Analysis of trade potential and factors influencing chili export in Indonesia

2.1 Data resources

The data used in this study are secondary data of 31 years, namely, data from 1990 to 2020. Data analysis was carried out from August to September 2021. Panel data are a combination of time series data and cross-section data. The time-series data include the volume of chili exports, the real GDP/capita of Indonesia, the real GDP/capita of the destination country, the rupiah exchange rate against the currency of the export destination country, the geographical distance between Indonesia and the export destination country, the real interest rate of the destination country, and involvement export destination countries in the WTO organization. The selected export destination countries are Nigeria, Saudi Arabia, and Malaysia because they are the main export destinations of chili [5]. In addition to these three countries, Singapore was chosen as an export destination because, according to the data from UN Comtrade [27], Singapore has consistently been an importer of chilies from Indonesia from 1991 to 2020, even though the quantity of chilies exported to Singapore is not as much as other export destination countries. Data on the volume of chili exports were obtained from the United Nations Commodity and Trade (https://comtrade.un.org/data/), data on the real GDP/capita of Indonesia and the real GDP/capita of the destination country were obtained from the World Bank (https://data.worldbank.org/indicator/NY.GDP.PCAP.KN), the exchange rate of the rupiah against the currency of the export destination country was obtained from the World Bank (https://data.worldbank.org/indicator/PA.NUS.FCRF), the real interest rate of the destination country was obtained from the World Bank (https://data.worldbank.org/indicator/FR.INR.RINR), the geographical distance between Indonesia and the export destination country was obtained from Indonesia Distance World (http://indonesia.distanceworld.com/country/ID), and data on the involvement of export destination countries in the WTO organization were obtained from the World Trade Organization (https://www.wto.org/english/thewto_e/whatis_e/tif_e/org6_e.htm).

2.2 Analysis method

This study uses four importers as Indonesia’s export destinations: Saudi Arabia, Nigeria, Malaysia [5], and Singapore [27]. Saudi Arabia, Nigeria, and Malaysia are Indonesia’s export destinations with higher export values than other countries [7]. In determining the factors that influence the volume of chili exports from Indonesia, the data were analyzed using panel data regression which was different from time-series data regression or cross-sectional data. Mathematically, panel data are described by the following equation:

where i = 1 , 2 , … , N , t = 1,2 , … , T , i indicates the unit of destination export country (cross-section), and t indicates the time period being analyzed (time-series). y it represents the response of the observations of the country i and the time period t . α is the intercept, β is the row vector of the regression coefficients for the observations of the country i , and the time periods t and μ it are confounding factors. Most of the data panel application devices use a model with a one-way error component which is written as follows:

where τ i represents the unobserved individual effect and ϵ it represents the confounding factor for the country i observations and time period T [33]. The panel data regression process is carried out through a pooled model approach, fixed effect model (FEM), and random effect model (REM).

1. Pooled Model is one of the models in panel data analysis where the regression coefficient (constant or slope) between the cross-section and time-series units is the same. Then, to estimate the parameters, the ordinary least squares (OLS) were used [34]. The equation of the pool model is as follows:

   (3) y it = α + ∑ i = 1 n β j x jit + u it ,

   where y _it is the dependent variable of country i in year t , x jit is the independent variable of country i in year j , α is the intercept, β j is the coefficient of the independent variable, and u it is the error component [35].

2. The FEM is a model that can be used by considering that the omitted variables can result in changes in the cross-section and time-series intercepts. This FEM model focuses on changes within the individuals rather than differences between individuals. This allows researchers to separate the differential effects of increasing and decreasing predictor factors [36]. The assumptions for this model are that τ i is assumed to be an estimated fixed parameter, ϵ it is an independent stochastic confounder and identically distributed (0, σ ϵ 2 ), and E ( X it , ϵ it ) = 0 , X it is assumed to be independent with ϵ it for all components i and t . The following equation describes the FEM:

   (4) y it = α + x it ′ β + τ i + ϵ it .

   If this equation is averaged over time, it will become equation (5).

