Trade Potential between India and Central Asia

Abstract

South–South bilateral trade partnerships have been growing increasingly important, and this article analyses the current patterns of bilateral trade between India and five Central Asian countries and the potential for improving them. We apply the gravity model of trade to construct a trade potential index from an Indian perspective and assess the potential for improving India’s trade with Central Asia. We find that the volume of trade between India and Central Asia in 2015 could be six to ten times greater than the actual volume. Political disturbances in the countries that fall along the direct trade route are a major barrier to trade in this region; therefore, we re-examine the potential for trade using alternate routes via Iran and China. We find that the potential for greater bilateral trade between India and Central Asia continues to be positive.

JEL Classification: F10, F14

Keywords

India Central Asia Trade Gravity Model Trade Potential Trade Routes

1. Introduction

The recent bilateral agreements between India and five Central Asian countries has placed a spotlight on trade in this region.¹ However, there is limited documentation about bilateral trade patterns between these countries. One may hypothesise that the volume of current bilateral trade is below the trade potential of this region due to the geopolitical situation in South Asia. We test this hypothesis using a gravity model framework to estimate the untapped potential of bilateral trade between India and five Central Asian nations—Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan. Further, the article attempts to quantify the loss in trade potential due of the lack of an accessible direct trade route.

Given the prolonged recession in Europe and the USA, South–South bilateral trading relationships are gaining importance, and nations would be prudent to diversify their trading partners and explore trading opportunities with emerging economies. India has recognised the need for developing stronger trade relations with other Asian nations and has made several attempts to develop free trade agreements with countries in South Asia, South East Asia and the Asia-Pacific region.² To improve bilateral trade between any two nations, it is necessary to address issues that are neither merely economic nor bilateral, but are entwined with geopolitical relationships in the entire region. One dimension of South–South trade that India has not yet explored fully is the potential for trade with Central Asia.

India, as well as the Central Asian nations, experienced a turning point in their economic history in the 1991. When the USSR collapsed in 1991, the Central Asian nations were formed, and they initially faced a severe economic downturn. After 1991, these economies became market-oriented—industries were privatised and globalised. Incidentally, India too adopted structural reforms, including liberalisation policies, in 1991. India and Central Asia faced a common challenge in 1991—how to integrate into the world economy? Each country, however, had a different set of resources and different economic circumstances; in turn, they met this challenge with different degrees of success (Dowling and Wignaraja, 2006a).

While there have been differences and similarities in the growth experiences of these countries since 1991, the 2008 global recession affected all. However, 5-year average annual growth rates continue to be stable in this region (Table 1). Hence, for India, one strategy to cope with recession in the west is to add diversity to its set trade and investment partners by developing stronger South–South ties with Central Asia.

Although India’s proximity to the Central Asian countries is obstructed by its unfavourable political relationship with Pakistan, trade in goods between India and Central Asia has increased manifold over the past decade. Nevertheless, Central Asia’s exports to India averaged only about 1 per cent of their total exports and did not exceed 2 per cent for any of the five countries (Figure 3). The average share of Indian exports in Central Asia’s imports is even lower (Figure 4). In fact, as it stands now, India is not a key trading partner for any of the five Central Asian countries. Developing a good trade relationship with the Central Asian countries is of high strategic importance for India, since these countries are sandwiched between major powers, such as Russia, China and Europe, and have enormous amounts of energy resources that could help meet India’s energy requirements and sustain its high economic growth. Given the volatile global economy, there is a great need to forge trade relationships with a range of developing countries. Cooperation is essential to maintain peace and harmony in the region, avoid terrorism and illegal activities, and promote mutually beneficial economic trade and investment ties (Rakhimov, 2010).

Table 1
Average Annual Growth Rate of India and Central Asian Nations (per cent)

India Kazakhstan Kyrgyz Republic Tajikistan Turkmenistan Uzbekistan

1996–2000 6.09 5.64 2.56 4.87 0.46 3.86

2001–2005 6.74 3.84 10.38 5.18 9.80 5.42

2006–2010 8.10 4.49 6.28 10.41 6.60 8.56

2011–2015 6.74 4.86 4.58 10.56 6.64 8.12

1995–2015 6.92 4.71 5.95 7.76 5.87 6.49

	India	Kazakhstan	Kyrgyz Republic	Tajikistan	Turkmenistan	Uzbekistan
1996–2000	6.09	5.64	2.56	4.87	0.46	3.86
2001–2005	6.74	3.84	10.38	5.18	9.80	5.42
2006–2010	8.10	4.49	6.28	10.41	6.60	8.56
2011–2015	6.74	4.86	4.58	10.56	6.64	8.12
1995–2015	6.92	4.71	5.95	7.76	5.87	6.49

Source: Author’s calculation using World Bank—World Development Indicators Dataset.

Note: GDP refers to gross domestic product, PPP (constant 2011 international $).

This is, to our knowledge, the first study to empirically estimate the potential to increase bilateral trade between India and Central Asia using a gravity model framework. We find that bilateral trade between India and the Central Asian countries could be six to nine times higher, given their economic conditions in 2015. Hence, investing in policies that promote trade in this region may lead to substantial gains. Geopolitical constraints render the direct route between India and Central Asia impassable and constitute a barrier to trade. In this article, we estimate the magnitude of the loss of trade due to this barrier. We attempt to re-evaluate the trade potential within this constraint and demonstrate that there is potential to increase trade substantially even via alternative trade routes that bypass politically unstable areas in the region. Our assessment of the positive trade potential between these nations provides scope for stepping up regional and bilateral trade agreements and investing in other policies that encourage trade in this region.

Figure 1

Source: International Monetary Fund—Direction of Trade Tables, 2016.

Figure 2

Source: International Monetary Fund—Direction of Trade Tables, 2016.

Figure 3

Source: Author’s calculations using International Monetary Fund—Direction of Trade Tables, 2016.

Figure 4

Source: Author’s calculations using International Monetary Fund—Direction of Trade Tables, 2016.

In the second section, we discuss the economies of the five Central Asian nations and India in depth—since economic growth is possibly the most critical factor for any nation to increase international trade—and analyse the patterns of bilateral trade between India and Central Asia including commodity-level trade patterns. In third section, we present the gravity model framework and the empirical strategies to calculate the trade potential index using this model, and present the results. In fourth section, we discuss alternative trading routes to circumvent political disturbances. Finally, in fifth section, we present our concluding remarks.

2. The Economy and Trading Patterns of the Central Asian Countries and India

To provide context to this analysis of trade potential between the countries under study, we discuss their historical growth patterns and the current state of their economies. We also study the aggregated trends in bilateral trade between India and each of the Central Asian nations.³

2.1. Growth Patterns in Central Asia and India

Both the exporter and importer countries need to have high economic growth for the expansion of international trade. A country that has a high GDP has the ability to produce more, and hence, export more. Also, a country with a high GDP will have the ability to consume, and hence, import more. International trade also feeds the economic growth of nations as it enables improvements in productivity and specialisation.

