Audit of Trade Facilitation Measures for Enhancing Pakistan Light Engineering and Made up Articles Exports to ECO Countries
By
Dr. Adil Khan Miankhel
Draft Report
November 2011
The author is Director at Pakistan Institute of Trade and Development (PITAD), Ministry of Commerce, Government of Pakistan. Please send your comments at
Pakistan’s exports have traditionally been concentrated in a few export destinations like USA, UK, Germany, Hong Kong, UAE and Afghanistan. Similarly Pakistan’s imports are also concentrated in few markets and imports mainly originate from UAE, Saudi Arabia, Kuwait, Malaysia, Japan, Germany, USA and UK. The details about direction of exports and imports indicate that except for Afghanistan, no other member country of Economic Cooperation Organization (ECO) is an important and major trading destination from Pakistan’s perspective (Economic Survey 2010-11). ECO is a ten member states organization comprising Afghanistan, Azerbaijan, Iran, Kazakhstan, Kyrgyz Republic, Pakistan, Tajikistan, Turkey, Turkmenistan and Uzbekistan. ECO exports during 2008 stood at US$276.0 billion while Imports amounted to US$ 306.8 billion. The intra industry trade in the region valued US$ 41.42 billion in 2008 (ECO Annual Report 2008). Pakistan being member of ECO needs to avail this opportunity and realize its full trade potential with other ECO member states.
Pakistan’s exports registered a growth rate of 29 percent during June-July 2010-11 compared to the same period of last year and as a result Pakistan’s exports crossed the US$ 25 billion mark for the first time (TDAP 2011). The major contributor to the export growth during the first ten months of current fiscal year compared to same period of last year was textile sector, constituting 55.3 percent of the total exports. Textile manufactures increased its exports by approximately US$2.7 billion. Similarly made up articles during the same period constituted 2.8 percent of total exports and registered an increase of approximately 30 percent amounting to US$ 132.8 million. In addition, engineering goods also increased by US$ 51.5 million during the same period. The engineering goods exports constituted 1.2 percent of the total exports June-July 2010-11 (Economic Survey 2010-11).
Economic Structure and Trade Links of ECO Countries
Table 1 Exports Structure US $ MillionProduct / July-April / Percentage (%) change / Absolute
Change / Percentage
Share of Total Exports
2010-11 / 2009-10
Textile Manufactures / 11,148.6 / 8,442.7 / 32.1 / 2,705.9 / 55.3
Made up Articles / 570.1 / 437.3 / 30.4 / 132.8 / 2.8
Engineering Goods / 239.9 / 188.4 / 27.3 / 51.5 / 1.2
Total / 20,154.2 / 15,773.2 / 27.8 / 4381.1 / 100
Source: Pakistan Economic survey 2010-11
ECO countries are relatively open economies especially the Central Asian Republics (CARS). Except for Turkey, Iran and Afghanistan with which Pakistan has established trade routes and trade facilitation infrastructure, Pakistan’s trade intensity is less than 1 for the rest of ECO countries (Table 2). This indicates that trading relations are not intense with Kazakhstan, Kyrgyz Republic, Tajikistan, Turkmenistan and Uzbekistan despite these being relatively open economies than Turkey, Iran and Afghanistan. However, this does not mean that CARs are not trading substantially with other countries or having low domestic demand for the imported products. The number of tariff lines at a 3 digit SITC level where import value for each tariff is higher than 100.000 dollars or represents more than 0.3 percent of total national imports in CARs is comparable with number of tariff lines in Afghanistan, Turkey and Iran. In addition, the concentration index also indicates that demand for imported products in CARs is fairly spread across imported products. However, the structure of imports by products in ECO countries differs considerably from average import structure of products in the world and this absolute deviation of the country share from the world structure for different products is more pronounced in CARs (Table 2).