   (5) y ˆ i = α + x ˆ i ′ β + τ i + ϵ ˆ i .

   The next step is to reduce equation (4) with equation (5) to equation (6).

   (6) y it − y ˆ i = ( x it ′ − x ˆ i ′ ) β + ( ϵ it − ϵ ˆ i ) .

   The model in equation (5) is a form of transformation estimated by the OLS method.

   The intercept in the FEM controls for each individual-specific and time-invariant, which is written in the following equation model:

   (7) y it = α i + ∑ i = 1 n β j x jit + u it ,

   where α i is the individual intercept in state i [37].

3. The REM is a model that incorporates different parameters between individuals and over time into the error to increase the efficiency of the estimated parameters. Unlike the fixed effects approach, random effects do not discard variation across individual units. The additional information on inter-individual variation provides several advantages, namely, less variability and narrower confidence intervals [37]. The assumption for this model is that τ i usually is distributed N( 0 , σ τ 2 ), ϵ it is an independent stochastic confounder and identically distributed, E ( X it , τ i ) = 0 and E ( X it , ϵ it ) = 0 , X it are assumed to be independent with ϵ it for all components i and t . REM allows to estimate the effect of variables that are individually invariant to time, as illustrated in the following equation model:

where

where u _it is the error component, μ i is the individual-specific random component, and ε it is an idiosyncratic disorder [35].

A consistent estimator is obtained through OLS, which can create a square standard error. The use of the generalized least square (GLS) model is better because it assumes that the intercept and the regression coefficient are constant for each individual. At the same time, u i changes according to the individual and is part of the error component ε i . This model is tested using the Hausman test and the Lagrange Multiplier (LM) method known as the Breusch–Pagan test [38]. Hausman test was used to test the significance between the FEM and REM. This test is critical in selecting and evaluating parameters [39].

H 0 : E ( τ i | X it ) = 0 (Random effect is the suitable model)

H 1 : E ( τ i | X it ) ≠ 0 (Fixed effect is the suitable model)

If H0 is rejected, then the FEM is more suitable. If H0 is accepted, then the results of the REM are better and should be continued with the Breusch–Pagan test to compare the REM with the pooled model. If the chi-square probability value is less than the significance value of α , then the REM is selected, and if the chi-square probability value is more than the α significance value, the pooled model is selected.

3.1 Factors affecting the volume of chili exports from Indonesia

Chili is one of Indonesia’s leading commodities. However, the growth rate of chili exports is still relatively low, so a study of the determinants of Indonesian chili exports is carried out using the gravity model. Table 2 presents the analysis of chili exports to several destination countries.

Based on the Hausman test results on data of chili export volume, real GDP/capita of Indonesia, real GDP/capita of the destination country, real rupiah exchange rate, geographical distance, actual interest rate of the destination country, and dummy membership of the WTO, the chi-square value is obtained. The chi-square value is 0.00 less than chi-square table, that is, 150.989, so H0 (H0 = random effect is the suitable model) is accepted and H1 (H1 = fixed effect is the suitable model) is rejected, meaning that the random effect is the suitable model compared to the fixed effect model. The test was continued with the Breusch–Pagan test with the result that the chi-square value was 46.36 less than chi-square table that is 150.989, so the random effect was the suitable model compared to the pooled model. Therefore, the appropriate model for this analysis is the random effect. Based on the model estimation results, it is known that the F-stat probability value is smaller than the 1% significance level (0.00 < 0.01), which means that overall, the independent variables in the model are feasible to use. There is at least one significant variable in the model. The R-squared value obtained is 0.9261, meaning that this model can explain the diversity of chili exports by 92.61%, while the remaining 7.39% is explained by other factors not included in the model.

Based on the t-test, it can be seen that from the six independent variables that make up the model, only four independent variables influence the export volume of Indonesian chili. The four independent variables are Indonesia’s real GDP/capita, the destination country’s real GDP/capita, the rupiah exchange rate against the destination country, and the geographical distance between Indonesia and the destination country [19–23,25,49]. Two independent variables are not significant, namely, the interest rate of the destination country (0.263 > 0.01) and the involvement of the destination country in WTO (0.205 > 0.01) [21,26]. It shows that changes in the independent variable do not affect the volume of Indonesian chili exports to each destination country. This is contrary to the previous research [19,22] which stated that involvement with trade unions affects export volumes.