In the first half of the 1990s, economic growth declined sharply and even became negative in the Central Asian countries as they were dealing with the aftermath of the USSR’s dissolution. When they were part of the USSR, they mostly supplied raw materials and energy to the USSR economy; for example, Hiro (2010) discusses how Uzbekistan had a ‘cotton monoculture’—it produced cotton, but had no textile industries. It processed barely 15 per cent of its total cotton output. Most manufacturing industries were located elsewhere in the USSR. This region never developed an independent industrial sector or managerial skills. International trade was planned and executed in Moscow (Dowling & Wignaraja, 2006b). Therefore, after the USSR collapsed, the Central Asian economies lost self-sufficiency, and the supply chains for consumption goods and raw materials were destroyed. Even from a geographic perspective, these countries were landlocked and isolated from the trading routes. The lack of access to resources resulted in increasing poverty, making it even more difficult for them to focus on industrial growth.

The end of the 1990s was a turnaround period, and most Central Asian economies began to report positive economic growth. Since then, the five economies have evolved in different ways, but their growth rates have remained fairly high. Kazakhstan has the highest GDP per capita in the region.⁴ The Kazakh economy switched from negative to positive growth in 1998 and has managed to sustain its economic progress (Agrawal, 2008). It has also the largest geographical area and a relatively small population. Kazakhstan is rich in oil and uranium and has a strong agriculture sector. Its average annual growth rate between 1991 and 2015 was 3.4 per cent. In contrast, Kyrgyzstan is the least developed country among the Central Asian nations—its GDP grew at 1.6 per cent on average over the same period. The absolute value of Kazakhstan’s GDP in 2015 was 22 times that of Kyrgyzstan.

Tajikistan is the other relatively poor country in this region. Although its growth rate has been increasing, its GDP per capita has been consistently the lowest among the five Central Asian countries. Its average growth rate was high during 2001–2005, primarily due to a very low initial GDP. After 2007, however, the oil price shock and global recession affected Tajikistan, the nation with the smallest area. It has a relatively higher population and the highest population density in this region apart from Uzbekistan.

Turkmenistan is the second-most developed economy after Kazakhstan in the region in terms of GDP per capita. Its average annual GDP growth rate over 1995–2015 was highest in the region at 7.6 per cent. The GDP per capita of Turkmenistan is five times that of Tajikistan and Kyrgyzstan. Turkmenistan has a relatively low population density. The country is large, but the area mostly comprises desert. While the growth rate fell during 2009, the country has managed to recover reasonably fast.

Uzbekistan is the most populous country in Central Asia and has the highest population density. While its GDP growth rate was negative prior to 1996, it has been growing steadily since then. The growth rate rose to over 8 per cent on average between 2006 and 2015. Uzbekistan ranks second in the region in terms of absolute value of GDP, but its GDP per capita is much lower than that of Kazakhstan or Turkmenistan.

Around the same time that the Central Asian nations were created, India enacted liberalisation policies following a balance of payments crisis. This ended industrial licensing; let the private sector participate in sectors, such as telecom, banking and infrastructure; and opened the economy to foreign trade and investment. In the period between 1996 and 2000, the average growth rate of the Indian economy was around 6 per cent; it increased to 8.1 per cent on average over 2006–2010. However, in the last 5 years (2010–2015), India’s average annual growth rate has dropped to 6.74 per cent. India also has a large geographical area, which is comparable to Kazakhstan, a large population (about 1.2 billion in 2012), and a huge supply of manpower. The population density of India is six times that of Uzbekistan, the most populated country in Central Asia, and about 65 times that of Kazakhstan. This reflects India’s potential as a trading partner for labour-intensive commodities and as a market for Central Asia and the world.

2.2. Evolving Patterns of Trade between India and Central Asia

To understand the potential for trade between India and Central Asia, we need to identify past patterns of bilateral trade between these nations. Therefore, we present the historical evolution of bilateral trade since 2000. The value of Central Asian exports are heavily driven by the prices of commodities in the international market. The export performance of Kazakhstan was driven by a favourable rise in prices (Pomfret, 2005). In the case of Uzbekistan, world cotton prices increased until 1996 and have dropped since then, and this correlates with the value of exports from the country. The circumstances in Turkmenistan have been much more complex.

Turkmenistan’s exports are dominated by energy products and the 1995–1996 values are inflated by over-reporting of natural gas exports to CIS destinations which were not paid for (the invoice value was recorded as exports, while the accumulating payment arrears were recorded as foreign assets); recognizing that the bills would never be paid in full, Turkmenistan stopped supplying gas in March 1997, after which export values (and GDP) collapsed until the flow was resumed in March 1999 (Pomfret, 2005).

Over the past decade or so, Central Asia’s imports and exports have steadily increased for the most part. In Kazakhstan, one can observe a consistent increase in exports between 2003 and 2014, except for a dip in the 2009 and 2015 (Table 2). The likely cause for this phenomenon was the recession in Europe and the USA and the decline in crude oil prices. Similarly, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan increased exports during the reference years, except for 2009–2010. This was followed by a recovery; however, another decline took place in 2015 in the case of Kazakhstan, Turkmenistan and Uzbekistan, largely due to the decline in crude oil prices, which is their major export. In 2015, the fall in exports was rather sharp in Kazakhstan, which has traditionally been a large exporter of crude oil. Imports have remained fairly stable in contrast, except in 2015 (Table 3).

Table 2
Total Exports of Central Asian Countries and India (USD billion)

India Kazakhstan Kyrgyz Republic Tajikistan Turkmenistan Uzbekistan

1996 32.32 5.93 0.51 0.77 1.69 2.62

1997 34.62 6.50 0.61 0.80 0.75 2.90

1998 33.67 5.51 0.51 0.60 0.59 2.31

1999 35.92 5.60 0.45 0.69 1.19 1.96

2000 42.63 9.88 0.50 0.77 2.51 2.18

2001 43.31 9.09 0.48 0.65 2.56 2.09

2002 50.50 9.67 0.49 0.74 2.82 1.55

2003 61.12 12.31 0.58 0.80 3.45 2.02

2004 75.39 20.10 0.71 0.91 3.53 2.73

2005 98.20 27.86 0.63 0.91 5.00 3.61

2006 120.53 38.26 0.80 1.40 5.70 5.26

2007 153.77 47.76 1.13 1.47 6.36 6.31

2008 177.70 71.20 1.62 1.44 10.27 7.51

2009 165.18 43.21 0.90 1.01 3.11 4.80

2010 222.92 56.97 1.26 1.19 3.32 5.68

2011 307.09 80.52 1.17 1.01 7.76 5.78

2012 297.26 85.45 1.50 1.15 10.66 5.93

2013 315.13 76.43 1.46 1.01 11.99 7.86

2014 317.73 72.98 1.52 0.85 12.39 6.86

2015 266.17 35.27 1.54 0.94 10.38 6.55

	India	Kazakhstan	Kyrgyz Republic	Tajikistan	Turkmenistan	Uzbekistan
1996	32.32	5.93	0.51	0.77	1.69	2.62
1997	34.62	6.50	0.61	0.80	0.75	2.90
1998	33.67	5.51	0.51	0.60	0.59	2.31
1999	35.92	5.60	0.45	0.69	1.19	1.96
2000	42.63	9.88	0.50	0.77	2.51	2.18
2001	43.31	9.09	0.48	0.65	2.56	2.09
2002	50.50	9.67	0.49	0.74	2.82	1.55
2003	61.12	12.31	0.58	0.80	3.45	2.02
2004	75.39	20.10	0.71	0.91	3.53	2.73
2005	98.20	27.86	0.63	0.91	5.00	3.61
2006	120.53	38.26	0.80	1.40	5.70	5.26
2007	153.77	47.76	1.13	1.47	6.36	6.31
2008	177.70	71.20	1.62	1.44	10.27	7.51
2009	165.18	43.21	0.90	1.01	3.11	4.80
2010	222.92	56.97	1.26	1.19	3.32	5.68
2011	307.09	80.52	1.17	1.01	7.76	5.78
2012	297.26	85.45	1.50	1.15	10.66	5.93
2013	315.13	76.43	1.46	1.01	11.99	7.86
2014	317.73	72.98	1.52	0.85	12.39	6.86
2015	266.17	35.27	1.54	0.94	10.38	6.55

Source: International Monetary Fund—Direction of Trade Tables, 2016.