Table 2 Economic Structure of ECO Countries 2010Indicator / Afghanistan / Azerbaijan / Iran / Kazakhstan / Kyrgyzstan / Tajikistan / Turkey / Turkmenistan / Uzbekistan
GDP
($ bn) / NA / 51.09 / NA / 142.9 / 46.16 / 56.40 / 735 / 21.07 / 38.98
Trade Openness (Avg2007-2009) / NA / 88.1 / 57.9 / 87.1 / 140 / 112 / 49.9 / NA / 70.6
Trade Intensity with Pakistan / 0.43 / 161 / 2.55 / 0.149 / 0.26 / 0.175 / 2.35 / 0.24 / 0.26
Absolute Value Imports (Tariff Lines at SITC 3 digit Level / 256 / 253 / 252 / 256 / 249 / 213 / 257 / 253 / 249
[1]Concentration Index Imports / 0.13 / 0.07 / 0.06 / 0.06 / 0.19 / 0.14 / 0.08 / 0.07 / 0.07
[2]Diversification index Imports / 0.51 / 0.47 / 0.4 / 0.36 / 0.48 / 0.55 / 0.29 / 0.45 / 0.43
Source: World Bank, UNCTAD, ITC Trade Map, WTO
The complementarity index[3] that measures the overlap and the degree to which the export pattern of a country (Pakistan) matches the import pattern of a country also indicates that this index despite being low with CARs at the start is rising over time demonstrating favourable prospects for trade integration (Figure 1). However, the performance of this index is different in Turkey, Iran and Afghanistan. In Turkey and Iran, it is fairly stagnant over time while with Afghanistan; it is rising sharply during the past three years highlighting prospects for further trade integration. ECO countries, in general, provide good opportunities for Pakistan export growth as these are open economies with the demand for imported products are fairly spread.
Light Engineering Products Exports to ECO Countries
Light Engineering as defined by the Pakistan Engineering Development Board includes 44 codes (products) at the HS 6 digit level. The detail of these codes (products) along with Pakistan’s total exports of these items to ECO countries and the World is given in Table A1 (Appendix). Pakistan’s total exports to world in 2010 in this category amounted to US$ 94.9 million that constituted only 0.44 percent of Pakistan total exports. Pakistan’s total exports of light engineering products to ECO countries stood at US$ 49.81 million that account for 52 percent of Pakistan’s exports in this category. In addition, Pakistan exports to ECO countries in only 18 codes (products) out of 44 codes.
Table 3 Pakistan Light Engineering Exports (US $ Mn)2008 / 2009 / 2010
Afghanistan / 22.764 / 32.988 / 49.55
Turkey / 0.096 / 0.191 / 0.215
Iran / 0.639 / 0.014 / 0.02
Turkmenistan / 0 / 0 / 0.013
Azerbaijan / 0.002 / 0.001 / 0.001
Uzbekistan / 0.003 / 0.013 / 0
Kazakhstan / 0 / 0 / 0
Tajikistan / 0 / 0 / 0
Kyrgyzstan / 0 / 0 / 0
ECO / 23.504 / 33.207 / 49.81
Source: Trade Map
As far as direction of exports of light engineering products in the ECO region is concerned, Afghanistan accounted for 99.3 percent of Pakistan’s total exports to ECO region (Table 3). Afghanistan was also the largest market for Pakistani light engineering products in 2010 accounting for 52 percent of Pakistan’s world exports (Figure 3).The large quantum of exports to Afghanistan can be explained in terms of proximity between the two countries and established trade routes and trade facilitation infrastructure with Afghanistan. Afghanistan is a landlocked country and most of the trade for Afghanistan is conducted through Pakistan via transit trade. In addition, most of supplies for the Allied forces in Afghanistan are also routed through Pakistan. Pakistan government has also allowed trade transactions with Afghanistan in local currency (Ministry of Commerce 2009). Given the flexibility of trade in local currency and demand effect due to non-existent or infant industry and reconstruction activities being undertaken by international developmental agencies, Afghanistan has become an important export destination for Pakistan. This can also be observed in the export of light engineering products.
Pakistan export of light engineering products to other ECO countries is insignificant. Even with the countries with which Pakistan has established trade routes and trade facilitation infrastructuresuch as Turkey and Iran, Pakistan export of light engineering products is negligible. In Iran and Turkey, China and Germany are major suppliers of this group. In Afghanistan, Russia and the US are bigger suppliers of this group than Pakistan while in CARs, Russia and China are the major suppliers of this group. CARs imported light engineering items to the tune of US$1.56 billion in 2010; however, their imports from Pakistan were only US$13,000. CARs are importing pumps from China, Germany and Russia. Fans from the Italy, the US and Russia and motors from Russia and China.
