

Principal components

First（Overall relationships with local communities）

Second（Societal

harmonization）

Third（Harmonization with local society）

COM1 Communication with local communities 0.3983 0.1235 0.5098 COM2 Compliance with laws and regulations 0.4335 0.6271 0.2569 COM3 Fairness of plantation management 0.4464 0.7343 －0.1820 COM4 Respect for local land rights 0.6285 －0.3462 －0.1326 COM5 Economic benefit to local communities 0.6577 －0.1394 0.2830 COM6 Environmental conservation 0.5283 －0.2839 －0.4149 COM7 Infrastructure development 0.6818 －0.1518 －0.0894 COM8 Transparency 0.5354 0.2527 －0.6449 COM9 Food availability in communities 0.6897 －0.1529 0.2496 COM10 Prioritization of benefit to communities 0.7077 －0.1638 0.1658

Note :We don’t use samples missing FFB production data and samples of imperfect response about relationships with local communities, either. The sample size of the analysis is subsequently reduced to 115 samples. Three principal components have eigenvalues exceeding unity. Cumulative energy contents up to the three principal components are 34％, 47％ and 58％ .

likely to be established in rural areas with lower population density and lower income because they need to be located near plantations. This condition naturally makes the influence of popu-lation density and community income less obvi-ous.14）

6.　Conclusions

　We carried out a questionnaire survey and collected 132 replies from members of GAPKI, which is the sole association of Indonesian palm oil producers. The average farmland area of the 132 respondents is much larger than that of the large-scale companies in Indonesia. It should, therefore, be noted that the sample may have upward bias in terms of scale and that all the implications from the survey may consequently be applicable only to large-scale companies. However, the results of the survey suggest that

CSR activities have become much more popular than was previously the case in this country. We also found that most companies place great-er importance on their relationships with local communities.　Statistical analyses of the results concluded that not only the size of the company, but also the performance of the NES scheme and the perceptions of relationships with local communi-ties are major driving forces for stimulating CSR activities. As for the NES scheme, we eval-uated the extent of the scheme from the scores of the first principal component in PCA analysis combining the variables of the NES programs implemented by the companies. We found that the overall effort of the NES scheme is an im-portant factor for stimulating CSR activities.　When the NES scheme was launched in 1977, the government strongly supported it with sub-

30

14）　Since the samples are from 13 provinces, heteroscedasticity in the standard error may exist. We conducted Goldfeld-Quant’s test to see if the standard errors for the regressions were heteroscedasticity robust. The re-sults of the test show that heteroscedasticity in the standard error was not a serious problem.

Table 11. Summary results of regression analysis by Equation （1）（n＝115）

Model 1 Model 2 Model 3 Model 4

X1 : Company size by FFB production （t） 0.2973 ＊＊＊ 0.3012 ＊＊＊ 0.3740 ＊＊＊ 0.3783 ＊＊＊

（0.0001）（0.0002）（0.0000）（0.0000）

X2 : Effort in NES scheme （composite） 0.4653 ＊＊＊ 0.4679 ＊＊＊ ─ ─（0.0000）（0.0000） ─ ─

X3 : Relation with local community （composite） 0.2896 ＊＊＊ ─ 0.2938 ＊＊＊ ─（0.0001） ─ （0.0007） ─

PS : Share of plasma farmers in terms of land area 0.0744 0.0801 0.0839 0.0898 （0.3113）（0.3077）（0.3305）（0.3216）

RF: Annual rainfall （mm）－0.0615 －0.0166 －0.0218 0.0240 （0.4551）（0.8491）（0.8208）（0.8104）

PD: Population density （people/km2）－0.1235 －0.1412 －0.0711 －0.0888 （0.1363）（0.1107）（0.4607）（0.3803）

RG: Regional GDP per capita （Million Rupiah）－0.0906 －0.1176 －0.0564 －0.0836 （0.2494）（0.1610）（0.5396）（0.3853）

Constant 6.9594 5.7237 5.2972 4.0338

Adjusted R squared 0.4341 0.3539 0.2205 0.1404 p value （0.0000）＊＊＊（0.0000）＊＊＊（0.0000）＊＊＊（0.0010）＊＊＊

AIC 229.4 243.7 265.3 275.6 ＊：（p＜0.1）, ＊＊：（p＜0.05）, ＊＊＊：（p＜0.01）Note : Degree of freedom of regression is 107 in Model 1.