Of the five Central Asian countries, India has had the largest value of trade with Kazakhstan, the largest economy in Central Asia. Kazakhstan’s exports to India rose after 2005 (Figure 1). There was a spike in Kazakhstan’s exports of to India in 2014, and this was primarily in crude oil. This might have been an outcome of specific bilateral agreements between the Ministry of Petroleum and Natural Gas, Government of India, and the Ministry of Energy, Government of Kazakhstan.⁵

The value of Kazakhstan’s imports from India has been consistently increasing since 2012 (Figure 2). The share of Uzbekistan’s Indian imports has also been increasing, but the rise is not as sharp as that for Kazakhstan. The rest of the Central Asian countries have not observed much of an increase, but in general, Central Asia’s imports from India are higher than the corresponding exports. In 2014–2015, there was a decline in the value of India’s exports to all the Central Asian countries (Figure 2). This drop in the value of trade between India and Central Asia in 2015–2016 is due to the fall in crude oil prices, because of which the value of Central Asia’s exports has declined. Consequently, these countries have lower export earnings, and this decline in purchasing power is likely to have subsequently affected imports into the nation.

Table 3

Total Imports of Central Asian Countries and India (USD billion)

	India	Kazakhstan	Kyrgyz Republic	Tajikistan	Turkmenistan	Uzbekistan
1996	36.05	4.25	0.80	0.67	1.31	4.85
1997	40.90	4.30	0.71	0.75	1.23	4.54
1998	42.16	4.37	0.84	0.71	1.01	2.93
1999	47.90	3.69	0.61	0.66	1.48	2.48
2000	50.34	5.06	0.55	0.67	1.79	2.07
2001	50.14	6.49	0.47	0.69	2.21	2.29
2002	58.91	6.60	0.59	0.72	2.13	2.08
2003	74.07	8.40	0.72	0.88	2.51	2.48
2004	99.84	12.80	0.94	1.19	2.68	3.16
2005	139.75	17.39	1.11	1.33	2.62	3.57
2006	176.60	23.71	1.71	1.73	2.55	4.49
2007	235.02	32.80	2.42	2.55	3.39	7.03
2008	281.47	37.93	4.07	3.27	5.51	10.13
2009	257.66	28.44	8.22	2.51	6.61	9.03
2010	350.78	23.66	7.24	2.66	5.74	9.31
2011	465.08	21.18	9.10	4.78	7.93	10.83
2012	490.41	26.75	10.18	4.72	9.95	11.99
2013	467.95	30.40	10.65	4.91	9.81	14.03
2014	460.51	26.90	10.55	5.93	10.17	14.74
2015	404.06	35.73	8.36	4.68	8.20	11.81

Source: International Monetary Fund—Direction of Trade Tables, 2016.

Despite the increase in trade levels over the last decade, India’s share in the trade of Central Asian countries, relative to their trade with the rest of the world, remains very low (Figures 3 and 4). The share of Tajikistan and Uzbekistan’s exports to India is less than 2 per cent of their total exports, and for the remaining Central Asian countries, it is less than 0.5 per cent. India’s share in the countries’ imports is less than 1.5 per cent; the only exception is Kyrgyzstan in 2003 and 2004. However, in the case of Kazakhstan, though its total exports to the rest of the world have declined, its bilateral exports to India have increased. This is a sign of the increasing importance of Kazakhstan’s trade with India.

Recent data show that pharmaceutical products, apparel, tea and coffee, and various kinds of machinery make up India’s exports to Central Asia. A list of India’s top exports to Central Asia at an industry (i.e., 2 digit HS Code) level in the year 2015–2016 is presented in Table 4. Mineral fuels from Kazakhstan and Turkmenistan largely accounted for imports. Fertilisers, vegetables and other metals and minerals are also some of India’s main imports from Central Asia. A selection of India’s top imports at an industry level from Central Asia in the year 2015–16 is listed in Table 5.

Table 4

A Selection of India’s Top Exports to Central Asia in 2015–16 (USD million)

HS Code	Product*	Kazakhstan	Kyrgyzstan	Tajikistan	Turkmenistan	Uzbekistan
30	Pharmaceutical products	46.4	6.88	13.5	18	47.9
9	Coffee, tea, mate, and spices	31.2	0.51	0.62	0.3	1.41
84	Machinery, mechanical appliances, etc.	17.5	0.42	0.25	8.7	13
61	Apparel (knitted)	13.4	6.4	0.41		2.58
90	Optical, medical instruments, etc.	12.6	0.31	0.16	0.5	2.88
69	Ceramic products	4.33			0	1.61
62	Apparel (not knitted)	4.29	6.76	0.27		4.33
85	Electrical machinery and equipment	2.35	0.6	0.43	1.2	2.96
29	Organic chemicals	0.84	0.14	0.29	0.5	1.83
42	Articles of leather, etc.	2.25	0.74			0.04

Source: Directorate General of Commercial Information and Statistics—Kolkata (2017).

Note: *For details of the product codes, please see Appendix A.

Table 5

A Selection of India’s Top Imports from Central Asia in 2015–16 (million USD)

HS Code	Product*	Kazakhstan	Kyrgyzstan	Tajikistan	Turkmenistan	Uzbekistan
27	Mineral fuels, oils, etc.	261			38
25	Salt; sulphur; etc.	61.1			2.1	0.2
72	Iron and steel	11.4
79	Zinc	9.98
28	Inorganic chemicals	5.25			1.7
7	Edible vegetables, etc.	1.92	1.31			16.8
26	Ores, slag, and ash	0.55	0.25	9.72
13	Lac, gums, resins, etc.	0.04		0.2		5.88
52	Cotton				4.5
31	Fertilisers					18.1

Source: Directorate General of Commercial Information and Statistics—Kolkata (2017).

Note: *For details of the product codes, please see Appendix A.

Tables 6 and 7 give a disaggregated picture of the top five exports and imports at the commodity (i.e., six digit HS Codes) level. The top commodity exported to these countries by India is the HS code 300490, which is a category of medicine for retail sale. Apart from commodities from the pharmaceutical industry, various kinds of machinery, instruments, parts and components of machinery make it to the list of the top commodities that are exported by India to Central Asia. In line with the findings above, commodities made of cotton and synthetic fibres are important exports of India to Kyrgyzstan. Petroleum oils are India’s largest imports from Kazakhstan and Turkmenistan while antimony ores and concentrates are the top imports from Tajikistan. The largest exports of Kyrgyzstan and Uzbekistan are different types of beans to India.

