sign in sign up
<<
Home , Spice trade

Competitiveness in the of : A global evidence

A. Jambor¹; A.T. Toth²; D. Koroshegyi³

1: Corvinus University of Budapest, Department of Agricultural Economics and Rural Development, Hungary, 2: Szent Istvan University, , Hungary, 3: Corvinus University of Budapest, , Hungary Corresponding author email: [email protected] Abstract: Comparative advantage is an important indicator in the analysis of international trade flow, however, in empirical studies on agriculture it is often neglected. In this article we aim to analyse comparative advantage in global spices trade and to test stability of trade indices as well as to identify the determinants behind different country performances. Our paper draws global spices trade data from the period 1991 to 2015. Results suggest that global trade is pretty much concentrated with Guatemala, and obtaining the highest comparative advantages in 1991-2015. However, duration and stability tests indicate that trade advantages have weakened for the majority of the countries concerned. Our model runs show that factor endowments, agricultural value added and regional trade agreements are negatively, while land as well as labour productivity are positively related to comparative advantages in global spices trade. Acknowledegment: This paper was supported by the National Research, Development and Innovation Office Grant No. 119669 titled ‘Competitiveness of Agriculture in International Trade: A Global Perspective’ as well as the UNKP-17-4-III-BCE-7 New National Excellence Program of the Ministry of Capacities of Hungary. JEL Codes: F14, Q02

#1306

Competitiveness in the trade of spices: A global evidence

Abstract

Comparative advantage is an important indicator in the analysis of international trade flow, however, in empirical studies on agriculture it is often neglected. In this article we aim to analyse comparative advantage in global spices trade and to test stability of trade indices as well as to identify the determinants behind different country performances. Our paper draws global spices trade data from the period 1991 to 2015. Results suggest that global spice trade is pretty much concentrated with Guatemala, Sri Lanka and India obtaining the highest comparative advantages in 1991-2015. However, duration and stability tests indicate that trade advantages have weakened for the majority of the countries concerned. Our model runs show that factor endowments, agricultural value added and regional trade agreements are negatively, while land as well as labour productivity are positively related to comparative advantages in global spices trade.

Keywords: comparative advantage, spice trade, determinants, development

JEL code: Q17, F14, Q02

Introduction

Competitiveness is one of the most used and abused word in economics, containing many kinds of different interpretations. One strand of the literature combines international trade theories with those of macro level competitiveness and argues that competitiveness of nations can be interpreted and measures via trade based indices. Balassa (1965) was one of the early supporters of this theory, elaborating his famous index of revealed comparative advantages. Since this seminal work, a vast amount of literature is dedicated to the analyses of revealed comparative advantages of global trade.

Despite the apparent importance of the topic, however, the majority of studies are focused on industrial products, while agri-food sectors are usually neglected in empirical works. The main reason is probably that agricultural markets are usually assumed to be perfectly competitive. The article analyses revealed comparative advantages in global spices trade – this approach, at least to our knowledge, is currently missing from the literature. This paper, therefore, contributes to the existing literature in three ways. First, it applies the theory of revealed comparative advantages on an ‘exotic’ agricultural product group. Second, it analyses a product which is important from a development economic perspective as spices are mainly produced and exported by developing countries. Third, the article aims to identify the factors lying behind comparative advantages.

The article is structured as follows. Section 2 presents an overview of the empirical literature, followed by a demonstration of methodology and data used. Section 4 summarizes the descriptive statistics of global spices trade, identifying key players and products. Section 5 describes the comparative advantage patterns of the major exporters together with stability test for analyzing the duration of their advantages. Section 6 discusses the possible factors behind different competitiveness potentials, while the last section concludes.

1. Empirical evidence

Tthe analysis of comparative advantages of agricultural and food products is limited in the international literature. In a regional context, Chingarande (2013) investigated the comparative advantage of the East-African Community (EAC) Member States and found advantage for some agricultural products such as green tea, coffee, ivory, fish fillet, and flowers. Similarly, Ndayitwayeko et al. (2014) analyzed the comparative advantage of the Eastern and Central African (EAC) coffee sector and revealed that EAC countries, though to a diminishing extent, had comparative advantage in global coffee exports from 2000 to 2012, with Uganda and Kenya leading the group.

In , Kuldilok et al. (2013) found a halt in the decline of Thailand’s tuna market share in global trade. Akmal et al. (2014) analyzed the competitiveness of Pakistan’s basmati rice exports and found that the country was losing its position to world markets in one of its biggest export products, calling for a change in its trade strategy. Astaneh et al. (2014) searched for comparative advantage in ’s stone fruits market and found that the country had strengthened her competitive positions, though it lacked comparative advantage in the majority of the years analyzed.

In , Bojnec and Fertő (2014) analyzed the competitiveness of agri-food exports of European countries, and found majority of countries and products to have an advantage globally. The most successful nations in this regard were the Netherlands, France and Spain. The article also predicted a more long lasting advantage for Western-European countries, compared to Eastern-European ones. Fertő (2008) analyzed the evolution of agri-food trade patterns in Central European Countries and found the trade specialization across the region to be mixed. For particular product groups, greater variation was observed, with stable (unstable) patterns for product groups with comparative disadvantage (advantage). Török and Jámbor (2013) also analyzed the agri-food trade patterns of New Member States, and highlighted that almost all countries experienced a decrease in their comparative advantage after the EU accession, though it still remained at an acceptable level for most cases.