Different reasons can be attributed for low Pakistan exports of light engineering products to ECO countries such as ‘behind and beyond the border factors’ and supply side constraints. It is hypothesized that beyond the border factors are less likely to affect export of such items as Pakistan is already exporting 52 percent of its exports to Afghanistan only. So it could be ‘behind the border factors’ that may be hindering the export of such products to other ECO countries. However, if it is ‘ beyond the border constraints’ hampering Pakistan exports, any improvement in it is likely to improve the export performance of all the players in the export markets. So Pakistan needs to focus on ‘behind the border constraints’ in first place as initiating reforms will be easy. In Turkey and Afghanistan, tariff structure on light engineering products provides level playing field to Pakistan and its competitors in this group. In CARs, Pakistan and China receive no preferential tariffs in this group but Russia has duty free access. With Iran, Pakistan shares a common border and receives preferential access under Pak Iran Preferential Trading Arrangement (PTA) and despite this, its exports are negligible pointing to either supply side constraints or ‘behind the border constraints’ and whereby our production is either consumed locally or could not compete with China (Tables A2-A5 Appendix).
As far as competitiveness of Pakistan exports of light engineering products group is concerned, it is weak as this group Revealed Comparative Advantage (RCA) is less than 1. However, the competitiveness of this group is improving over time (Figure 5). Therefore, Pakistan needs to focus on improving ‘behind the border’ factors and increase supply base to capture bigger share of light engineering product market in the ECO region. By catering to bigger market, Pakistani exporters can reap the benefits of increasing returns to scale by focussing on economies of scale. This would help the exporters to not only specialise in certain varieties of this product group but also would stimulate investment for developing new products. Specialisation would lead to intra-industry trade and would integrate the region through trade links.
Made up Articles Exports to ECO Countries
Table 4 Pakistan Exports of Made up Articles US$Mn2008 / 2009 / 2010
Afghanistan / 1.947 / 4.878 / 2.974
Turkey / 6.069 / 3.925 / 4.143
Iran / 0.725 / 1.003 / 1.825
Turkmenistan / 0 / 0 / 0
Azerbaijan / 1.007 / 0.684 / 0.204
Uzbekistan / 0.117 / 0.052 / 0.39
Kazakhstan / 0.005 / 0.073 / 0.055
Tajikistan / 0.017 / 0.189 / 0.076
Kyrgyzstan / 0.023 / 0.036 / 0.492
ECO / 9.91 / 10.84 / 10.159
Source: Trade Map
Pakistan mainly exports made up products to the US and EU. Pakistan exported in 2010 made up articles to the world amounting to approximately US$ 3.28 billion. Exports to ECO countries are, however, negligible that constitutes 0.3 percent of Pakistan global export of made up articles. Among ECO countries, Turkey is a traditional export market for made up articles while Afghanistan and Iran are emerging markets. Exports of made up articles to CARs is low (Table 4). A disaggregated analysis for the four markets of CARS, Iran, Turkey and Afghanistan is also presented in the Appendix (Table B1-B5 Appendix). In the CARs, for example, Pakistan is exporting bed sheets and dish cloth. CARs are importing most of the made up products from China and Turkey. Towels which are important items in the Pakistan export basket were exported in small quantities in 2010 to CARS. It is important to mention that both Turkey and China have not been given any special tariff preferences in CARS markets. This raises the question that even though Pakistan is competitive in the made up articles and both Turkey and China does not have any preferential advantage or any geographical advantage, then why our exports for made up items to CARs are negligible (Figure 6).The competitiveness of Pakistan made up articles is very high and it demonstrates the strength of these products in the export basket of Pakistan. Despite this, the share of made up articles to ECO countries is very low. One of the reasons could be that there exist ‘behind the border’ obstacles that are hindering the export growth to these countries. The analysis for Iran also point towards the existence of ‘behind the border’ factors as the major reason for low exports of made up articles in Iranian market. However, as far as Afghanistan is concerned, Pakistan is a major supplier of these items due to close proximity with Afghanistan and established land routes for trade and we are also a major supplier in Turkish market. As far tariffs on made up articles in the ECO countries are concerned, our products are not placed at a disadvantageous position with respect to other exporting countries.
Pakistan Export Potential of Light Engineering and Made up Articles in the ECO Region
Export potential is defined as the level of exports that could be achieved with open and frictionless trade possible between countries. It is also called the maximum level of exports for a given level of determinants of exports. Frictionless trade is possible between countries when there are no institutional and infrastructural rigidities in home and trading partner countries. The rigidities that exist within home country are named as ’behind the border’ factors, while the rigidities that exist in trading partner countries are named as ’beyond the border’ factors.