Table 6

A Selection of India’s Top Exports to Central Asia in 2015–16 (million USD)

HS CODE	Description	Value
KAZAKHSTAN
300490	Other medicine put up for retail sale	32.62
090240	Other black tea (fermented) and other partly fermented tea	29.92
901580	Other instruments and appliances	10.9
300420	Medicaments contacting other antibiotics and put up for retail sale	9.52
841950	Heat exchange unit	4.66
KYRGYSTAN
300490	Other medicine put up for retail sale	4.05
610910	T-shirts etc., of cotton	2.46
620443	Dresses of synthetic fibres	1.49
620630	Blouses, shirts and shirts-blouses of cotton	1.22
300420	Medicaments containing other antibiotics and put up for retail sale	1.16
TAJIKISTAN
300490	Other medicine put up for retail sale	8.65
020230	Meat of bovine animals, frozen boneless	3.59
300220	Vaccines for human medicine	2.35
730890	Other structures and parts of structures etc.	1.55
300450	Other medicaments containing vitamins or other products of heading 2936	1.01
TURKMENISTAN
020230	Meat of bovine animals, frozen boneless	33.62
300490	Other medicine put up for retail sale	14.14
840219	Other vapour generating boilers, including hybrid boilers	3.22
300420	Medicaments containing other antibiotics and put up for retail sale	1.54
842199	Other parts of filtering/purifying machinery	1.38
UZBEKISTAN
300490	Other medicine put up for retail sale	30.44
300220	Vaccines for human medicine	6.56
870600	Chassis fitted with engines, for motor vehicles of headings nos 8701 to 8705	4.83
300420	Medicaments containing other antibiotics and put up for retail sale	4.27
300450	Other medicaments containing vitamins or other products of heading 2936	3.02

Source: DGCIS—Kolkata (2017).

In the future, India may try to expand its exports in the pharmaceutical sector and in various kinds of machinery. At the same time, India’s imports of raw materials, metals and minerals may increase in the process. In the agriculture and allied industries sector, coffee, tea and spices are commodities that are in high demand in Central Asia. Another potential value chain that could be strengthened is the apparels sector. India can import cotton, wool and silk from Central Asia and export apparel.⁶

India is an emerging market in the process of becoming industrialised and Central Asia is rich in minerals and raw materials. At the same time, Central Asian countries are small economies, and they depend on foreign trade to meet many of their needs. This situation indicates the immense potential for trade between India and Central Asia.

3. Trade Potential Between India and Central Asia: An Analysis using the Gravity Model

This section briefly describes the gravity model of trade, which is used to analyse bilateral trade between India and Central Asia, its relevance to this research study and to create the trade potential index.

3.1. The Gravity Model Framework

The gravity model has been used extensively to study bilateral trade patterns and determinants.⁷ In its basic form, the model suggests that bilateral trade between two countries is directly proportional to the GDP of both nations and inversely proportional to the distance between them. If the two trading partners are i and j, Trade_ij indicates flow of commodities from i to j. Then the gravity model can be represented as,

T r a d e_{i j} = α G D P_{i}^{β_{1}} G D P_{j}^{β_{2}} / D i s \tan c e_{i j}^{β_{3}}

(1)

As elaborated by Bergstrand and Egger (2011), this equation explains bilateral trade flows reasonably well. This model has received theoretical backing in the form of different research papers, including Anderson and Van Wincoop (2003).

Table 7

A Selection of India’s Top Imports from Central Asia in 2015–16 (million USD)

HS CODE	Description	Value
KAZAKHSTAN
270900	Petroleum oils and oils obtained from bituminous minerals, crude	261.03
252490	Asbestos; other than crocidolite (blue asbestos)	61.12
720241	Ferro-alloys; ferro-chromium, containing by weight more than 4% of carbon	6.15
790112	Zinc; unwrought, (not alloyed), containing by weight less than 99.99% of zinc	5.29
790111	Zinc; unwrought, (not alloyed), containing by weight 99.99% or more of zinc	4.69
KYRGYZSTAN
071333	Vegetables, leguminous; kidney beans, including white pea beans (phaseolus vulgaris), shelled, whether or not skinned or split, dried	1.31
760120	Aluminium alloys	0.25
080252	Pistachios: shelled	0.11
080251	Pistachios: in shell	0.05
847990	Machines and mechanical appliances; parts, of those having individual functions	0.04
TAJIKISTAN
261710	Antimony ores and concentrates	9.72
130190	Natural gums, resins, gum-resins and oleoresins, n.e.c.	0.2
081310	Apricots, dried	0.04
850440	Electrical static converters	0.01
901812	Ultrasonic scanning apparatus	0.01
TURKMENISTAN
270900	Petroleum oils and oils obtained from bituminous minerals crude	38.27
520100	Cotton, not carded or combed	4.49
250300	Sulphur of all knds othr thn sublimed sulphur pcptd sulphur and colloidal sulphur	2.14
280120	Iodine	1.71
390210	Polypropylene	0.22
UZBEKISTAN
071331	Vegetables, leguminous; beans of the species vigna mungo (l.) hepper or vigna radiata (l.) wilczek, shelled, whether or not skinned or split, dried	16.8
310420	Potassium chloride	9.24
310210	Urea, whether or not in aqueous solution	8.84
130190	Natural gums, resins, gum-resins and oleoresins, n.e.c.	5.87
780110	Refined lead	1.92

Source: Directorate General of Commercial Information and Statistics—Kolkata (2017).

The GDPs of the two countries represent the size of the economies engaged in bilateral trade. Multiplication of the masses of two entities makes this model analogous to Newton’s Law of Gravitation. However, in the gravity model in international economics literature, GDP_i reflects the output of the economy and acts as a proxy for production capabilities in the exporting country. It is the aggregation of the supply of traded goods across all sectors in the exporting nation. In contrast, GDP_j represents income or purchasing power in the importing country. It denotes the size of the market for commodities—countries with a higher GDP are likely to have a greater demand for goods. Although advances in communication and transportation technologies have facilitated international trade, several factors still inhibit it. The cost of transportation is an obvious barrier. The variable ‘distance’ is a proxy for cost of transportation (Anderson & Van Wincoop, 2003; Head & Mayer, 2014).

3.2. Gravity Model Estimation

This article contributes to a small and emerging area in the literature on the analysis of trade potentials. The methodology of this article is similar to that of Kabir and Salim (2011), which attempts to understand the potential for economic integration between the European Union (EU) and the Association for South East Asian Nations (ASEAN). In their paper, the authors analyse the level of trade integration between the ASEAN and EU and assess the untapped trade potential. They suggest that the projected potential for trade is the amount of trade that can take place between ASEAN and EU if they can attain the same level of regional trade integration as achieved by the countries within the EU. They use the coefficients of a gravity model of intra-EU trade to generate a benchmark against which potential trade is estimated. They calculate the ‘undiscovered trade potential’, which is the ratio of trade potential to actual trade, and find that the trade potential in the case of the ASEAN and EU is both substantial and untapped.

Batra (2006) uses the gravity model to estimate India’s trade potential. He uses global trade flows to estimate the benchmark against which India’s trade potential is measured. He aims to find the trading partner with which India has the greatest trading potential and finds that India has the maximum trading potential with the Asia-Pacific region and with China. In his analysis, he also demonstrates that India has the potential of increasing trade with Turkmenistan, Uzbekistan and Tajikistan about tenfold. However, the study uses a cross-sectional analysis that does not reveal the direction of change in trade potential. Finally, another paper that uses the gravity model to predict bilateral trade and assess trade potential for Hungary, Poland and the Czech Republic is Jakab, Kovacs and Oszlay (2001). They find that Poland and the Czech Republic had positive trade potential in 1997 and demonstrate that the direction of trade had changed from east to west in the case of Hungary and Poland. They highlight the importance of analysing the dynamics of actual and potential trade, rather than their point estimates, for accurate results.