sidies. However, recent policies have favored market-based approaches resulting in decreased direct support for the scheme （McCarthy et al., 2011）. The diversified NES performance by companies may be a consequence of reduced government intervention. On the other hand, re-cent policy has enabled the companies to hold the majority of their estates, which has resulted in increased tension between the companies and the local communities. This situation implies that the government can play an important role in this arena. In order to mitigate the negative consequences from the policy change mentioned above, there is scope for the government to cre-ate a more supportive environment to promote CSR in the palm oil industry.　Since the history of CSR in Indonesia is short, this study was directed toward the palm oil companies’ activities. Collecting information about their current CSR activities and the com-panies’ attitudes is a first step toward a more in-depth evaluation of the effects of those activities in the formulation of better relationships be-tween the companies and the communities. Fur-ther studies will be necessary to identify the im-pacts of these activities. It would be particularly useful to know the reciprocal relationships, i.e., the local residents’ perceptions of the company’s CSR activities.15） We hope that the present study will be followed up by such studies, which could contribute to the greater impact of CSR activities in terms of improving the welfare of the local communities as well as contributing to environmental conservation.

Ackowledgements　The authors gratefully acknowledge the financial support provided by JSPS KAKENHI Grant Number 23405035.

ReferencesArli, D. I. and H. K. Lasmono （2010） Consumers’

Perception of Corporate Social Responsibility in a Developing Country, International Journal of Con-sumer Studies 34（1）: 46─51.

Badan Pusat Statistik （BPS）（2008） Statistical Year Book of Indonesia 2008, BPS.

Badan Pusat Statistik （BPS）（2010） Statistical Year Book of Indonesia 2010, BPS.

Badan Pusat Statistik （BPS） Statistics Indonesia,

http://www.bps.go.id/eng/index.php （accessed on January 24, 2014）.

Baker, M. （2004） Corporate Social Responsibility : What does it mean? http://www.mallenbaker.net/csr/definition.php （acessed on March. 17, 2014）.

Baughn, C. C., N. L. （D.） Bodie and J. C. McIntosh （2007） Corporate Social and Environmental Re-sponsibility in Asian Countries and Other Geo-graphical Regions, Corporate Social Responsibility and Environmental Management 14 : 189─205.

Besley, T., and M. Ghatak （2007） Retailing Public Goods : The Economics of Corporate Social Re-sponsibility, Journal of Public Economics 91 : 1645─1663.

Blackman, A. （2010） Alternative Pollution Control Policies in Developing Countries, Review of Envi-ronmental Economics and Policy 4（2）: 234─253.

Carter, C., W. Finley, J. Fry, D. Jackson and L. Willis （2007） Palm Oil Markets and Future Supply, Eu-ropean Journal of Lipid Science and Technology 109（4）: 307─314.

Chambers, E., W. Chapel, J. Moon and M. Sullivan （2003） CSR in Asia : A Seven Country Study of CSR Website Reporting, International Centre for Corporate Social Responsibility, Nottingham Uni-versity Business School, Nottingham University, 1─43.

Direktorat Jenderal Perkebunan （DJP）（1992） Bab III : Pola Perusahaan Inti Rakyat Perkebunan

（PIR-BUN）, Sistim, Latar Belakang, Perkem-bangan dan Masalah-masalahnya （Chapter III : Nuclear Estate Smallholders Plantation Scheme

（PIR-BUN）, System, Background, Development and Problems）, in Perusahaan Inti Rakyat Perke-bunan : Pelaksanaan dan Penilaian （Nuclear Estate Smallholders Plantation : Implementation and Eval-uation）, DJP, 3.1─3.19. （in Indonesian）

Direktorat Jenderal Perkebunan （DJP）（2011a） In-donesian Sustainable Palm Oil （ISPO）, DJP. （in In-donesian）

Direktorat Jenderal Perkebunan （DJP）（2011b） Pelaksanaan Program Corporate Social Responsi-bility （CSR）（Implementation of Corporate Social Responsibility Program）, DJP. （in Indonesian）

European Commission （EC）（2011） Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions : A Renewed EU Strategy 2011─14 for Corporate So-cial Responsibility, EC, 1─15.

Fifka, M. S. （2013） Corporate Responsibility Report-ing and its Determinants in Comparative Perspec-tive : A Review of the Empirical Literature and a


15）　In another survey conducted later in 2013, we found that the main causes of the conflicts among companies and farmers are rooted in the farmers’ lack of trust of the companies. Such trust could only be achieved gradu-ally by companies’ efforts to enhance farmers’ income.

Meta-analysis, Business Strategy and the Environ-ment 22 : 1─35.