In Latin America and the Caribbean, McLean et al. (2014) investigated regional integration in the Caribbean and found many countries and products to have a comparative advantage and potential to prosper. Korinek and Melatos (2009) analyzed revealed comparative advantages of MERCOSUR countries and found margarine, vegetable oils and coffee as the most competitive products in 1988 to 2004. In particular, Brazil and Argentina are leaders in comparative advantage in beef, both in fresh and preserved form.

In North America, Málaga and Williams (2006) found a lack of comparative advantage in agricultural and food export in . At the product group level, however, results suggested vegetables and fruits to have competitive positions. However, this competitiveness was decreasing for vegetables and increasing for fruits with time. Sparling and Thompson (2011) investigated the Canadian agri-food sector and concluded that despite its overall competitiveness, the country was losing its position in food processing. Sarker and Ratnasena (2014) analyzed the comparative advantages of Canadian wheat, beef and pork sectors between 1961 and 2011, and found only the wheat sector to be competitive.

Disdier al. (2015) analyzed comparative advantages of agri-food products in the Asian and Pacific region and found that Australia and had strong comparative advantages in fruit and vegetables, beverages and the dairy market. This was partly due to the opening of some of their important but overly protected markets (e.g. Canada and Japan). Linehan et al. (2012) also found that Australia has had a strong comparative advantage in her agricultural sector and especially in grains, beef and semi-prepared foods.

On a product basis, we have not found any article analyzing the competitiveness of spices in international trade.

2. Methodology As discussed in the theoretical framework, probably the most well-known index analyzing trade-based competitiveness of nations is Revealed Comparative Advantage (RCA), calculating the proportion of a country’s share of exports for a single commodity to the exports of all commodities and the similar share for a group of selected countries, expressed by Balassa (1965) as follows:

 X ij   X nj  RCAij      (1)  X it   X nt  where, X means export, i indicates a given country, j is a given product, t is a group of products and n is the group of selected countries. Hence, a revealed comparative advantage (or disadvantage) index of exports can be calculated by comparing a given country’s export share by its total exports, with the export share by total exports of a reference group of countries. If RCA>1, a given country has a comparative advantage compared to the reference countries, or in contrast, a revealed comparative disadvantage if RCA<1. Although the Balassa (RCA)-index is criticized from many aspects, it is beyond the scope of this paper to go into details – see Jambor and Babu (2016) for a thorough review on this issue.

The paper also checks the stability and duration of the RCA index in two steps. First, Markov transition probability matrices are calculated and then summarized by using the mobility index, evaluating the mobility across countries and time. Second, following Bojnec and Fertő (2008), survival function S(t) can be estimated by using the non-parametric Kaplan–Meier product limit estimator, pertaining to the product level distribution analysis of the RSCA index. Following Bojnec and Fertő (2008), a sample contains n independent observations denoted (ti; ci), where i = 1, 2 , . . . , n, and ti is the survival time, while ci is the censoring indicator variable C (taking on a value of 1 if failure occurred, and 0 otherwise) of observation i. It is assumed that there are m < n recorded times of failure. We denote the rank-ordered survival times as t(1) < t(2) < … < t(m). For the purpose of our analysis let nj indicate the number of subjects at risk of failing at t(j) and let dj denote the number of observed failures. The Kaplan–Meier estimator of the survival function is then (with the convention that ˆS(t) = 1 if t < t(1)) as follows:

n  d Sˆ(t)   j j (8) t(i)t n j

In order to calculate indices above, the article uses the World Bank WITS software based on COMTRADE, an international trade database developed by the United Nations at the HS six digit level as a source of raw data. Spice trade is defined as trade in the following products. The chapter works with trade data for the period of 1991 to 2015.

However, we are aware that the methodology above has a number of limitations. First, trade data is not fully reliable due to various reasons. These include the following: trade values may not necessarily sum up to the total trade value for a given country dataset; countries may not necessarily report their trade values for each and every year; trade data may differ by the selection of classification; and imports reported by one country may not coincide with exports reported by its trading partner. Second, Balassa-based indices are sensitive to zero values (see equation 1, for instance). Third, outliers in results get omitted, dropping inconsistent indices and some useful data. However, based on the literature review and previous empirical works, our results well fit into past findings.

3. Global spice trade patterns

Global spice trade has been continuously increasing in the previous 25 years (Figure 1). In 2015, for instance, the world traded 8 times more spices ($38 billion) than it did in 1991 ($7 billion), which can be attributed mostly to the growing population (in 2015 the world population almost reached 7.5 billion while in 1991 it was 5.3 billion). This growth was mainly in line with that of agricultural and all products, also showing a significantly increasing trend with total merchandize trade growing faster. This was primarily due to overall relatively higher increase in production of merchandize products as compared to, agricultural and food products. Further, a significant decline in transport and transaction costs has also contributed to this trend. Figure 1: The evolution of global trade of spices, agricultural and all products, 1992-2015 (1991=1), nominal trade value

10 8 6 4 2 0 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014

Agricultural World total Spices

Source: Own composition based on World Bank WITS (2016) database

The analysis of trade in global spices by country gives further insights to the trends above. Ten countries gave the vast majority of global spices trade in the period analysed with changing concentration patterns. As evident from Table 1, India, China and were the biggest spice exporters in the world in 2011-2015, accounting for 38% of global spices trade. The case of Vietnam should be highlighted here as an example: from 1991 to 2000, Vietnam was missing from the list of biggest spice exporters, while it was among the three biggest between 2006 and 2015 with continuously growing positions, mainly due to pepper exports. Note that between 1991 and 2014, the Vietnamese pepper production has increased 15 times (FAO, 2016).