In this article, we use a strategy that is similar to this literature. A gravity model analysis is undertaken for India’s bilateral trade with the rest of the world between 1996 and 2015 to derive our benchmark estimates. We use the log linear version of the gravity model to derive our estimation equation, as shown below:

\ln T r a d e_{i j t} = α + β_{1} \ln (G D P_{i t} * G D P_{j t}) + β_{2} \ln D i s \tan c e_{i j} + ε_{i j t}

(2)

In our case, the country i is India and country j is a conglomerate of more than 170 countries. We have only one country i in this model, and hence, GDP_i does not vary across different panels at each time point. In this situation, one possible approach to estimate the gravity model is to use the product of the GDPs as a dependent variable. The main hypothesis of this model is that β₁ would be positive and the coefficient of distance, β₂ would be negative. Another approach is to drop India’s GDP and use only the GDP of the partner country in the gravity model. While we have chosen to use the former approach, we demonstrate that both approaches have similar outcomes in our choice of sample.

The statistics on bilateral trade have been taken from the IMF’s Direction of Trade data. The variable ‘trade’ encompasses the annual total of the value of India’s bilateral exports and imports with each of the trading partners.

GDP data have been sourced from the World Bank’s database (World Bank, 2017). Note that bilateral trade data are nominal, and hence it is appropriate to use nominal GDP as the explanatory variable. As elaborated in Baldwin and Taglioni (2006), any inappropriate deflating of the GDP variable may lead to biased estimates.⁸ Further, the use of time dummies or panel-fixed effects would be necessary to factor in the effect of price changes. Hence, in the gravity model, we use GDP in current US dollars.

Distance and other bilateral control variables are from the CEPII’s gravity dataset.⁹ The distance variable is the distance between two countries in kilometres, weighted by the geographic distribution of populations in the two countries.

The model presented in Equation (2) has been augmented to incorporate some standard gravity model controls, such as common official language, common colonial history, and contiguity, in order to test for robustness. Common official language enables smoother communication regarding the terms and conditions of a trade deal and the execution of transactions. Common colonial history is a proxy indicator of path dependence based on pre-existing colonial ties. Contiguous countries are more likely to trade with each other, as they share a border and it is easier to transport commodities, thus reducing transaction costs. Thus, we hypothesise that a common language, common colonial history and contiguity have a positive and significant effect on bilateral trade.

We undertake estimation of the gravity model and the extended gravity model using data on India’s trade with over 170 partner countries using different estimation strategies: the pooled OLS (POLS), fixed effects (FE) and random effects (RE) models. We also discuss the merits and demerits of each approach and identify the appropriate model for the subsequent sections of this article. The results are presented in Tables 8 and 9.

Table 8

OLS Regression

Dependent Var: Log of Total Bilateral Trade	(1)	(2)	(3)	(4)	(5)	(6)
Dependent Var: Log of Total Bilateral Trade	POLS	POLS	POLS (Includes Time Dummies)	POLS (Includes Time Dummies)	POLS (Includes Time Dummies)	POLS (Includes Time Dummies)
Log GDP Product	0.957	0.959	0.951	0.953
Log GDP Product	(0.009)***	(0.009)***	(0.010)***	(0.010)***
Log Distance	–1.145	–1.120	–1.147	–1.121	–1.147	–1.121
Log Distance	(0.040)***	(0.044)***	(0.040)***	(0.044)***	(0.040)***	(0.044)***
Common Official Language	0.587	0.591	0.581	0.585	0.581	0.585
Common Official Language	(0.053)***	(0.054)***	(0.054)***	(0.054)***	(0.054)***	(0.054)***
Contiguity		0.218		0.221		0.221
Contiguity		(0.150)		(0.150)		(0.150)
Common Colonial History		–0.379		–0.343		–0.343
Common Colonial History		(0.315)		(0.316)		(0.316)
Log GDP Partner Country					0.951	0.953
Log GDP Partner Country					(0.010)***	(0.010)***
Constant	–20.730	–21.037	–20.702	–21.019	4.704	4.433
	(0.615)***	(0.636)***	(0.653)***	(0.675)***	(0.453)***	(0.479)***
R ²	0.78	0.78	0.78	0.78	0.78	0.78
N	3,484	3,484	3,484	3,484	3,484	3,484

Source: Authors’ calculations.

Notes: (1) * p < 0.1; ** p < 0.05; *** p < 0.01.

(2) Standard errors in brackets.

(3) Time dummy coefficients are suppressed.

(4) All models satisfy F-test.

Table 8 presents some of regression results from the POLS model. As expected, the coefficient of the GDP product variable is positive and significant, while that of the distance variable is negative and significant. All the regressions include the variable common official language, which is positive and significant as expected. In column (2), we attempt to include contiguity and common colonial history, which are standard gravity control variables. However, we find that the coefficients of these variables are not significant. This is not entirely unexpected. India is located in a subcontinent where Pakistan and Bangladesh share most of the common borders, and India’s trade linkages with Pakistan are curtailed due to political reasons. In the north-east, India shares a border with China, but the terrain does not facilitate the standard benefits of contiguity in the context of international trade. Hence, the coefficient of the variable for contiguity is not significant. The coefficient of the common colonial history variable is also not significant, indicating that India has moved away from its colonial trading partners over time.

Table 9

Panel Regression

Dependent Var: Log of Total Bilateral Trade	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Dependent Var: Log of Total Bilateral Trade	RE Model	FE Model	RE Model	FE Model	RE Model (with time dummies)	FE Model (with time dummies)	RE Model (with time dummies)	FE Model (with time dummies)
Log GDP Product	0.989	.991	0.990	.991	0.915	.838	0.915	.838
	(0.010)***	(0.010)***	(0.010)***	(0.010)***	(0.030)***	(0.048)***	(0.030)***	(.048)***
Log Distance	–1.187	Omitted	–1.173	Omitted	–1.214	Omitted	–1.195	Omitted
	(0.155)***		(0.170)***		(0.147)***		(0.161)***
Common Official Language	0.677	Omitted	0.685	Omitted	0.596	Omitted	0.595	Omitted
	(0.203)***		(0.206)***		(0.195)***		(0.198)***
Contiguity			0.125	Omitted			0.156	Omitted
			(0.559)				(0.529)
Common Colonial History			–0.562	Omitted			–0.138	Omitted
			(1.234)				(1.178)
Constant	–22.044	–32.369	–22.186	–32.369	–18.331	–25.013	–18.506	–25.013
	(1.467)***	(0.552)***	(1.599)***	(0.552)***	(2.031)***	(2.390)***	(2.125)***	(2.390)***
N	3,484	3,484	3,484	3,484	3,484	3,484	3,484	3,484
Hausman Test	RE Model Preferred		RE Model Preferred		RE Model Preferred		RE Model Preferred

Source: Authors’ calculations.

Notes: (1) * p < 0.1; ** p < 0.05; *** p < 0.01.

(2) Standard errors in brackets.

(3) Time dummy coefficients are suppressed.

(4) All models satisfy F-test.