Food and Agriculture Organization of the United Nations （FAO） Production, in FAOSTAT, http://faostat3.fao.org/home/index.html （accessed on No-vember. 27, 2013）.

Gabungan Pengusaha Kelapa Sawit Indonesia （GAPKI）（2009） Build Indonesia with Palm Oil, Mitra Media Nusantara, 1─28.

Gillespie, P. （2011） How Does Legislation Affect Oil Palm Smallholders in the Sanggau District of Kali-mantan, Indonesia? Australasian Journal of Natu-ral Resources Law and Policy 14（1）: 1─37.

Gillespie, P. （2012） The Challenges of Corporate Governance in Indonesian Oil Palm : Opportunities to Move Beyond Legalism? Asian Studies Review 36 : 247─269.

Greening, D. W. and D. B. Turban （2000） Corporate Social Performance as a Competitive Advantage in Attracting a Quality Workforce, Business and Society 39（3）: 254─280.

Henson, I. E. （2011） Oil Palm : Ecophysiology of Growth and Production, in S. A. Penna ed., Oil Palm : Cultivation, Production and Dietary Compo-nents, Nova Science Publishers, 179─211.

Hettige H., M. Huq, S. Pargal and D. Wheeler （1996） Determinants of Pollution Abatement in Develop-ing Countries : Evidence from South and Southeast Asia, World Development 24（12）: 1891─1904.

Hsu, J. L. and M. C. Cheng （2012） What Prompts Small and Medium Enterprises to Engage in Cor-porate Social Responsibility? : A Study from Tai-wan, Corporate Social Responsibility and Environ-mental Management 19（5）: 288─305.

Integrated Regional Information Networks （IRIN）（2010） Indonesia : Demand for Palm Oil Fuels Land-grabbing, http://www.irinnews.org/Report/ 89747/INDONESIA-Demand-for-palm-oil-fuels-land-grabbing on 13/06/2012 （accessed on November. 27, 2013）.

Jelsma, I., K. Giller, and T. Fairhurst （2009） Small-holder Oil Palm Production Systems in Indonesia : Lessons Learned from the NESP Ophir Project, Plant Production Systems, Plant Sciences Group, Wageningen University, 1─95.

Kathuria V. （2007） Informal Regulation of Pollution in a Developing Country : Evidence from India, Ecological Economics 63 : 403─417.

Kemp, M. （2001） Corporate Social Responsibility in Indonesia : Quixotic Dream or Confident Expecta-tion? （Technology, Business and Society Pro-gramme Paper 6）, United Nations Research Insti-tute for Social Development, 1─41.

Kitzmueller, M. and J. Shimshack （2012） Economic Perspectives on Corporate Social Responsibility, Journal of Economic Literature 50（1）: 51─84.

Marti, S. （2008） Losing Ground : The Human Rights Impacts of Oil Palm Plantation Expansion in Indo-

nesia, Friends of the Earth, Life Mosaic and Sawit Watch, 1─108.

McCarthy, J., P. Gillespie and Z. Zen （2011） Swim-ming Upstream: Local Indonesian Production Net-works in “Globalized” Palm Oil Production, World Development 40（3）: 555─569.

McWilliams, A. and D. Siegel （2001） Corporate So-cial Responsibility : A Theory of the Firm Perspec-tive, Academy of Management Review 26（1）: 117─127.

Nunnally, J. C. （1978） Psychometric Theory （2nd ed.）, McGraw-Hill.

OECD （2005） Corporate Responsibility Practices of Emerging Market Companies : A Fact Finding Study （Working Papers on International Invest-ment, Number 2005/3）, OECD, 1─32.

Paoli1, G. D., Y. Betsy, P. L. Wells and A. Sileuw （2010） CSR, Oil Palm and the RSPO: Translating Boardroom Philosophy into Conservation Action on the Ground, Tropical Conservation Science 3

（4）: 438─446.Pargal, S. and Wheeler, D. （1996） Informal Regula-

tion of Industrial Pollution in Developing Coun-tries : Evidence from Indonesia, Journal of Political Economy 104（6）: 1314─1327.

PT Astra Agro Lestari （2012） Community Develop-ment, in Reflection for the Future : 2011 Sustaina-bility Report, PT Astra Agro Lestari, 97─117.

Rist, L., L. Feintrenie and P. Levang （2010） The Livelihood Impacts of Oil Palm : Smallholders in Indonesia, Biodiversity and Conservation 19（4）: 1009─1024.