Table 1: Top spice exporters in the world, 1991-2015, in percentage of total spice export (nominal value) 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 14% 13% China 10% India 14% India 18% Singapore 12% India 13% India 8% China 11% China 10% India 12% Indonesia 12% 6% Vietnam 7% Vietnam 10% China 9% China 8% Indonesia 6% Indonesia 6% Indonesia 6% Spain 6% Netherlands 5% Singapore 5% Guatemala 4% Netherlands 5% Germany 5% 4% Vietnam 5% Netherlands 4% Singapore 4% Madagascar 5% Brazil 4% Germany 4% Germany 4% Germany 4% Brazil 4% Spain 4% Netherlands 4% Spain 4% Brazil 4% Netherlands 4% Germany 4% Brazil 3% Brazil 3% Sri Lanka 3% United States 3% Turkey 3% Spain 3% Sri Lanka 3% Guatemala 3% TOP10 74% 70% 54% 60% 67% Source: Own composition based on World Bank WITS (2016) database

Over the last decade, India also has had a significant and growing share in spice exports thanks to its virtual monopoly on spice oils made from tropical aromatics but is also one of the biggest exporters of , and spice mixtures. China, Vietnam and Indonesia are located at the same climate zone so they have very similar and intensive flavoured cuisines based on spices. Rice is the main staple food and is served with side dishes of meat and vegetables. Their fundamental, major ingredients are pepper, paprika, cumin, spice mixtures and (Matthews and Jack, 2011).

As to global spice imports, the United States, Germany and Japan were leading the line in the majority of the period analysed with India being the third in the last sub-period. The concentration of the ten biggest spice importers of the world has continuously been decreasing together with the changes of the countries in the list. Note that the overall concentration of spice imports are lower than that of spice exports, suggesting diversified import patterns.

Table 2: Top Spice Importers, 1991-2015, in percentage of total spice import (nominal value) 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 United United United United United 22% 19% 17% 13% 15% States States States States States Japan 10% Japan 9% Japan 6% Germany 6% Germany 6% Germany 9% Germany 7% Germany 6% Japan 5% India 4% United Singapore 8% Singapore 7% Singapore 5% 3% Netherlands 4% Kingdom Saudi 4% Netherlands 5% France 4% Malaysia 3% Japan 4% Arabia Saudi Spain 4% France 4% Netherlands 4% Netherlands 3% 4% Arabia United United United Netherlands 4% 4% 3% Spain 3% 4% Kingdom Kingdom Kingdom Canada 3% Spain 3% India 3% India 3% Singapore 3% United Saudi Saudi 3% 2% Spain 3% 3% Spain 3% Kingdom Arabia Arabia Saudi Mexico 3% Canada 2% 3% Singapore 3% France 3% Arabia TOP10 70% 62% 54% 45% 50% Source: Own composition based on World Bank database (2016)

The list of biggest spice importers are also related to the overall level of development, suggesting that it is mainly developed economies importing while developing countries exporting spices. Moreover, the case of the Netherlands or Germany also suggest that without being a typical producer, the procurement of raw materials together with their conversion to high value added processed spices can result in leading export positions.

The product structure of global spices trade is also worth to be investigated (Table 3). In 2011- 2015, the biggest spice export products were or badian, , , and cumin, altogether giving 63% of global spices trade, suggesting a high level of concentration (TOP10 products gave 85% in the same period). The product structure of global spice exports has changed little over time. Concentration of these products are also high by country – for instance, Guatemala, India, Nepal, Singapore and Indonesia exported 93% of world’s cardamom in 2011-2015. It is almost the same situation with paprika, ginger, and , coming from relatively few countries.

Table 3: Top spice export products in the world, 1991-2015, in percentage of total spice export (nominal value) 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 Anise or Anise or Anise or 27% 37% Caraway 17% Caraway 19% 24% badian badian badian Anise or Anise or Caraway 18% Caraway 14% 17% 19% Caraway 17% badian badian Cardamon 8% Cardamons 7% Spice mix 10% 9% Cardamons 9% s Curry 6% Ginger 6% Ginger 8% Ginger 9% Ginger 8% Cardamons 6% 4% Cardamons 7% Thyme 6% Cumin 5% and Capsicum Ginger 6% juniper 4% Cumin 6% Curcuma 5% 5% or Pimenta berries Thyme 4% 3% Thyme 4% Nutmeg 5% Thyme 5% Capsicum or Capsicum Nutmeg 3% 3% Nutmeg 4% 4% Nutmeg 5% Pimenta or Pimenta Fennel and Cumin 2% Curry 3% Curcuma 4% juniper 4% Curcuma 4% berries Fennel and Pepper, Fennel and 2% Spice mix 3% juniper 4% crushed or 3% juniper 3% berries ground berries TOP10 82% 84% 81% 83% 85% Source: Own composition based on World Bank database (2016)

Regarding global spice import by product (Table 4), dried pepper, capsicum or pimenta, ginger, pepper and other spices took the lead in 2011-2015, giving 63% of global spices import (TOP10 gave 84% in the same period). Similarly to exports, the product structure of imports has changed little over time.