The independent variable of real GDP/capita in Indonesia has a significant effect at the 10% level on the export volume of Indonesian chili (0.071 < 0.1) with a coefficient of 40.6395. GDP is the total income of a country or the entire population. An increase in a country’s per capita income will increase citizens’ purchasing power and consumption. Based on the results, it is known that the independent variable real GDP/capita of the country of origin positively influences a country’s export volume. In Indonesia, real GDP/capita is increasing with an average growth of 4.12%/year. Based on the previous studies [19,50], the increase in GDP will increase the purchasing power and consumption for various goods and services. This happens because if GDP increases, per capita income also increases, which indicates prosperity. Additional income will open consumption choices for other substitute goods that are better than what is usually consumed. According to Chao et al. [51], people with low income will like the spicy taste and get used to the taste. This is because they do not have many choices of food that can be reached with their income. Chili is a commodity that is easily accessible in terms of price and can be found anywhere. In addition, people with low incomes tend to allocate more of their income to food, while people with high incomes will allocate their income to secondary, tertiary, and investment needs [52]. Therefore, the higher the real GDP/capita, the less chili peppers will be consumed. Chili stock will be allocated more for export activities.

In contrast to the independent variable of real GDP/capita of Indonesia, the independent variable of real GDP/capita of the destination country has a negative and significant effect on the export volume of Indonesian chili (0.002 < 0.01). However, the high real GDP/capita of the export destination country will make the country choose chili with high-quality standards [53]. Meanwhile, in Indonesia, the quality of chili is not better than in other exporting countries, so the destination country prefers to import chili from other countries. Due to changes in quality standards in chili importing countries, imports from Indonesia are reduced. Chili quality can be seen from the level of productivity. Countries with productivity levels above Indonesia are China, India, and Pakistan [54]. This result is different from that of previous studies [22,55] which state that improvement in real GDP/capita of the destination country would create demand for Indonesia’s chili as one of the largest chili-producing countries. Generally, the percentage growth of real GDP/capita in destination countries shows positive growth of export.

The independent variable of the rupiah exchange rate against the export destination country has a significant effect on the 5% level of significance (0.015 < 0.05) with positive effect. This result is in contrast to the study by Gachena et al. [22] which reported that weaker currency makes costs of domestically produced goods lower, which consequently enhances competitiveness in exports to destination country. If the exchange rate of the rupiah against the destination country increases, the value of the rupiah weakens and results in a decrease in the value of exports. As the price of chili becomes cheaper, the demand for chili from importing countries will increase [56]. As an exporting country, Indonesia benefits from the increasing demand for chilies from other countries, and export destination countries benefit from low chili prices.

The independent variable geographical distance between Indonesia and the export destination country significantly affects the 1% significance level (0.000 < 0.01). The farther Indonesia is from the export destination country, the more chili exports will decrease. It happens because the farther the distance, the longer it will take for chili to reach the destination country. This will affect the quality of chili. Besides, it also makes shipping costs expensive which causes increase in chili prices in countries that are far away. Making it unable to compete with the price of chili produced by neighboring countries around the export destination country [41]. According to Chaabouni [57], chili exporters are negatively sensitive to far markets, which is consistent with the literature specifying that distance acts as a resistance factor in trade since it embodies transport fee and transaction costs. Shipping chilies to remote locations require special management and packaging so that the costs will increase. This is because chili is a horticultural commodity easily damaged and rotten if it is not at an optimal temperature [58]. Geographical distance in trade cooperation still has a high impact and tends to be detrimental. Anticipating improvements in transportation, goods packaging technology, providing insurance, and others are ways to deal with these losses. Even so, this variable is still one of the main factors influencing a country to export certain commodities to importing countries [48].

