EFFECTS OF AN EXPORT TAX ON THE FARM-GATE PRICE OF INDONESIAN COCOA BEANS. Authors: Jeff Neilson*, Kartika Fauziah**, and Alexander Meekin*

Transcription

1 EFFECTS OF AN EXPORT TAX ON THE FARM-GATE PRICE OF INDONESIAN COCOA BEANS Authors: Jeff Neilson*, Kartika Fauziah**, and Alexander Meekin* *School of Geosciences, University of Sydney, Australia, 2006 **South Sulawesi Institute for the Assessment of Agricultural Technology, Makassar, Indonesia, ABSTRACT The Government of Indonesia has introduced an export tax on unprocessed cocoa beans to improve the supply of raw materials to the domestic processing sector, thereby enhancing its international competitiveness and increasing the total value-added retained within Indonesia. This policy initiative has been successful in encouraging local investment in the processing sector and increasing both the value and volume of processed cocoa exports. However, the extent to which industrial policy objectives have been achieved through an effective subsidy transfer from Indonesian cocoa farmers continues to be debated. This paper presents empirical evidence of the effect of the policy on the cocoa price received by Indonesian farmers on the island of Sulawesi, calculated as a ratio of farm-gate price against the prevailing international cocoa price. This analysis draws on household survey data collected within a major cocoa-producing district in West Sulawesi before and after the introduction of the export tax. This study does not find a significant negative effect on farm-gate rices resulting from the tax and consequently does not support claims that the policy has contributed to recently depressed levels of cocoa production in Indonesia. 1

2 INTRODUCTION In April 2010, after years of public debate, the Government of Indonesia introduced a sliding tariff (between 0 and 15%) on the export of unprocessed cocoa beans. The stated aim of the policy was to improve the supply of raw materials to the domestic processing sector, thereby increasing the total value-added retained within Indonesia. The International Cocoa Organisation (ICCO, 2008) had earlier recommended an optimal export tax of 11% for Indonesia, whereby optimal referred to the maximum combined income for both cocoa producers and for government revenue. This was based primarily on the assumption that the tax would result in reduced levels of domestic production that would be compensated for by having a positive impact on world prices. Consequently, government tax revenue would be generated at the expense of domestic producers with net welfare for the producing country increased. The rationale of the tax in Indonesia, however, was presented almost exclusively in terms of improved competitiveness for domestic processors. Econometric modelling on the impact of export taxes generally conclude that such taxes constitutes a subsidy of domestic processors at the expense of producers, who receive lower prices for raw materials (Akiyama, 1992). This has also been predicted in the case of Indonesian cocoa (Arsyad, 2007, Permani et al. 2011). Furthermore, in a paper prepared for the Indonesian Ministry of Trade prior to the introduction of the tax, Marks et al. (2005) argued that an export tax on beans would significantly increase the poverty rate among cocoa-growing households in Sulawesi due to lower farm-gate prices. Since the introduction of the export tax, there has been significant increased investment in the domestic processing sector. According to the Indonesian Minister of trade (Ministry of Trade, 2013), six new cocoa processing factories with a total capacity of 430 thousand tonnes per year have been established since the tax was introduced. Official trade data suggests a sharp shift in the composition of cocoa-based exports between 2009 and 2012 towards processed products with reported exports of raw beans falling dramatically by 50% in the same period (Figure 1). Moreover, the total combined value of processed and unprocessed cocoa exports has declined. 2

3 Figure 1. Changing composition of cocoa-based exports ( ). Data Source: UNCOMTRADE (2013) According to some industry representatives in Indonesia, the export tax had an immediate impact on farmer prices and was threatening to seriously disrupt supply in the coming years (for example, Jakarta Globe, 5/9/2010). While the production of cocoa beans in Indonesia has indeed stagnated in recent years and may even be declining (Table 1), it is problematic to identify the tax as the primary or even a contributing cause when there are several other factors influencing production. Farm production has been notoriously affected by various pests and disease since around 2000, and the ability of farmers to manage these problems is a key determinant of production levels. Then, commencing in 2009, the Indonesian government embarked on a large-scale smallholder cocoa replanting and rejuvenation scheme (GERNAS) with various subsidies made available to participating farmers. In the case-study area of Polewali-Mandar (Polman), West Sulawesi, presented in this paper, 94 per cent of farm household respondents had participated in the GERNAS program. The government made bold predictions about the GERNAS program s potential to double production within three years (Ditjen Perkebunan, 2008). Despite this program, production has stagnated (Table 1). Moreover, an independent report (Gusli et al., 2012) suggests that the program may have even had a negative impact on production levels due to the dissemination of poor quality planting material. The projected mechanism through which national production levels are thought to be affected by the tax is through suppressing farm-gate prices, thereby creating disincentives for cocoa farming. The relatively elevated international cocoa prices in 2011 and 2012 (Figure 4), however, would have had a nullifying effect on the outright prices received by farmers during the period. 3