As described earlier, the model includes the nominal variables of trade and GDP product, and hence we introduce time dummies from column (3) onwards. Note that the coefficients of our key gravity model variables continue to support the hypotheses of the gravity model. In columns (5) and (6), we replace the GDP product with the GDP of the partner country only. We find that the GDP of the partner country is positive and significant in these regressions. Also, note that the magnitude of the GDP_j coefficients in columns (5) and (6) is fairly similar to those of GDP product in columns (3) and (4). This indicates that the impact of GDP_i in this model is minimal, as expected, since there is only one home country in this dataset.¹⁰ We continue to use the GDP product variable to achieve a complete depiction of the gravity model.

In Table 9, we use panel data models to estimate the gravity equation. The random effects model is more suitable in situations where we need to estimate time invariant variables; in our case, distance and common official language are the main time invariant variables. However, on the other hand, the benefit of a fixed effects model is that it captures the effects of different trade policies (including tariffs, quotas, etc.) and other peculiarities of different trading partners. In standard gravity models, one could also explore the option of including interacted country and time-fixed effects (Baldwin & Taglioni, 2006). However, given that we have only one home country in this model and that we require the inclusion of the distance variable in order to assess the impact of alternate trade routes, we would prefer a random effects model. However, we estimate both the RE and FE models and compare them in the Table 7. Note that the coefficient of the GDP product is positive and significant in each case. The columns (2), (4), (6) and (8) are results of the fixed effects model, and hence time invariant variables like distance and other gravity controls are not estimated. In the rest of the columns, we find that the coefficients of the variable distance are negative and significant, and that of common official language are positive and significant. We also find that the other gravity controls—contiguity and common colonial history—continue to be insignificant in both the fixed effects and random effects models. In columns (5) and (7), we incorporate time-fixed effects in order to adjust for the nominal trade and GDP variables. Finally, we perform the Hausman test to check whether RE is the preferred model—and find that in every case it is the preferred model for our sample. The coefficients of these models continue to be significant even with the use of cluster-robust standard errors.

It is necessary for us to select the most appropriate form of the gravity model for the next steps of our study, that is, calculating the trade potential index and assessing alternate trade routes between India and Central Asia. We select the random effects model along with time-fixed effects. The gravity model is presented in Table 9; column (5) is thus used as our benchmark to estimate the predicted trade or potential trade against which we compare the actual bilateral trade between India and Central Asian countries.

3.3. Trade Potential Index

The potential trade between India and any other trade partner in a particular year can be estimated using this model, utilising information about the GDP of India and the partner country in that year and the distance between the two countries. The estimated potential trade between countries can be used to create a trade potential index, which is the ratio between the predicted trade and the actual trade at time t.

(Trade Potential Index)_t = (Predicted total trade from the model)_t/ (Actual bilateral trade)_t

Using the above procedure, we first estimate the potential trade between the five Central Asian nations and India from 1996 to 2015. Then (for any given year) by dividing the estimated potential trade between any two countries by the actual trade between them for the same year, we can calculate the trade potential index between them for each year.

If the trade potential index is 1, then the actual trade is exactly equal to the estimated predicted trade given the GDP of India and of the trading partner and the distance between them. If the trade potential index is less than 1, then the actual trade is more than the predicted trade from this model. This is particularly true for trade-friendly countries, such as the UAE and the USA, which are also hubs of international trade. Trade is easier with these countries, and other unobserved trade frictions are likely to be lower in these nations than most developing countries. Finally, a trade potential index of more than 1 implies that actual trade is lower than the predicted trade expected from the model. This also implies that under the current conditions—the size of the economies and the distance between them—there is a potential to increase bilateral trade. We expect this to be more likely with other developing countries.

It is essential to keep in mind certain shortcomings of the trade potential index. One is that trade in services, an important aspect of India’s exports, is not included in the data. Therefore, this model suggests that as the GDP of India increases, India’s bilateral trade with any other country in the world should increase. However, with an increase in GDP, there might be an increase in services trade, which is unfortunately not captured at a bilateral level.

It is also instructive to focus on the direction of change of the index. If the trade potential index is larger than one and declines towards one, this indicates that actual trade is converging towards the predicted trade and that the trade potential is being tapped. On the other hand, an increase in the trade potential index implies that actual trade is falling short of the trade predicted by the model, and that the untapped potential for trade is increasing. The possible implication of this scenario is that there are obstacles to trade and one could benefit from exploring the benefits from trade. This has policy implications for the governments of both countries in question and signifies a need to identify and resolve problems affecting trade between them.

3.4. Results and Implications

Using the procedure explained above, we have calculated the trade potential index between India and the five Central Asian countries for the period 1996–2012; The results are plotted in Figure 5. For most parts, the trade potential index is far greater than one, indicating that actual trade between India and the Central Asian countries is much lower than the expected potential trade between them given the circumstances of distance and GDPs. Broadly speaking, the trade potential index is U-shaped over the time-frame of two decades (1996–2015).

Kazakhstan has a relatively lower trade potential index than the other countries, to some extent because it already has substantial bilateral trade with India. The trade potential index shows an upward trend from 2000 onwards, which implies that there is an increasing gap between the actual and prospective trade. Between 2011 and 2014, there seems to have been some ‘tapping’ of the trade potential, nevertheless, there continues to be further scope to improve bilateral trade. Kyrgyzstan has the lowest trade potential index among the Central Asian nations for the most part of these two decades, but the untapped trade potential between India and Kyrgyzstan has been increasing since 2004. The trade potential index was rather high for Tajikistan, Turkmenistan and Uzbekistan in 1996, but actual and potential trade appear to tend towards convergence until about 2002–2003, after which there has been a divergence. Actual trade has increased in this time frame, but is much lower than the potential predicted by this model.

Figure 5

Source: Author’s calculations.

The main implication of this is that the GDPs of India and the Central Asian countries increased in this period (as evidenced by the growth rate), but bilateral trade has not kept up. There has also been an increase in trade frictions—the political problems in Afghanistan started around this time, and the war has caused instability in the region. India’s relationship with Pakistan continues to be tense, causing hurdles in finding the most efficient trade route between India and the Central Asian region.

However, after 2010, there has been a decline or stagnancy in the trade potential index, possibly indicating a reorganisation of trading partners. As developed countries face a continuing recession, there may be some shift in trade in favour of the Central Asian countries. Finally, a tendency towards an increase in the index has been observed for 2015. The only exception is Turkmenistan, where actual trade has appeared to have caught up with predicted trade. The volatile nature of the index over the last 5 years warrants a cautious interpretation and prediction on the basis of these trends. In spite of this decline, the index lies between 5 and 13 for these countries in 2013 and between 3 and 9 in 2014. This denotes that trade in 2013 was 5 to 13 times lower than what it could have been, and in 2014, it was 3 to 9 times lower than its potential. Though the rise in the trade potential index has been arrested, there is still huge potential to improve bilateral trade between India and this region by more than five times, given global economic conditions in 2015.

4. Trade Potential Under Political Constraints

The distance between trade partners is the cause of some of these trade frictions. There may be several more frictions, notably tariff and non-tariff barriers to trade. Further frictions to trade may be caused by national or international policies that vary across geography and time; Bergstrand and Egger (2011) call these ‘unnatural’ or ‘man-made’ costs of trade. One possible cost to trade is political turmoil in the region, which has not been explored much in the literature.