The independent variable, the interest rate of the destination country, has no significant effect at the 1% significance level (0.263 > 0.01), which means that any increase or decrease in the interest rate of the destination country will not affect Indonesia’s chili exports. The interest rate of the destination country does not affect the export volume of chilies because exporters may not need loans to the destination country for working capital, so even if the interest rate of the destination country increases, it will not affect the volume of chili exports [59]. In addition, in less developed credit markets, exports are not very sensitive to changes in interest rates [60]. The result is in contrast to that reported by Cecilia et al. [61] who stated that exchange rate positively affects the price and amount exported, with exports more competitive within the foreign market.

The independent variable of membership of export destination countries in the WTO does not significantly affect the export volume of Indonesian chili (0.205 > 0.01). Membership of export destination countries in the WTO does not affect the export volume of Indonesian chilies because within the organization, there are several major world-producing countries, such as China, Mexico, and Turkey. Indonesia is not the highest chili exporting country in the world, so its competition with other big countries is still relatively low. On the other hand, Indonesia’s chili importing country may not make Indonesia the leading country for importing chili products. Developed countries have benefited the most from their involvement as members of the WTO since it was founded in 1995. There was no rule to eliminate agricultural export subsidies before this rule was finally enacted in 2015. Several developing countries still apply agricultural export subsidies [62]. Even though there has been an agreement to eliminate agricultural export subsidies, developed countries, in particular, still maintain subsidies and protect their agricultural sector. This condition has impoverished developing countries, so Indonesia’s involvement as a WTO member has not yet had a significant impact, especially on chili commodities [2].

3.2 Export potential of Indonesian chili in export destination countries

The estimation results of the gravity model are used to analyze the trade potential of Indonesian chili in each destination country. Based on the analysis results, it is known that each country’s average trade potential ratio is more significant than one (Table 3), which shows that Indonesia’s chili trade with chili importers from Indonesia has not exceeded the existing trade potential, so Indonesia still has the potential to expand trade to these countries. Indonesia can develop its market in countries that have become its trading partners [63].

Sequentially, the countries with the largest to smallest trade potential are Singapore, Nigeria, Malaysia, and Saudi Arabia. Singapore is a destination country for Indonesian chili exports with the highest average value of potential trade, which is 10.61 times the actual trade potential. Thus, Indonesia has the potential to expand trade with Singapore because based on geographical distance, the distance between Indonesia and Singapore is the closest compared to the other three countries, namely, 2,850 km. While the distance between Indonesia and Malaysia is 3,360 km, Saudi Arabia is 7,898 km, and Nigeria is 11,688 km [64]. Moreover, in 2020, the actual trade volume of Indonesian chilies to Singapore is only 151,696 kg, while the predicted trade volume will reach 193102.3 kg. Cooperation between Indonesia and Singapore has positive potential; the portion of Indonesia’s non-oil and gas exports to Singapore reaches 6.36% of total exports. Singapore is the destination country with the highest export value within ASEAN [65]. As a country with limited natural resources, Singapore relies on imports of agricultural products, especially chili from Indonesia. The percentage of chili export value to Singapore is 4.44% of the total chili export value, or around US$1.1 million [7]. The relationship between Indonesia and Singapore in chili trading activities has been going on since 1991 or for approximately 28 years until 2020. Singapore started the cooperation by importing fresh Indonesian and ground chilies since 2012. Even though Singapore had experienced a recession for two consecutive quarters in 2020, it did not affect the performance of trade with Indonesia.