4 2006/ / / / / / /13 Estimates Forecasts 545, , , , , , ,000 Table 1. Production of Cocoa Beans in Indonesia (tonnes) Source: ICCO (2013a) The aim of this study, therefore, is to assess the effects of the export tax on the farmgate price being received by farmers in a major cocoa-growing region in Sulawesi through an empirical analysis, remembering that a negative impact has been strongly suggested theoretically through econometric modelling. The logic of this econometric argument is that exporters paying the tax would simply pass the cost on to regional traders, who would also pass it on to their suppliers, so that ultimately it would be borne by the farmers. An implicit assumption here is that Indonesian supply is unlikely to be sufficient to influence global cocoa prices (ie. supply is price inelastic), which is an assumption countered elsewhere in the literature (eg. Akiyama, 1992; ICCO, 2008). Regardless of these assumptions, our paper examines the empirical evidence of an effect. A recent report (Rifin, 2012) examined a similar question by also drawing on an empirical price data supplied by a major exporting firm with an upcountry buying station in the Luwu District of South Sulawesi. Rifin (2012) found that the effect of the export tax was not being transmitted to the price paid at the buying station. Our current study will further Rifin s (2012) analysis by adding further empirical data collected instead through a farmer survey and taking into account marketing conditions such as market chain structures and moisture levels. MATERIALS AND METHODS This study draws on empirical survey data in Polman, West Sulawesi. An earlier farm household survey (n=197) was undertaken by two of the authors across eight subdistricts within Polman between December 2008 and February 2009 (ie. about 16 months before the export tax was introduced, and referred to hereafter as the 2008 survey for ease of presentation). Part of that survey collected information about farmgate prices, extent of drying at the point of sale, and information about cocoa marketing structures. Data from that earlier survey is treated here as a pre-tax baseline. From October to December 2012 (ie. twenty months after the export tax), a repeat survey was conducted with the specific aim of assessing the farm-level effects of the tax. A similar number of respondents (n=200) were randomly selected from the same villages surveyed in 2008 to avoid the potential local effects of market accessibility on prices. Respondents were asked about other factors that might influence the farm-gate prices being received, such as their participation in certification schemes (Utz or Rainforest Alliance), the structure of the local marketing chain and their first-stage buyer, and moisture levels (indicated here by the extent of drying). There was a significant increase in the number of farmers selling directly to exporters (Figure 2) and an associated general trend towards longer drying periods prior to sale (Figure 3) between the 4

5 surveys. As the cocoa content of beans with lower moisture levels is higher, the price received by farmers needs to be adjusted accordingly to indicate the price paid for actual cocoa content. The direct sales to exporters are the result of direct-buying programs introduced by some exporters in Polman since 2008, often associated with a move towards sustainably certified supply chains. In 2008, none of the respondent farmers were involved in a certification scheme, whereas by 2012, 37 percent of respondents were involved in programs certified by either Utz or Rainforest Alliance. Figure 2. Cocoa marketing by farmers 1 Figure 3. Days cocoa beans were dried We assume that fluctuations in the global cocoa price are transmitted relatively efficiently to farmers in Sulawesi (as demonstrated by Rifin, 2012). Farm-gate prices were then converted to USD using the prevailing exchange rate taken from and compared as a percentage of the prevailing international price (ICCO monthly average indicator price, ICCO, 2013b). Prices were adjusted for bean moisture levels, using the following assumptions: beans dried for one day contained 40% moisture, 2 days = 20%, 3 days = 12%, 4 days = 10%, 5 days = 8%. The assumptions were suggested by a local industry researcher who had experimented with moisture levels in relation to fermentation trials (these estimates are, of course, also highly variable depending on local weather conditions). While international cocoa prices had fluctuated significantly between 2007 and 2012 (Figure 4), they were quite similar at the time of both surveys. The exchange rate between the US dollar and the Rupiah however was significantly higher during the 2008 survey (Figure 4). 1 A trader was defined here as a situation where farmers take their produce to a warehouse, whereas a collector would collect cocoa from farmers at their farm or house. 5