In general, Central Asian countries are landlocked and have a challenging topography, which hinder trade in this region (Asian Development Bank [ADB], 2006). However, the problem of trade frictions between India and Central Asia is peculiar. The distance variable tacitly implies that trade can easily take place through the shortest trade route through Afghanistan and Pakistan. But this is not the ground reality for political reasons. While political disturbances within a trading partner can be captured in the form of a decline in GDP, disturbances in other countries that lie along the trade route are not directly captured in the model. Some of the frictions to trade between India and Central Asia are the political turmoil and security problems in Afghanistan and the adverse political relationship between India and Pakistan. These do not enable trade via the shortest routes through these countries. This leads us to the question: What happens to the trade potential if the shortest trade route is infeasible?

Given this situation, we consider two optional routes for trade between India and Central Asia—via Iran or China. In fact, trade via Iran has been under discussion for some years now. Balooch (2009) explores the possibility of increased trade after the ‘North–South corridor’ connecting India and Central Asia is created. Very recently, in May 2016, India signed an agreement with Iran towards investment in the Chabahar Port and for setting up a railway line for trade access to Central Asia (Roy, 2016). In contrast, trade via China is a more distant reality, given the political situation. While trade via Iran or China has its own international relations problems for India and Central Asia, it would be instructive to understand the implications of these options from an economic standpoint.

This situation highlights one drawback of the empirical methodology of gravity models, where the geographical distance between two countries is incorporated into the model, as opposed to the actual cost of transportation of the traded goods. There is no enough data on the actual distance of the trade route for different commodities. This may vary depending on the mode of transportation used (air, sea, roads, railways or a combination of these). Even for a particular mode of transport, commodities may have different features that may cause transportation costs to vary for goods with identical origin and destination locations. Given the lack of information, distance is used as a simple proxy for transportation costs. We further assume that trade via alternate routes is an aggregation of the distance between the bilateral country pair in terms of each route.

Literature that analyses the role of distance in international trade is rather sparse. Marimoutou, Peguin and Peguin-Feissolle (2009) demonstrate that while distance has a negative effect on trade, the magnitude of the effect of distance decreases as the partner’s GDP increases. This idea supports the findings of Fratianni and Kang (2006), which show that distance elasticity is much lower when the trading partners are Organisation for Economic Co-operation and Development (OECD) members. Neither India nor the Central Asian countries are very rich economies, and hence distances are likely to matter considerably. There is insufficient research on the role of distance in specific cases, such as ours. We attempt to understand the role of distance, which varies according to the alternate trading route, in the case of bilateral trade between India and Central Asian countries.

We create a trade potential index for three cases: the actual distance (the shortest route through Afghanistan and Pakistan); the distance via Iran and the distance via China. The actual distance is the distance between each of the Central Asian nations and India as defined earlier in Section 3.2. The distance via Iran refers to the sum of distances from each Central Asian country to Iran and from Iran to India. Similarly, the distance via China refers to the sum of distance from each Central Asian country to China and from China to India. Then, a trade potential index can be calculated for each of the three alternative trade routes.

The results are presented in Figure 6. As expected, the trade potential indices for all three routes show a similar pattern in each bilateral country pair. It is also expected that the index would fall as the distance increases. However, the question is whether there the magnitude of decline in the index is small enough for the trade potential to remain positive. In other words, does the trade potential index remain greater than one in these alternate routes?

The increase in the index after about 2000 shows that trade has not kept pace with the increasing GDP of the countries. We find that the potential trade index is very high when trade takes place via the shortest route (and potential trade is estimated using the shortest distance), often exceeding 10 at different time points. This suggests that there is potential for about a tenfold increase in trade between India and Central Asia if political and security problems could be sorted out, and if trade could be conducted through the shortest route via Afghanistan and Pakistan.

The index drops dramatically when trade takes place through the alternate routes, and the potential trade is estimated using the distances involved in these routes. In the case of Kazakhstan, the index remains greater than one in the recent past in the case of trade via Iran; however, trade via China has not seen a trade potential index of less than one given the situation in 1997–2000 and in 2014. Uzbekistan is the only nation where the trade potential index remains higher than one with few exceptions, even when the trade route is via China. But, China does not appear to be a feasible stopover in the case of trade between India and Tajikistan, Turkmenistan and Kyrgyzstan. For the latter two nations, trade via China became slightly more feasible since 2008. The trade potential index via Iran is generally greater than one, albeit by narrow margins. The one exception is Kyrgyzstan, in which case trade via Iran remained infeasible until 2006.

The average trade potential index over the five countries and two decades was 4.7 in the case of actual distances. This fell to 2.3 and 1.05 in the case of trade via Iran and China, respectively. The implication is that if trade via the direct route were possible, then bilateral trade could have potentially increased by 4.7 times on average, but this potential is reduced to 2.3 times or by less than half in the case of trade via Iran. Further, there is not much scope for improving trade potential for trade via China. Although this analysis is limited by the lack of data on routes and modes of transportation for specific commodities, the above model throws some light on the extent to which the trade potential between India and Central Asia is untapped. It also suggests that as the trade route through Iran opens, trade with Central Asia is likely to more than double.

Figure 6

Source: Author’s calculations.

5. Policy Implications and Conclusion

The recent thrust by India to encourage trade with Central Asia—a resource-rich and strategically located region—is a welcome move. There has been significant untapped trade potential over 1996–2015, which is a signal that there is likely to be positive potential for trade in the near future as well.

The current volume of trade between India and Central Asia has been very low. Further, the decline in trade in 2015–2016 is a cause of concern. We find, using the trade potential index that trade between India and Central Asia was three to nine times below the potential in 2014, indicating a huge scope for increasing bilateral trade between these countries in future.

One reason for the decline in trade in this region is the decline in commodity prices in international markets and the persistent unfavourable global economic climate. Another problem specific to this region is that, while India shares congenial political ties with Central Asia, political differences tend to occur among different countries that fall on the trade path between India and Central Asia. However, if the trade route via Iran becomes operational, trade with Central Asia can be expected to increase by about two to three times, as the distance would be significantly shortened. Hence, investment in this route would benefit the economies of Central Asian nations and India. Should security and political issues in and with Afghanistan and Pakistan be resolved over time, India’s trade with Central Asia could increase by about five to ten times the current levels, as that would dramatically reduce distances and the cost of transporting goods.

The primary policy suggestion that emerges from this study is that the Indian economy could gain from trade policies that encourage trade in this region. Bilateral or regional trade agreements could help increase trade up till its full potential. There would also be some merit in pursuing a policy of encouraging the North–South trade corridor between Central Asia and Iran.

Given the low probability of resolving security issues in Afghanistan and political problems with Pakistan soon, we propose three alternate suggestions to enhance trade and economic relations with Central Asia. The solutions are based on strategies that focus on alternate modes of trade to overcome the geopolitical barriers to trade in this region. One solution could be enhanced trading in high-value, low-volume commodities via air cargo. India could export coffee, tea, spices, frozen meat and pharmaceutical products. Importing from Central Asia is more challenging, since the most highly traded commodities are metals, minerals and oils. Nevertheless, India could import merchandise such as gold, silver and gemstones complementary to the growth of the jewellery industry in India. Another solution could be to expand FDI ties. Indian firms could set up industries in Central Asian countries that produce commodities for the local market there and vice versa. This idea is particularly feasible in the case of manufactured products, where new factories can be set up in the partner country. The goods would then be sold in the local market directly, bypassing the need for transportation through unreliable routes. India can also participate in oil, gas and minerals exploration in Central Asia. India’s share of the find could be sold to China and European countries, and the proceeds could be used to buy goods that are located closer by. This would solve the problem of transporting India’s share of the output of these items to India.