Nigeria is a country that has the second-highest average value of potential trade after Singapore, with a potential of 5.85 times its actual trade. Nigeria’s geographical distance is the farthest from Indonesia, but it country imports chili with the largest average amount per year compared to the other three countries, even the percentage of export value to Nigeria reaches 13.35% of the total chili export value of around US$3.36 million. As the only chili importer country from the African continent, Nigeria’s cooperation with Indonesia is quite potential. Nigeria has the most extensive total trade among other African countries such as South Africa, Egypt, Algeria, and Angola [66]. Although chili commodities are not the top export products from Indonesia, the export value contributed by chili commodities provides considerable potential, especially since Nigeria is the newest country to become an importer of Indonesian chilies compared to the other three countries, since 2013 or approximately 7 years. Since 2001, about 795 million people have experienced food crises, most of which are in Sub-Saharan Africa and Asia. In 2009, the world’s production decreased by 32.2%, and the Sub-Saharan Africa region experienced a decline of 4.5 million tons of food commodities, which impacted their financial condition where most of the GDP comes from agricultural commodities. Nigeria is a country on the African continent that contributes more than 50% of production and consumption activities. Nigeria has experienced a lot of decline in the production of chili commodities. This country can produce 50% of the total production in Africa and accounts for 1.5% of the total world chili production. However, the food crisis has pushed Nigeria to import chilies from outside its territory, Indonesia. The quantity of chili exports to Nigeria experienced a significant increase, which amounted to 568,025 tons from 225 tons in 2019 to 793,025 tons in 2020 [27]. Therefore, this condition provides an advantage for Indonesia to become an exporter in the region of Nigeria [67].

Meanwhile, Malaysia has been an importer of Indonesian chili from 1990 for approximately 20 years. Malaysia is the only chili importing country that has been importing every year since 1990. Malaysia is a country that has the third-highest average potential trade value after Nigeria, with a potential of 4.46 times its actual trade. Although the value of the trade potential index is far below Singapore’s, the value of chili exports to Malaysia is higher than Singapore’s, which is US$1.8 million. In contrast, the export value of chilies to Singapore is US$1.1 million [7]. The value of chili exports to Malaysia accounts for 7.4% of the total chili exports [7] The geographical distance between Indonesia and Singapore, and Indonesia and Malaysia is relatively close compared to other importing countries [68]. As a country without agricultural land, Singapore depends entirely on Indonesian chili production [69]. In addition, the area adjacent to Indonesia provides an advantage for Singapore to meet the needs of chili in the country. The high demand, the closer distance, and the condition of the absence of agricultural land in Singapore provide a higher trade potential for Indonesia compared to the trade potential in Malaysia.

Chili exports to Saudi Arabia started in 1992 but did not continue and started again in 2001, then stopped and again started in 2010 until 2020. Even though it had stopped importing chilies from Indonesia, Saudi Arabia has been the primary destination for chili exports for the last 3 years since 2017. In 2020, the value of Indonesian chili exports to Saudi Arabia reached US$9 million. This value decreased in 2018, reaching only US$7 million [7]. The index value of chili trade potential in Saudi Arabia is the smallest among the other three export destination countries, indicating that chili exports to Saudi Arabia have lower trade potential than Nigeria, Singapore, and Malaysia. Saudi Arabia has the highest percentage of chili export value compared to other export destination countries. Types of chilies exported to Saudi Arabia include fresh, dried, and crushed chilies. However, the largest proportion exported has been crushed chili. As a country visited by pilgrims every year, the demand for food ingredients is always high; moreover, mashed chili spices are very much needed for processed food ingredients in Saudi Arabia. Most pilgrims come from countries that consume chilies in their food, such as Indonesia, Pakistan, India, Bangladesh, and Egypt [70]. This is a potential that can be developed to be able to expand the trading market in the region of Saudi Arabia. Based on the TPI, Indonesia has the potential to expand its trade to Singapore and Nigeria.

3.3 Implications of Indonesian chili export policy in the international market

Chili farming in Indonesia is profitable, both privately and socially. The chili that Indonesia exported are fresh chilies and ground chilies. From a trade perspective, adding value for agricultural products can be created if Indonesia is involved in global exchange. It can stimulate economic activity in the country and contribute more to even income distribution [71]. Chili export is one sector that is being boosted by the government. The fact that chili exports do not grow optimally compared to domestic chili production makes this sector interesting to study. The diverse increase in import of Indonesian red chili among countries presents it as consideration-worth export opportunity. This opportunity should be utilized by the government, to start exporting chili intensively [15]. In fact, if exports run well, it will not only improve the Indonesian economy, but also the competitive and comparative value of chili in the eyes of world trade [16,17]. Based on Table 1, it can be seen that several factors affect the volume of chili exports. The estimation results of the gravity model show that the variables of Indonesia’s real GDP/capita, real GDP/capita of the destination country, the rupiah exchange rate against the destination country, and the geographical distance between Indonesia and the destination country have a significant effect on Indonesia’s chili exports. The rupiah exchange rate against the destination country has the largest coefficient and positively affects the Indonesian chili trade.