6 Figure 4. Fluctuations in world cocoa price and exchange rate of the Rupiah ( ), with survey period shaded. Data: ICCO (2013b); RESULTS While the average raw price received by farmers was very similar in each period (Table 2), the price actually increased in the post-tax survey when taking into account moisture levels and prevailing exchange rate, and was much higher as a percentage of the ICCO price (Figure 5) Average raw price (Rp/kg) 15,776 15,779 Average price adjusting for moisture (Rp/kg) assumption 1 Average price adjusting for moisture and exchange rate (USD/kg) Average adjusted price (% of prevailing ICCO price) Average Price differential between farmgate price and ICCO price (USD/tonne) 19,509 18, % 79% Table 2. Average farm-gate prices before and after the export tax 6

7 Figure 5. Farm-gate price as a percentage of ICCO world price Indonesian cocoa beans are routinely sold at a discount price to the world market due to inferior quality and the lack of fermentation. While farm-gate prices will fluctuate depending on international markets, the differential between the farm-gate price and the ICCO price would be expected to remain relatively constant over time. While the imposition of an export tax would be expected to increase transaction costs along the value chain and to widen this differential, our survey actually found a considerable decrease in this differential between 2008 and 2012 (Table 2). DISCUSSIONS This empirical study has not identified a noticeable downward impact of the tax on farm-gate prices as otherwise predicted through modelling. While the precise reasons for this remain unclear, it appears that significant restructuring has taken place within the Indonesian cocoa value chain during the period under investigation that has increased the farm-gate price for cocoa more significantly than any potential negative effect induced by the tax. The most fundamental change in the value chain relates to the increased upstream activity of cocoa exporters within the study region (as indicated in Figure 2 above). Exporters and domestic processing firms, traditionally based in export hubs such as Makassar, have developed direct supply chain relationships with farmers in the intervening period. These relationships are driven, partly at least, by international buyer demands for sustainably-certified cocoa. Participating farmers may be receiving a direct price premium from certification, or they may be benefitting indirectly from an altered trade relationships associated with certification, which has effectively eliminated middle-men. Direct farmer relationships have also been established by at least one local processing firm in an attempt to influence quality attributes (ie. fermented beans) in their supply chain (Sulkani, 2013). 7

8 It is possible that the lower proportion of ICCO prices received by farmers in 2008 was due to profiteering along the value chain in 2008 where local traders refused to pass on higher Rupiah prices during a temporary depreciation of the Rupiah. Transactions along the Indonesian value chain for cocoa do not operate perfectively, and information asymmetries exist between individual actors that ultimately shape transaction costs and price transmission. The influence of interlocked credit markets between farmers and collectors is also likely to distort market dynamics. The declining differential between farm-gate prices and international prices reported here (Table 2) appears to support the conclusions drawn by Rifin (2013) that the tax is being primarily borne by the exporter, or at least by smaller traders and collectors, some of whom may have been squeezed out of the market entirely by a variety of factors, potentially including the tax. Malaysia had been the major destination for Indonesian cocoa bean exports leading up to 2009, and reported exports of beans to Malaysia declined significantly following the introduction of the tax. It is plausible that new illegal trade networks across this relatively porous border may have emerged to circumvent tax payment, which could have negated the negative effect of the tax. ICCO data (2013a), however, suggests a marked increase in imports to Malaysia from West Africa in 2011 and 2012 allowing processing activities to continue and the extent of smuggling does not appear to be great. CONCLUSIONS This empirical study has not identified a noticeable downward impact of the tax on farm-gate prices as otherwise predicted through economic modelling. While these are preliminary results and the assumptions regarding moisture levels and price transmission between international markets and farm-gate prices require further examination, subsequent refinement of the analysis is unlikely to fundamentally alter the key findings. It is, therefore, considered extremely unlikely that the export tax has had a negative impact on cocoa production levels in Indonesia. The primary reason offered to explain the absence of an observed effect at the farmlevel is the considerable restructuring of the value chain that has taken place during the period In particular, the emergence of new marketing channels for farmers direct to exporters, linked to the introduction of a sustainability program in the case study location, has diminished the profits of middle-men over the same period. These savings have been transmitted to farmers by way of higher prices. This situation is not unique to the Polman District and appears to reflect be a nation-wide development towards direct-buying programs and certification. This development was unrelated to the introduction of the tax and has coincidently taken place during the intervening period. Another contributing factor to elevated farm-gate prices observed in the second period, however, could be directly attributed to the tax. The increased investment in the 8