Given these trends, the rest of the article concentrates on TOP10 major spice exporters and products, representing the vast majority of respective global trade.

Table 4: Top spice import products in the world, 1991-2015, in percentage of total spice import (nominal value)

1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 Capsicum or Dried Capsicum or Capsicum Dried 20% 32% 19% 20% 24% Pimenta pepper Pimenta or Pimenta pepper Dried Capsicum or Dried Dried Capsicum 19% 16% 16% 18% 15% pepper Pimenta pepper pepper or Pimenta Other Vanilla 8% Ginger 7% Vanilla 11% 10% Ginger 9% spices, nes Other Other Ginger 8% 5% Ginger 8% Ginger 8% 8% spices, nes spices, nes Pepper, Other Cinnamon 8% Cinnamon 5% 7% Spice mix 5% crushed or 7% spices, nes ground Other 6% Vanilla 4% Cardamons 5% Cardamons 5% Cloves 5% spices, nes Pepper, Cumin 5% Cardamons 4% Cloves 5% crushed or 5% Spice mix 5% ground Pepper, Cardamon Cardamons 4% crushed or 3% Spice mix 5% Cumin 4% 4% s ground Spice mix 2% Cumin 3% Cinnamon 4% Cinnamon 4% Cinnamon 4% Cloves 2% Spice mix 3% Cumin 4% 4% Cumin 3% TOP10 82% 82% 84% 83% 84% Source: Own composition based on World Bank database (2016)

4. The specialisation of global spices trade

With calculation of Balassa indices, the specialisation of global spices trade becomes apparent (Table 5). The paper only focuses on the original Balassa index due to high correlations (not presented here) among different indices. It is obvious that Guatemala, Sri Lanka and India had the highest Balassa indices in the period analysed, while six countries out of the ten biggest exporters had a comparative advantage in 2011-2015. Indonesia and Singapore experienced the biggest fall in the period analysed, while Germany and the Netherlands had a comparative disadvantage in global spices trade, despite being one of the biggest exporters.

Table 5: Revealed comparative advantage of global spices trade by the Balassa index, 1991-2015, by country Country 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 Brazil 1.27 1.31 1.80 1.17 1.09 China 1.54 1.38 1.17 0.79 0.58 Germany 0.27 0.31 0.28 0.29 0.33 Guatemala 59.50 51.09 74.46 65.96 44.08 India 12.44 12.78 9.79 7.71 6.43 Indonesia 3.83 3.05 2.23 1.93 1.65 Netherlands 0.43 0.56 0.61 0.80 0.99 Singapore 2.14 1.86 1.33 0.89 0.85 Sri Lanka 5.35 6.20 9.98 15.49 13.11 Vietnam 0.00 2.57 4.74 4.88 4.37 Source: Own composition based on World Bank database (2016)

Between 1991 and 2015 the TOP10 gave more than 80 percent of global spice trade.

When analysing comparative advantages by product, further specialisation patterns become available (Table 6). It is apparent that cardamoms have the highest comparative advantages in among product groups, followed by cloves, dried pepper and cumin seeds. Consequently, countries exporting these products (Guatemala-cardamon, Sri Lanka-pepper, India-cumin) had the highest comparative advantages, while concentrating on the export of other spices have not proved to be beneficial.

Table 6: Revealed comparative advantage of global spices trade by the Balassa index, 1991-2015, by product Product 1991-1995 1996-2000 2001-2005 2006-2010 2011-2015 Dried pepper 4.17 4.51 7.71 7.52 7.77 Pepper, crushed or ground 1.02 1.01 1.79 3.59 3.01 Capsicum or Pimenta 0.00 1.90 1.47 1.46 1.40 Vanilla 1.59 1.34 0.66 0.81 0.69 Cloves 3.77 3.62 9.13 13.47 9.48 Cardamons 48.20 63.88 78.57 78.86 56.65 Seeds of cumin 11.35 10.74 8.83 6.25 5.97 Ginger 3.00 2.10 1.10 0.75 0.80 Spice mixtures 1.27 1.31 1.04 1.04 1.14 Other spices 2.41 1.90 1.70 1.32 1.25 Source: Own composition based on World Bank database (2016)

The degree of mobility in Balassa indices is estimated by using the mobility index based on the Markov transition probability matrices (Figure 2). Results show a relatively low mobility of the Balassa index in global spices trade for Germany, the Netherlands and India, suggesting stable patterns of comparative (dis)advantages. Besides these countries, more than 70% of product groups with a comparative advantage remained persistent for China, Sri Lanka and Singapore, while lowest mobility measures pertained to Guatemala, Vietnam, Brazil and Indonesia, implying highly changing competitive potentials (in line with results observable in Table 5).