Earlier studies have noted the difference between observed exports and the predicted value as potential exports. Generally, studies have used the OLS estimation of gravity equation to find trade potential (For example, Baldwin 1994; Nilsson 2000) between a pair of countries. The OLS estimation procedure produces estimates that represent the centered values of the data set. But, potential trade refers to free trade with no restrictions to trade. Thus, for policy purposes, it is rational to define potential trade as the maximum possible trade that can occur between any two countries, which have liberalized trade restrictions the most, given the determinants of trade. This means that the estimation of potential trade requires a procedure that represents the upper limits of the data and not the centered values of the data set (Kalirajan 2007).
The realized actual trade is measured by the current level of determinants of trade with existing level of restriction and institutions. Thus, there may exist a gap between potential and actual trade, which is directly related to various socio-political and institutional factors that are hindering the actual trade to grow to its potential level. By knowing the trade potential, countries could engage in bilateral and multilateral processes to make efforts to minimize or at least mitigate the effect of existing restrictive measures to trade growth. Therefore, the objective of any country is to achieve its full trade potential through reducing or eliminating institutional and infrastructure rigidities through trade policy reforms and alsoby taking part in bilateral and multilateraltrade negotiations.
Therefore, it is of significant importance that each country may know its full trade potential with other countries or other regions in order to get the reforms and negotiations process started. However, countries are diverse not only in their level of development but also in terms of endowment, openness, strategies, level and pattern of engagement with the rest of the world. Therefore, anycountry’s inability to export to its potential is a reflection firstof ‘behind the border’ factors that exist within the exporting country, and secondly of ’beyond the border’ factors that exist within the partner countries.
In this study, Stochastic Frontier Gravity model will be applied to determine Pakistan’s trade potential in textile made ups and light engineering products with ECO countries. This will enable us to examine how effective are the ‘behind the border measures’ as major trade constraints. The suggested approach provides potential trade estimates that are closer to frictionless trade estimates (Drysdale, et al 1997 and Armstrong, et al 2008). Because, this approach represents the upper limits of data, which come from those economies that have liberalized their trade restriction the most. Finally, the suggested method bears strong theoretical and trade policy implications towards finding ways of improving the performance of the ‘behind the border measures’. The gravity equation for exports can be estimated as for example,
,
where the term represents the actual exports from country i to country j. The term is a function of the determinants of potential bilateral trade (), which include distance, GDP and population to represent supply and demand conditions, and is a vector of unknown parameters. The single sided error term, is the economic distance bias referred to by Anderson (1979), which is due to the influence of the ‘behind the border measures’. takes a values between 0 and 1 and it is usually assumed to follow a truncated (at 0) normal distribution, . The double-sided error term is usually assumed to be. It captures the influence of omitted variables on trade flows including ‘beyond the border’ factors that are present in the importing countries in addition to measurement errors that are randomly distributed across observations in the sample.
The stochastic frontier gravity model is estimated at HS 2 digit level. The sample consisted of ECO countries for the period 2008-2010. The Gross Domestic Product (GDP), bilateral export data, population data, tariffs as well as GDP deflators’ for analysis were obtained using World Integrated Trade Solutions (WITS) from the UNCOMTRADE database and TRAINS. The bilateral trade weighted distances have been downloaded from ( The model estimation for light engineering products and made up articles are presented in Table 5 that indicate that sigma-squared (), which is a measure of total variation in the model is significant in all estimations.This variation in potential exports can be due to just random factors () or to country specific socio-political and institutional factors () between Pakistan and its ECO trading partners. The nature of variation in the model can be obtained by examining the gamma coefficient. We estimate gamma ‘ ’, which is . The value of close to 0 reflects that the bias or country-specific ‘behind the border measures’ are not important. If takes a value other than 0, it implies that country specific ‘behind the border measures’ are important. Behind the border obstacles incur additional transaction costs on the smooth flow of goods such as institutional costs due to attitude of institutions, regulatory and legislative costs, equipment and training costs and political costs due to inability of the governments to take trade facilitation measures due to geo-strategic reasons. Behind the border measures could range from product standards and conformity assessment measures, business facilitations, trade finance to hard (physical) and soft (regulatory) infrastructure such as efficient transport links and networks, logistics in form of efficient freight forwarders, distributors and efficiency of telecommunication system.