Rosser, A. and D. Edwin （2010） The Politics of Cor-porate Social Responsibility in Indonesia, The Pa-cific Review 23（1）: 1─22.

Roundtable on Sustainable Palm Oil （RSPO）（2007） RSPO Principles and Criteria for Sustainable Palm Oil Production : Including Indicators and Guidance, October 2007, RSPO.

Sedyono, C. H. （2007） Current Situation of CSR in Indonesia, http://www.adbi.org/conf-seminar- papers/2007/10/24/2377.csr/ （accessed on No-vember 27, 2013）.

Shauki, E. （2011） Perceptions on Corporate Social Responsibility : A Study in Capturing Public Confi-dence, Corporate Social Responsibility and Envi-ronmental Management 18 : 200─208.

Sugino, T., H. Mayrowani, Supadi, Sunarsih and H. Kobayashi （2013） The Impact of Supports by a Plantation Company on Small Scale Oil Palm Farmers’ Productivity : A Case of The Nuclear Es-tate Small Holders Program in Riau Province, In-donesia, Journal of Rural Economics : Special Issue 2013 : 319─326 （in Japanese）.

Tiominar, B. （2011） Plantations and Poverty : Notes from a Village Deep in Oil Palm Territory, in C. Marr ed., Down to Earth.

Tres, A., G. van der Veer, M. Alewijn, E. Kok and S.

32

M. van Ruth （2011） Palm Oil Authentication : Classical Methodology and State-of the-Art Tech-niques, in S. A. Penna ed., Oil Palm: Cultivation, Production and Dietary Components. Nova Science Publishers, 1─44.

United Nations （2014） United Nations Global Com-pacts, https://www.unglobalcompact.org/index.html （accessed on August 31, 2014）.

Welford, R. （2005） Corporate Social Responsibility in Europe, North America and Asia, Journal of Cor-porate Citizenship 17 : 33─52.

World Bank （1993） Nucleus Estates and Smallhold-ers Projects in Indonesia, in Independent Evalua-tion Group （IEG）, http://lnweb90.worldbank.org/oed/oeddoclib.nsf/DocUNIDViewForJavaSearch/ 95D104DD2107D21D852567F5005D8461（accessed on November 27, 2013）.

World Business Council for Sustainable Development （WBCSD）（1999） Corporate Social Responsibility : Meeting Changing Expectations. WBCSD, 1─30.

Zen, Z., C. Barlow and R. Gondowarsito （2005） Oil Palm in Indonesian Socio-Economic Improvement : A Review of Options （Working Paper No. 11）, Re-search School of Pacific and Asian Studies, Aus-tralian National University, 1─25.

（Received November 28, 2013 ; accepted December 17, 2014）

Appendix 1 : The Nuclear Estate Smallholder （NES） Scheme

　The Nuclear Estate Smallholder （NES） scheme was launched by the Indonesian govern-ment in 1977. It is a kind of land allocation pro-gram intended to promote collaboration be-tween small-scale farmers and plantations operated either by private companies or as state enterprises. Although the primary crop covered by the NES program in its early stages was rubber, it was soon expanded to include other tree crops, and oil palm has been the ma-jor crop since the middle of the 1980s （DJP, 1992）. In cases where oil palm is the crop in an NES scheme, the process begins with a partici-pating company developing a palm oil refinery and plantation. The plantation consists of two parts : one area named nuclear estates and a sep-arate area planted by smallholders. The nuclear estates are directly managed by the company. The smallholders, called plasma farmers, are newly established under the NES scheme, and the areas they manage are called plasma farms. The company also deals with the fresh fruit bunch （FFB） products from other farmers out-side the plantation, called independent farmers, as well as from their own estate and the plasma