Figure 2: The mobility of Balassa indices, 1991-2015, by country, % 0 10 20 30 40 50 60 70 80 90 100

Germany

Netherlands

India

China

Sri Lanka

Singapore

Guatemala

Vietnam

Brazil

Indonesia

Source: Own composition based on World Bank database (2016)

Regarding the duration of revealed comparative advantages in global spice exports, the non- parametric Kaplan–Meier product limit estimator was estimated. As described in the methodology section, equation 9 was on our panel dataset and results confirm that in general the survival times are not persistent over the period analysed (Table 7). Survival chances of 99% at the beginning of the period fell to 2-18% by the end of the period, suggesting that fierce competition exists in global spices trade. Results vary by product, though the highest survival times exist for dried pepper and the lowest for Capsicum or Pimenta. The equality of the survival functions across the top 10 countries can be checked using two non-parametric tests (Wilcoxon and log-rank tests). Results of the tests show that the hypothesis of equality across survivor functions can be rejected at the 1% level of significance, meaning that similarities in the duration of comparative advantage across most important global spice exporters are absent (Table 7).

Table 7: Kaplan-Meier survival rates for Balassa indices and tests for equality of survival functions in global spices trade, by most exported product, 1991–2015 Capsicu Spice Survivor Dried Vanill Cardam Other Years Pepper m or Cloves Cumin Ginger mixtur function pepper a ons spices Pimenta es 1991 0.9878 0.9959 0.9878 0.9799 0.9837 0.9837 0.9918 0.9918 0.9919 0.984 0.9837 1992 0.9688 0.9833 0.9709 0.9511 0.9586 0.9671 0.9747 0.975 0.975 0.9635 0.9629 1993 0.9458 0.9702 0.9451 0.9218 0.933 0.9371 0.9485 0.9576 0.9577 0.9384 0.9373 1994 0.9234 0.9523 0.9275 0.8962 0.8983 0.9111 0.9307 0.9398 0.9355 0.9170 0.9156 1995 0.9014 0.9338 0.9048 0.8744 0.8764 0.8801 0.9078 0.9215 0.9173 0.8952 0.889 1996 0.8789 0.9148 0.8818 0.8522 0.8495 0.8533 0.8844 0.9027 0.8986 0.8728 0.8619 1997 0.8559 0.8953 0.8582 0.8296 0.8223 0.8215 0.8605 0.8835 0.8794 0.8544 0.8344 1998 0.8315 0.8753 0.8293 0.8017 0.7947 0.7940 0.836 0.8637 0.8597 0.8307 0.8064 1999 0.8057 0.8599 0.8047 0.7735 0.762 0.7753 0.8110 0.8383 0.8344 0.8014 0.7732 2000 0.7721 0.8384 0.7744 0.7444 0.7191 0.7414 0.7800 0.8016 0.8030 0.7663 0.7297 2001 0.7402 0.816 0.7432 0.7147 0.6808 0.7117 0.7482 0.7696 0.7708 0.7357 0.6908 2002 0.7084 0.7985 0.7164 0.6789 0.6419 0.6812 0.7099 0.7366 0.7376 0.7147 0.6513 2003 0.6771 0.7740 0.6999 0.6424 0.6073 0.655 0.6708 0.7083 0.7034 0.6817 0.6163 2004 0.6452 0.7546 0.6766 0.6103 0.5719 0.6223 0.6421 0.6846 0.6622 0.6533 0.5752 2005 0.6083 0.7272 0.6397 0.5714 0.5303 0.5883 0.6121 0.6535 0.6138 0.6236 0.5386 2006 0.5700 0.6981 0.6077 0.5257 0.4932 0.553 0.5806 0.6143 0.5707 0.5862 0.4955 2007 0.5301 0.6671 0.5807 0.4790 0.4548 0.5162 0.5339 0.5802 0.526 0.5406 0.4569 2008 0.4847 0.6337 0.5371 0.4311 0.4207 0.4775 0.4992 0.5294 0.4665 0.4933 0.4113 2009 0.4335 0.5885 0.4834 0.388 0.3726 0.4365 0.4396 0.4765 0.4061 0.4439 0.3643 2010 0.3852 0.5394 0.4351 0.3298 0.3416 0.3929 0.4010 0.4209 0.3515 0.3922 0.3218 2011 0.3344 0.4963 0.3829 0.2704 0.3074 0.3300 0.3584 0.3704 0.3018 0.3373 0.2703 2012 0.2809 0.4342 0.3254 0.2096 0.2767 0.2723 0.3112 0.3056 0.2559 0.2782 0.2162 2013 0.2199 0.3763 0.2712 0.1467 0.2306 0.2178 0.2556 0.2445 0.1865 0.2133 0.1586 2014 0.1503 0.3011 0.217 0.0807 0.1499 0.1525 0.1884 0.1589 0.1291 0.1387 0.0951 2015 0.0703 0.1806 0.1302 0.0242 0.03 0.0762 0.113 0.0794 0.0612 0.0693 0.0381 Source: Own composition based on World Bank database (2016)

5. Hypotheses and econometric specifications

Latruffe (2010) identified four different methods to investigate the determinants of competitiveness. First, one can run econometric regressions on the competitiveness scores obtained. This is a standard approach, frequently used in the efficiency and productivity literature. Second, some authors prefer to apply simple correlation analysis between competitiveness scores and their determinants without establishing any causality. Third, farm level competitiveness is sometimes analyzed by using different samples and comparing their results with various statistical tests. Finally, farm level competitiveness can also be investigated by using cluster analysis and comparing average values. On the basis of the methods and data available, this article uses regression analysis to illustrate the process of identifying the determinants of competitiveness in global spices trade. The following hypotheses are tested based on previous theoretical and empirical research.