Third, we should consider enhancing trade in services, which largely eliminates the need for physical transportation. A strong, secure Internet connection would facilitate BPO services exports. India is progressing as a country that specialises in trade in services and, as a newly industrialising region, Central Asia needs technology services. Services exports from India, especially of information technology, have increased steadily since 2000. Also, Kazakhstan’s imports of services have been increasing during this period, and services trade in Central Asia will likely increase in the years to come.

Footnotes

Appendix

HS Code	Detailed Description Of HS Code
2	Meat and edible meat offal
7	Edible vegetables and certain roots and tubers
9	Coffee, tea, mate, and spices
13	Lac; gums, resins and other vegetable saps and extracts
20	Preparations of vegetables, fruit, nuts, or other parts of plants
21	Miscellaneous edible preparations
25	Salt; sulphur; earths and stone; plastering materials, lime and cement
26	Ores, slag, and ash
27	Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes
28	Inorganic chemicals; organic or inorganic compounds of precious metals, of rare-earth metals, or radi. elem. or of isotopes
29	Organic chemicals
30	Pharmaceutical products
31	Fertilizers
32	Tanning or dyeing extracts; tannins and their deri. dyes, pigments and other colouring matter; paints and ver; putty and other mastics; inks
38	Miscellaneous chemical products
39	Plastic and articles thereof
40	Rubber and articles thereof
42	Articles of leather, saddlery and harness; travel goods, handbags and similar cont. articles of animal gut (othr thn silk-wrm)gut
52	Cotton
61	Articles of apparel and clothing accessories, knitted or corcheted
62	Articles of apparel and clothing accessories, not knitted or crocheted
68	Articles of stone, plaster, cement, asbestos, mica or similar materials
69	Ceramic products
72	Iron and steel
79	Zinc and articles thereof
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof
85	Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts
89	Ships, boats and floating structures
90	Optical, photographic cinematographic measuring, checking precision, medical or surgical inst. and apparatus parts and accessories thereof;
96	Miscellaneous manufactured articles

Acknowledgements

We thank the anonymous referees, participants of the seminar held at the Institute of Economic Growth for helpful feedback, Piyush Mahajan for research assistance and the IDRC-TTI for supporting this study through the Think Tank Initiative Fellowship.

Notes

References

Agrawal

(2008). Economic growth and poverty: Evidence from Kazakhstan. Asian Development Review, 24(2), 90–115.

Agrawal

& Sangita

(2014). India and Central Asia: Trade prospects. In Bhatia

R. K.

Abusseitova

Yerekesheva

& Zafar

(Eds), India and Central Asia: Exploring new horizons for cooperation (pp. 106–128). New Delhi: Shipra Publications.

Anderson

J. E.

& Van Wincoop

(2003). Gravity with gravitas: A solution to the border puzzle. American Economic Review, 93(1), 170–92.

Asian Development Bank (ADB). (2006). Central Asia: Increasing gains from trade through regional cooperation in trade policy, transport, and customs transit. Philippines: Asian Development Bank.

Baldwin

& Taglioni

(2006). Gravity for dummies and dummies for gravity equations (NBER Working Paper No. 12516). Cambridge, US: NBER.

Balooch

(2009). Iran and India’s cooperation in Central Asia. China and Eurasia Forum Quarterly, 7(3), 25–29.

Batra

(2006). India’s global trade potential: The gravity model approach. Global Economic Review, 35(3), 327–61.

Bergstrand

J. H.

& Egger

(2011). Gravity equations and economic frictions in the world economy. In Bernhofen

Daniel

Falvey

Rod

Greenaway

David

& Kreickemeier

Udo

(Eds), Palgrave handbook of international trade (pp. 532–570). Hampshire, UK: Palgrave Macmillan.

Chaudhury

D. R.

(2015). PM Narendra Modi’s Central Asia tour: India set to push trade with resource-rich Kazakhstan. The Economic Times. Retrieved from http://economictimes.indiatimes.com/articleshow/47966320.cms?utm_source=contentofinterest&utm_medium=text&utm_campaign=cppst

10.

DGCIS-Kolkata. (2017). Export import data bank version 7.1—TRADESTAT. Ministry of Commerce, Government of India. Retrieved from http://commerce.nic.in/eidb/Default.asp

11.

Dowling

& Wignaraja

(2006a). Central Asia: Mapping future prospects (Singapore Management University Economics and Statistics Working Paper Series Paper No. 05-2006). Singapore: Singapore Management University. Retrieved from http://ink.library.smu.edu.sg/soe_research/869

12.

Dowling

& Wignaraja

(2006b). Central Asia after fifteen years of transition: Growth, regional cooperation, and policy, choices (ADB Working Paper Series on Regional Economic Integration No. 3). Phillipines: ADB. Retrieved from https://www.adb.org/sites/default/files/publication/28460/wp03-cars.pdf

13.

Fratianni

& Kang

(2006). Heterogeneous distance–elasticities in trade gravity models. Economics Letters, 90(1), 68–71.

14.

Head

& Mayer

(2014). Gravity equations: Workhorse, toolkit, and cookbook. In Gopinath

Helpman

& Rogoff

(Eds), Handbook of international economics (Vol. 4, pp. 131–195). North Holland: Elsevier.

15.

Head

Mayer

& Ries

(2010). The erosion of colonial trade linkages after Independence. Journal of International Economics, 81(1), 1–14.

16.

Hiro

(2010). Inside Central Asia. Noida, India: Harper Collins.

17.

Jakab

Z. M.

Kovacs

M. A.

& Oszlay

(2001). How far has trade integration advanced? An analysis of the actual and potential trade of three Central and Eastern European countries. Journal of Comparative Economics, 29(2), 276–92.

18.

Kabir

& Salim

R. A.

(2011). Parallel integration and ASEAN-EU trade potential: An empirical analysis. Journal of Economic Integration, 26(4), 601–23.

19.

Lee

(2016). India, Kazakhstan nurture millenia—Long trade relationship. The Astana Times. Retrieved from http://astanatimes.com/2016/03/india-kazakhstan-nurture-millennia-long-trade-relationship/

20.

Marimoutou

Peguin

& Peguin-Feissolle

(2009). The “distance-varying” gravity model in international economics: Is distance an obstacle to trade? Economics Bulletin, 29(2), 1157–73.

21.

Pomfret

(2005). Trade policies in Central Asia after EU enlargement and before Russian WTO accession: Regionalism & integration into the world economy. Economic Systems, 29(1), 32–58.

22.

Rakhimov

(2010). Internal and external dynamics of regional cooperation in Central Asia. Journal of Eurasian Studies, 1(2), 95–101.

23.

Reyaz

(2016). Is India finally getting serious about its connect Central Asia plans? The Wire. Retrieved from https://thewire.in/88943/india-central-asia/

24.

Roy

(2016). India to sign agreement for Chabahar port with Iran. The Indian Express. Retrieved from http://indianexpress.com/article/india/india-news-india/india-to-sign-agreement-for-chabahar-port-with-iran/

25.

The World Bank. (2017). World development indicators 2017 . Washington, DC: The World Bank. Retrieved from http://data.worldbank.org/data-catalog/world-development-indicators