farmers. Shares of FFB procurement from the respective sources differ by company （Sugino et al., 2013）.　According to DJP （1992）, a 2-hectare plot of oil palm is normally allocated to each plasma farmer after the plantation company launches the NES scheme and transplants the seedlings. In addition, each plasma farmer is also allocated 0.25 hectares of land for residential use and 0.75 hectares of farmland for food crops. Local resi-dents living near the plantation as well as trans-migrates from densely populated Java Island are eligible to be plasma farmers.　The company is responsible for providing the plasma farmers with soft loans to purchase the farmland. The farmland can then be used by the farmers as collateral, while the interest on the loans is subsidized by the government. Af-ter the commencement of commercial produc-tion, which is normally three years after the transplanting, the farmer is obliged to sell her/his products to the company. The company then deducts a percentage, usually 30％ , off the price of the FFB products it purchases from the plasma farmers as repayment of the loan. Once the loan has been cleared, full entitlement is transferred to the farmer.　Under the NES scheme, the company is also responsible for supporting the plasma farmers through various kinds of activities, such as pro-viding soft loans for purchasing farm inputs and technical support ; assisting with infrastructure development ; purchasing FFB from farmers at a guaranteed price ; and establishing Koperasi Unit Desa （KUDs）, which are a kind of agricul-tural co-operative. However, not all of these sup-port activities are necessarily provided in each of the NES schemes. Rist et al. （2010） reported in their studies conducted in Sumatra and Kali-mantan that although palm oil companies made various promises, including technical assistance with plantation management, they often failed to fulfill these promises and faced conflicts with the local communities as a result.　Zen et al. （2005） evaluated the NES highly for its contribution to improving the rural economy, although there were significant differences in the level of performance from one scheme to an-other. In contrast, Jelsma et al. （2009） reported that by the middle of the 1980s, it was widely recognized that many NES schemes had failed or underperformed. Possible reasons for these failings were identified by the World Bank


（1993） in three noteworthy observations : first, some of the land area would often be left unallo-cated, thus failing to fulfill its potential ; second, the passing of land titles to plasma farmers was often delayed ; and third, the farmers did not al-ways possess the necessary skills, particularly in the area of fertilizer application.　Throughout the 1980s and into the early 1990s, plasma farms took up the majority of the land, typically up to 70─75％ of the total land area （Zen et al., 2005）. However, after the Asian economic crisis erupted in 1997, nuclear estates began to dominate the allocation of the farmland at the expense of the area taken by smallhold-ers in the newly started projects （Gillespie, 2011）. Agriculture Ministerial Regulation No. 26/2007 accepted a request by companies to hold the majority of their estates. In recent times, the companies have owned and managed up to 80％ of their estates （Gillespie, 2011）.　The reduced land share of plasma farms in oil palm plantations after the Asian economic crisis means that the local farmers are benefiting less now than in the previous projects. The reduc-tion in land area allocated to plasma farms natu-rally limits the opportunities for local farmers to participate in the NES scheme.

Appendix 2 : CSR Activities in the Indonesian Palm Oil industry

　In 2007, Indonesia legislated CSR under Law No. 40/2007 on Limited Liability Companies

（Rosser and Edwin, 2010）. However, heavy op-position from the Kamar Dagang dan Industri, or KADIN, the Indonesian National Chamber of Commerce, as well as from other representa-tives from the business and industrial world, put pressure on parliament to amend the law, making it applicable only to the natural re-source sectors （Sedyono, 2007）. However, one of the serious problems with this proposal is that there is no clear definition of what is covered by the natural resource sectors. Rosser and Ed-win （2010） concluded that the future direction of CSR policy in Indonesia will depend on

strong political leadership which supports en-forcement of the law and that without such leadership, the legal mandatory requirement of CSR will be difficult to implement. In spite of an unclear legal definition of the natural resource sectors which are obliged to practice CSR activi-ties, some oil palm companies as shown in Table 1 state in their annual or CSR reports that they are implementing CSR activities as their man-date based on law No. 40/2007.　We can point out three important factors to enhance CSR activities in the Indonesian palm oil industry. First, although government orient-ed, the NES schemes have ─to some extent at least─ paved the way for establishing mutual relationships between palm oil companies and local farmers. Second, palm oil companies have strong incentives to avoid any conflict with local communities and to establish better relation-ships so as not to hinder their profit-making en-deavors in the long run. A similar view is ex-pressed in PT Astra Agro Lestari （2012）, which states that community development is one of the non-tax obligations undertaken by palm oil companies. Gillespie （2012） also argued that oil palm plantations which fail to move beyond le-gal compliance in their relationship with local smallholders will continually suffer from ten-sions with those smallholders, who often feel in-justice in terms of the plantation companies’ dis-tribution of benefits. Third, palm oil companies are seriously concerned with their external rep-utation in both national and international con-texts. Due to the nature of their considerable impacts on resources and the environment, ex-ternal pressures tend to be stronger. Activist campaigns directed at high-profile palm oil pro-ducers and buyers have prompted the cancella-tion of contracts with these producers （Paoli et al., 2010）. The theoretical model by Besley and Ghatak （2007） concluded that CSR could be harmonized with a company’s profit making in the long run due to the enhancement of the company’s reputation among its consumers that such CSR activities would create.

34

determinants for csr in developing countries: the case of

Documents