H1: Higher factor endowments increase comparative advantages.

The difference in factor endowments is usually measured by inequality in per capita GDP, in line with Falvey and Kierzkowski (1987). It seems reasonable that higher factor endowments of a country leads to higher comparative advantages based on the higher number of resources available. Factor endowments are proxied by the logarithm of per capita GDP (lnGDPPERCAP), which is expected to be positively related to comparative advantages. Per capita GDP is measured in PPP in constant 2010 US dollars, where the data come from the World Bank WDI database.

H2: Agricultural sector growth leads to higher comparative advantages.

The larger the international market, the larger the opportunities for production of differentiated goods and the larger the opportunities for trade in such goods. The logarithm of agricultural value added (AGVA) is used as a proxy for the average size of sectors/markets. AGVA is measured in PPP in 2010 international dollars and the source of data is also the World Bank WDI database. A positive sign for comparative advantages is expected.

H3: More productive economies in agriculture have better comparative advantages

According to the original theory of comparative advantages, productivity differences leads to specialisation, resulting in different levels of comparative advantages. Productivity here is proxied by land and labor productivity, where agricultural value added is divided by agricultural area or the labor force, respectively. The indices are measured in constant 2010 international dollars per hectare or per persons and data are coming from the World Bank and FAO databases.

H4: Regional trade agreements foster comparative advantages.

According to trade theories, regional trade agreements foster economic integration and growth. Consequently, the article presumes that they also foster comparative advantages by establishing free trade zones, encouraging the full use of available resources via specialisation. The number of regional trade agreements a country has is used here, available from the WTO database.

The paper applies the gravity equation approach to test these hypotheses and identify determinants of comparative advantages of global spices trade in 1991-2015. On the whole, we estimate the following regression model: lnRCAit= α0+ α1lnGDPPERCAPit + α2lnAGVAit + α3lnLANDPRODit + α4lnLABORPRODit + α5RTAit + vi + εit (9) where, i is a unit of observation (country) and t is the time period (year). Vi is an error term suggesting country-level effect that does not vary across time, and 휀it is an error term that varies across countries and time. Table 8 provides a description of variables used in the above model.

Table 8: Description of independent variables Expected Variable Variable description Data source sign Gross Domestic Product per capita in lnGDPPERCAP World Bank + constant 2010 thousand US$ Agricultural value added in constant 2010 lnAGVA World Bank + thousand US$ Land productivity: Agricultural Value World Bank, lnLANDPROD Added/Utilised Agricultural Area + FAO (constant 2010 thousand US$)/ha) Labor productivity: Agricultural Value Added/Total Economically Active World Bank, lnLABORPROD + Population (constant 2010 thousand FAO US$)/person) Number of regional trade agreements in RTA WTO + goods Source: Own composition

Many static panel data techniques are available for testing the competitiveness of global agri- food trade. These include, pooled OLS, fixed effects (FE) and random effects (RE), feasible generalized least squares (FGLS), and the panel-corrected standard errors (PCSE) method. Here we apply the random effects GLS regression method as a static panel estimation technique. We use this because pooled OLS lead to biased results as it does not account for unobserved cross- country heterogeneity (Turkcan and Ates, 2010).

Moreover, Leitao (2012) suggest using dynamic panel framework to solve problems of serial correlation, heteroskedasticity and endogeneity of some explanatory variables. These econometric problems were solved by Arellano and Bover (1995) and Blundell and Bond (1998) who developed the first-differenced GMM (GMM-DIF) estimator and the GMM system (GMM-SYS) estimator. Windmeijer (2005) proposes a finite sample correction that provides more accurate estimates of the variance of the two-step GMM estimator (GMM-SYS). As the t-tests based on these corrected standard errors are found to be more reliable, the paper estimates the coefficients using the finite sample correction.

6. Determinants of competitiveness of global spices trade

Results of our model runs seem to be identical. Based on static and dynamic panel data econometrics, it is obvious that factor endowments, agricultural value added and regional trade agreements are negatively, while land as well as labour productivity are positively related to comparative advantages in global spices trade. In other words, differences in factor endowments, in sector size or in the number of regional trade agreements do not seem to count in acquiring a higher comparative advantage for spices globally. However, higher productivity appears to be a crucial determinant of comparative advantages for spices. On the whole, the vast majority of the variables are significant at all levels with both models generally fitting well to the sample.