9 domestic processing sector (indeed to the point that total installed capacity appears to be exceeding national production) has resulted in the emergence of increased competition for beans between exporters and domestic processors, the latter of which are also developing farmer-linkage programs and resulting in higher farm-gate prices. From a policy perspective, the apparent benefits from the tax in terms of the development of industrial capacity, employment generation in downstream manufacturing, government revenue generation and value-retention need to be assessed against the absence of observable farm-level effects of the tax. It is also important to assess any effects on farmer prices against the much greater fluctuations in the international cocoa market, ranging from a low of 1900USD in April 2007 to a high of 3500 USD in January 2010 (Figure 4). These price fluctuations are far greater than the price effect would be even if the full tax was passed on entirely to farmers. 9

10 REFERENCES Akiyama, T.(1992). Is there a case for an optimal export tax on perennial crop. Policy Research Working Paper WPS 854. Akiyama, T. and Larson, D.F. (1994). The adding-up problem. Strategies for primary commodity exports in Sub-Saharan Africa. Policy Research Working Paper No Arsyad, Muhammad. (2007). The Impact of Fertilizer Subsidy and Export Tax Policies on Indonesia Cocoa Exports and Production. Ryokoku Journal of Economic Studies, Vol 47(3), pp1-27. Ditjen Perkebunan (2008). Program Revitalisasi Kakao Nasional, Paper presented at the National ASKINDO Working Group Meeting in Denpasar, Bali, 31 October, Dirjen Perkebunan, Departemen Pertanian Republik Indonesia. Gusli, S., Samsuar, Useng, D.,Darmawan, Sarisi, L. (2012). Field Performance of cocoa trees produced by somatic embryogenesis, Sulawesi, Pusat Penelitian dan Pengembangan Sumberdaya Alam, Hasanuddin University, Makassar ICCO (2008). Optimal Export Taxes in Cocoa Producing Countries, Consultative Board on the World Cocoa Economy Fifteenth Meeting, Monday 14 January 2008, International Cocoa Organisation, London. Available at (accessed 12 August, 2013) ICCO (2013a). ICCO Quarterly Bulletin of Cocoa Statistics, Vol. XXXIX, No. 2, Cocoa year 2012/13, International Cocoa Organisation, London. ICCO (2013b). ICCO Monthly Averages of Daily Prices, International Cocoa Organisation, London. Available at Jakarta Globe (2010), National Cocoa Export Tax Could Grind Farmers Down, September 5, 2010, Available at (Accessed 20 August, 2013) Marks, S.V., Paryadi, T. A., Wicaksono, Y. Y. and Bird, K. (2005). An Assessment of Major Issues in the Cocoa Sector and Policy Proposals to Support Cocoa Processors, Final Report June 30, Paper prepared for the Ministry of Trade, Republic of Indonesia. Ministry of Trade (2013). Export Duty of Cocoa Boost Development of Indonesian Downstream Industry. Ministry of Trade Press Release, June Permani, R. and Setyoko, N. R. (2011). Optimum Level and Welfare Effects of Export Taxes for Cocoa Beans in Indonesia: A Partial Equilibrium Approach, Paper presented at the 2011 AARES Annual Conference 8-11 February 2011 in Melbourne Rifin, A. (2012), Impact of Export Tax Policy on Cocoa Farmers and Supply Chain, SEADI Discussion Paper No. 1, USAID, United States Agency for International Development (USAID), Jakarta Sulkani (2013). Kemitraan Usaha Antara Petani Kakao dengan Asosiasi Eksportir dan Industri Kakao Indonesia (AEIKI), Friday 24 May, 2013, available at Dirjen Perkebunan, Departemen Pertanian Republik Indonesia. UNCOMTRADE (2013). United Nations Commodity Trade Statistics Database. Available at (accessed 15 August, 2013). United Nations, New York. 10