Chili is mainly exported to Malaysia, Singapore, Saudi Arabia, and Nigeria. Figure 2 shows the volume of export to Malaysia, Singapore, Saudi Arabia, and Nigeria from 1990–2020. It is shown that there is an increasing trend in the volume of chili exported and extremely high increase in the volume of chili exported to Nigeria and Saudi Arabia from 2013–2020. The potential for chili exports in Nigeria, Saudi Arabia, Malaysia, and Singapore is under trade status, so there needs to be a market expansion policy in the destination country. Singapore and Nigeria are the most potential countries for export expansion. Figure 3 shows that the largest Indonesia chili export value are to Nigeria and Singapore. The total value of chili that Nigeria imports from Indonesia is about US$15.8 million for only 7 years. Then, followed by Singapore, which is about US$5.3 million, Saudi Arabia, which is about US$1 million, and Malaysia, which is about US$816 thousand.

Figure 2

Volume export of Indonesia chili to destination countries during 1990–2020.

Figure 3

Value export of chili of Indonesia during 1990–2020 (USD).

Singapore has the closest distance to Indonesia compared to other destination countries, thus facilitating the delivery of goods in terms of financing and chili quality. Chilies are extremely sensitive to the seasons and weather, which will have an impact on the instability of the supply amount of red chili, and will affect the price of red chili if it is not managed well. Singapore, with the absence of optimal agricultural land, causes the food sector to depend on other countries, including Indonesia as an exporter of chili. This is a distinct advantage for the Indonesian agricultural sector, where most of the Indonesian people’s livelihoods are farming. Comparative advantage in the agricultural sector, in this case chili, makes chili commodity special in the world trade. In the end, to maintain the country’s economic stability, it is necessary to increase exports and reduce imports [72].

On the other hand, the food crisis in Africa impacts import of food from other countries, especially, Nigeria, the largest producer of the agricultural sector in Africa. Chili as a food and spice needed for the people of Nigeria has decreased in production, which has become a potential for Indonesia to supply chili needs in the country. International trade requires a mutually agreed currency. The rupiah exchange rate must be equal in value to what has been agreed upon. This exchange rate is important to global trade including to Singapore, Malaysia, Saudi Arabia, and Nigeria. Based on the estimation results of the gravity model, the rupiah exchange rate is the variable with the highest and most significant coefficient in influencing the export volume of chili. Efforts to increase exports can be made by increasing the rupiah exchange rate (depreciation) to lower the price of export goods and benefit export destination countries [73]. The depreciation of the rupiah exchange rate will worsen the trade balance in the short term but provide a better condition for the trade balance in the long term. Depreciation causes imported goods to be relatively expensive and the prices of exported goods to be relatively cheaper; with the assumption that changes in the volume of imports and exports take place slowly, it will cause the trade deficit to be larger or the trade surplus to decrease.

However, after some time, export volumes will increase along with relatively lower prices for export products, followed by a decrease in import volumes due to relatively high prices for imported goods. Therefore, the trade balance surplus will increase and exceed the surplus condition before the exchange rate depreciation. The relatively lower price of export chili due to the depreciation of the rupiah exchange rate will benefit export destination countries, especially Nigeria and Singapore, as countries with the highest TPI compared to Malaysia and Saudi Arabia. From this exposure, there is no one-size-fits-all strategy for increasing exports and trade benefits. As has been increasingly often pointed out in the most recent literature, it is crucial to consider the actual economic environment in order to make the best use of existing opportunities.