Table 9: Determinants of global spices trade competitiveness Variables GLS model GMM-SYS model -0.0340 L1.lnRCA - (0.283) -0.6127 -0.6536 lnGDPPERCAP (0.0000) (0.0000) -0.4910 -0.5186 lnAGVA (0.0000) (0.0000) 0.0729 0.0793 lnLANDPROD (0.0000) (0.0000) 0.7438 0.7044 lnLABORPROD (0.0000) (0.0000) -0.0326 -0.0307 RTA (0.0000) (0.0000) 14.4089 15.5541 Constant (0.0000) (0.0000) Observations 223 223 R-squared 0.8098 - Number of countries 10 10 Source: Own composition based on World Bank database (2016)

From a development perspective, note that most spices are grown in the subtropical and tropical zones where the vast majority of the biggest exporters actually pertain. Spices are almost exclusively grown by small local farmers without any significant investments or capital endowments. Their cultivation is based on traditions and experience, so women in households can easily deal with their gardens near the house. There isn’t any kind of technology for harvesting spices, because spice crops are small, sensitive or vulnerable, so they need almost only manual labour. In some villages local farmers cooperate to represent a bigger and much formal community, so they can easier reach national and international markets as well. They create a coordinated, strategic approach to sell spices – and products made by spices - those, who can help them to meet the international requirements (Matthews and Jack, 2011). These facts can partly explain our results presented above.

On the whole, based on Table 9, all hypotheses except the third are rejected by both model specifications.

Conclusions The patterns, stability and determinants of comparative advantage in global spices trade were analysed in the paper, generating a number of conclusions. First, our results indicate that global spice trade has been continuously increasing in the previous 25 years with a high concentration on both the export and import sides by country and by product. , India, China and Vietnam were the biggest spice exporters in the world in 2011-2015, while the United States, Germany and Japan were leading the line in global spice imports. Most traded products were anise or badian, caraway, cardamoms, ginger and cumin, altogether giving 63% of global spices trade in 2011- 2015, suggesting a high level of concentration (TOP10 products gave 85% in the same period).

Second, our results also suggest that Guatemala, Sri Lanka and India had the highest comparative advantages in the period analysed, while at the product level, cardamoms led the line, followed by cloves, dried pepper and cumin seeds. It seems evident that countries concentrated on the export of these products were the most competitive in global spice markets. Third, duration and stability tests indicated that trade advantages had weakened for the majority of the countries concerned.

Fourth, the article was in search for the determinants of comparative advantages in global spice trade and concluded that factor endowments, agricultural value added and regional trade agreements are negatively, while land as well as labour productivity are positively related to comparative advantages in global spices trade. Research in the future might check other products and variables to extend these results and make them more valid.

References

Akmal, N., Akhtar, W., Shah, H., Niazi, M.A. and Saleem, T. (2014): The Structure and Competitiveness of Pakistan’s Basmati Rice Exports. Asian Journal of Agriculture and Rural Development, 4(4): 304-312.

Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of error- components models. Journal of Econometrics, 68(1): 29–51.

Astaneh, H.K., Yaghoubi, M. and Kalateharabi, V. (2014): Determining Revealed Comparative Advantage and Target Markets for Iran's Stone Fruits. Journal of Agriculture, Science and Technology, 16: 253-264.

Balassa, B. (1965): Trade Liberalization and Revealed Comparative Advantage. The Manchester School, 33: 99-123.

Bhawsar, P. and Chattopadhyay, U. (2015): Competitiveness: Review, Reflections and Directions. Global Business Review, 16(4): 665-679.

Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1): 115–143.

Bojnec, Š., I. Fertő (2008): European Enlargement and Agro-Food Trade. Canadian Journal of Agricultural Economics 56, no. 4: 563-579.

Bojnec, Š., I. Fertő (2014): Agri-Food Export Competitiveness in European Union Countries. Journal of Common Market Studies, no. 3: 476–492

Chingarande, A., Mzumara, M. and Karambakuwa, R. (2013): Comparative Advantage and Economic Performance of East African Community (EAC) Member States. Journal of Economics, 4(1): 39-46.

Dalum, B., Laursen, K. and G. Villumsen (1998): Structural change in OECD export specialisation patterns: De-specialisation and ‘stickiness’, International Review of Applied Economics, 2: 423–443.

Disdier, A-C., Emlinger, C. and Fouré, J. (2015): Atlantic versus Pacific Agreement in Agri- food Sectors: Does the Winner Take it All? Selected Paper prepared for presentation at the 2015 Agricultural & Applied Economics Association and Western Agricultural Economics Association Annual Meeting, San Francisco, CA, July 26-28.

Domazet, T. (2012): Regional cooperation striving for competitiveness and finance. Ekonomika preduzeća, 60(5-6): 290-300.

Falvey, R., Kierzkowski, H. (1987): Product Quality, Intra-Industry Trade and (Im)Perfect Competition, in H. Kierzkowski, eds., Protection and Competition in International Trade. Essays in Honor of W.M. Corden. Oxford University Press, Blackwell.

Fertő, I. (2008): The evolution of agri-food trade patterns in Central European countries. Post- Communist Economies 20, no. 1: 1-10.

Heckscher, E. (1919): The Effect of Foreign Trade on the Distribution of Income. Ekonomisk Tidskrift, 21: 497–512.

Hoen, A.R. and Oosterhaven, J. (2006): On the Measurement of Comparative Advantage. Annals of Regional Science. 40(3): 677-691.

Korinek, J. and Melatos, M. (2009): Trade Impacts of Selected Regional Trade Agreements in Agriculture, OECD Trade Policy Working Papers, No. 87, OECD publishing

Kuldilok, K.S., P.J. Dawson, J. Lingary (2013): The export competitiveness of the tuna industry in Thailand. British Food Journal, no. 3: 328-341.

Latruffe, L. (2010): Competitiveness, Productivity and Efficiency in the Agricultural and Agri- Food Sectors, OECD Food, Agriculture and Fisheries Papers, No. 30, OECD Publishing.

Leitão, N. C. (2012). GMM Estimator: An Application to Intraindustry Trade, Journal of Applied Mathematics, pp. 1-12.

Linehan, V., Thorpe, S., Andrews, N., Kim, Y. and Beaini, F. (2012): Food demand to 2050. Opportunities for Australian Agriculture. Australian Government, Department of Agriculture, Fisheries and Forestry (ABARES). Paper presented at the 42nd ABARES Outlook conference 6–7 March 2012, Canberra, ACT.

Lucas, R. (1988): On the mechanics of economic development. Journal of Monetary Economics, 22: 3-42.

Málaga, J.E. and Williams, G.W. (2006): Mexican agricultural and food export competitiveness. TAMRC International Market Research Report No. IM-01-06.

Matthews, M. and Jack, M. 2011. Spices and for home and market. FAO Diversification Booklet No. 20, FAO, Rome, Italy

McLean, S., Pantin, M. and Skerrette, N. (2014): Regional integration in the Caribbean. The role of trade agreements and structural transformation. ECLAC Studies and Perspectives Series no. 37.

Ndayitwayeko, W.M., Odhiambo, M.O., Korir, M., Nyangweso, P.M. and Chepng’Eno, W. (2014): Comparative Advantage of the Eastern and Central in the Coffee Export Sector: the Case of Burundi. African Crop Science Journal, 22(4): 987-995.

Ohlin, B. (1933): Interregional and International Trade. Harvard University Press, Cambridge, USA.

Porter, M. E. (2004): Competitive Advantage: creating and sustaining superior performance. New York: Free Press.

Proudman, J. and Redding, S., 1998. Openness and growth. London: The Bank of England.

Ricardo, D. (1817): On the Principles of Political Economy and Taxation. John Murray, London, United Kingdom.

Romer, P. (1990): Endogenous technological change. Journal of Political Economy, 98: 71- 102.

Sarker, R. and Ratnasena, S. (2014): Revealed Comparative Advantage and Half-a-Century Competitiveness of Canadian Agriculture: A Case Study of Wheat, Beef, and Pork Sectors. Canadian Journal of Agricultural Economics, 62(4): 519-554.

Smith, A. (1776): An Inquiry into the Nature and Causes of The Wealth of Nations. Methuen & Co., London, United Kingdom.

Sparling, D. and Thimpson, S. (2011): Competitiveness of the Canadian Agri-Food Sector. The Canadian Agri-Food Policy Institute.

Török, Á., A. Jámbor (2013): Agri-Food Trade of the New Member States since EU Accession. Agricultural Economics (Czech), no. 3: 101–112.

Turkcan, K. and Ates, A. (2010). Structure and Determinants of Intra-Industry Trade in the U.S. Auto-Industry. Journal of International and Global Economic Studies, 2(2): 15-46.

Vollrath, T. L. (1991): A Theoretical Evaluation of Alternative Trade Intensity Measures of Revealed Comparative Advantage. Weltwirtschaftliches Archiv 130(2): 265–279.

Wijnands, J. H. M., Bremmers, H. J., Van Der Meulen, B. M. J., & Poppe, K. J. (2008): An economic and legal assessment of the EU food industry's competitiveness. Agribusiness, 24(4): 417-439.

Windmeijer, F. (2005). A finite sample correction for the variance of linear efficient two-step GMM estimators. Journal of Econometrics, 126(1): 25–51.

World Economic Forum (2015): The Global Competitiveness Report 2014-2015. Geneva, Switzerland.

World Economic Forum (1996): The Global Competitiveness Report 1995-1996. Geneva, Switzerland.

Yap, J.T. (2004): A Note on the Competitiveness Debate. Philippine Journal of Development, 57(1): 1-23.

Appendix

Appendix 1: Spice product codes and associated descriptions at the HS6 level HS6 Product code Description 090411 Dried pepper (exc. crushed nor ground) 090412 Pepper, crushed or ground Fruits of the genus Capsicum or Pimenta, dried, neither crushed 090420 nor ground 090500 Vanilla 090610 Cinnamon, neither crushed nor ground 090620 Cinnamon and cinnamon-tree flowers, crushed or ground 090700 Cloves (whole fruit, cloves and stems) 090810 Nutmeg 090820 Mace 090830 Cardamons 090910 Seeds of anise or badian 090920 Seeds of 090930 Seeds of cumin 090940 Seeds of caraway 090950 Seeds of fennel; juniper berries 091010 Ginger 091020 Saffron 091030 (curcuma) 091040 Thyme, bay leaves 091050 Curry 091091 Spice mixtures 091099 Other spices, nes Source: Own composition based on World Bank database (2016)