1.1. Theoretical Framework
The general idea of free trade agreement of growth was developed in advocacy of free trade based on neoclassical trade theory (Solow, 1956) and from recent endogenous growth theory (Romer P. , 1990). The support for free trade is drawn from Ricardian principles of comparative advantage (Viner, 1937). Similar idea is drawn from the notion of perfect competition and the believe of neoclassical economists who argues on the importance of efficient capital allocation due to free trade (Krugman, 1986; Corden W. , 1974).
The phenomenon of free trade came under severe scrutiny in the face of Great Depression. Hence, theoretical foundations of ‘optimum tariff' were developed in support of protection (Johnson, 1950; Kaldor, 1940). Johnson (1958, 1971) advocated trade protection in three groups in his classical exposition. They are the economic arguments, non-economic arguments and non-arguments. Economic arguments raise infant industry argument, optimal tariff argument and correction of domestic market distortions, while non-economic arguments emphasize on self-sufficiency for domestic economy. Non-arguments attempt to resolve balance of payment distortions through trade protection. Johnson concluded that ‘optimal tariff' protection is the only valid argument, while in other cases such arguments will only inflict distortions.
The neoclassical economists refute the notion of protection as an alternative, as this would result in intra-industry effects. The increased barrier to entry would make domestic traders to engage in monopolistic competition, while small enterprises will be left inefficient. Intra-industry effects are the source to welfare loss (Tybout J. D., 1991). In addition, Bhagwati(1988) and Kruger(1974), raises the theory of directly unproductive and profit (DUP) seeking activities, which will cause waste to national resources. Additionally, the Solow-growth model embodies technology as an endogenous factor (Agion, 1992; Romer P. , 1989), which argue that international trade ensures faster diffusion of technology, that is embodied into the better intermediate goods which results in higher productivity and growth for domestic economy (Grossman, 1991). This will result in learning by doing effect and technological know-how is surpassed. In addition, management is more efficient and all will combine in high growth (Krugman, 1987; Young, 1991; Lucas, 1988) .
1.2. Objective of the study
A high number and standard of studies have been conducted on Export-led growth, trade openness, “manufacturing exports as a new engine of growth”, specifically in the last decade, on different economies, ranging from developed to poor countries, drawing interesting conclusions. The present study seeks to investigate the effects of export, openness on growth in the context of Bangladesh.
Firstly, the study will seek for stable effects of policy shifts and implementation, in Bangladesh, which will be determined by stable changes in the determinants. Stationarity conditions, if satisfied, will ensure the stability of economy and productivity, towards a particular goal.
Secondly, the study will investigate the current association between growth and trade openness. While it is desired, that the adopted outward-looking trade policies of Bangladesh to result in positive association of productivity to liberalization, trade openness might be effected by other variables and may render different conclusions.
Thirdly, the study will examine, if the export led growth hypothesis is still applicable to Bangladesh, as before, while many countries, such as Sri-Lanka, Philippines, Nigeria have seen opposite relationships. Additionally, Hossain and Karunaratna (2004) have argued that “manufacturing exports have become new engine of growth” which is a disciple of the de novo hypothesis. In contrast, Adelman (1984) suggests that, agricultural exports should have dominant effect for a pro-agricultural society as Bangladesh. It is important to see if, ‘manufacturing exports is an engine of growth', or still other factors are dominant as before.
Investment is an endogenous factor that should imply the increased import of intermediate goods, as a result of increased export, and more openness, would consequently; render higher productivity (Krugman, 1987; Lucas, 1988; Young, 1991).
1.3. Relevance and limitations of the study
Relevant studies have been conducted in the context of Bangladesh, in last decade and have drawn interesting remarks. However, the major drawback is the timeframe of earlier studies, which did not cover analyses from the last ten years. In the last ten years, econometric methods have changed and improved rigorously. Hence, many studies have been rendered invalid due to absence of proper methodology. The world economy has seen dramatic events in politics, international trade and global economy. The trends in global economy, which were much more rigorous, in the last ten years, have affected Bangladesh magnificently, as Bangladesh emerges as a high power economy in Asia, and have interested researchers, due to high deviations and high rises to productivity. It is necessary to embody recent econometric techniques of Johansen's maximum likelihood cointegration analysis and vector error correction methodology, which will inform on recent associations, among the interested indicators. Hence the state-of-art econometric techniques will provide reliable results that would help the policy makers to observe the relationships and bring sufficient changes, in trade policy to render profit.
Among the few limitations of the study was the absence of first hand secondary sources. Most data sources for Bangladesh are not available online and are preserved in paper based format. The lack of proper technology and internet, withdraw the authorities of the country, to provide data directly. Therefore, data are collected from World Bank sources, which may not correct for errors, and sometimes fail to provide detailed data series as an intermediary.
1.4. Structure of the dissertation
The second section will contain a brief country profile and approaches to liberalization. The third phase will contain literature review that will discuss literatures in support to export-led growth hypothesis and trade liberalization. This section will bring forth studies that contrast and significance of the study. The fourth and fifth section will contain methodology to estimation and results of analysis. The final section will give conclusion and remarks to the dissertation.
2. Process of trade openness and Export-growth in Bangladesh
After independence in 1971, Bangladesh has gone through three phases of policy changes, towards deregulation and openness to trade. The first phase was marked by severe control on exports and imports. The policy implemented in 1972 to 1975, put the country in a socialist framework, with a fixed exchange rate system. Industrial enterprises, banking and trade infrastructure was massively nationalized as an inward-looking, import substitution approach was adopted. Agricultural inputs and outputs were controlled. Empirical literature suggests that this was a good decision for the researched timeframe (Ahmed N. , 2000).
The second phase of policy shifting began in 1976 and continued up to 1990. This phase of denationalization, deregulation and trade liberalization lacked a good direction to work out the process. Nationalized trade barriers were reduced, and a ‘free trade' approach was undertaken. Privatization of industries and banking sector was allowed and price controls over nationalized firms were lifted. Abolishment of state trading was initiated.
The third phase of policy shift, were introduced in the beginning of 1991 and continued up to 2002 with significant remarks. In the recently developed policies toward export-promotion and trade openness, ‘trade barriers' have been removed as a flexible exchange rate regime is adopted. To encourage further reduction of anti-export bias, export processing zones have been established, to co-operate manufacturing exports. This recent policy shift has moved towards complete privatization of banks, infrastructure and agricultural sector. During this phase, Bangladesh continued to experience rapid liberalization.
In order to further opening up the boundaries, Bangladesh has entered into a ‘free trade agreement' among countries in the region. As discussed, to encourage EP trade policies, and free trade, Bangladesh entered into bi-lateral agreements with India, Pakistan, Sri-Lanka, and continues to trade and investment framework agreement with United States. Bangladesh is an active member of SAARC, Developing-8, and Bangkok Agreement. The country is rigorously seeking to import in developing of a regional co-operation among Bangladesh, Bhutan and seven northeastern states of India and Nepal.
From the graph presented above, it can be seen that, starting form 1971, imports were increasing at a low rate, with a fall in 1976-77, but moved up after that. The import line took a peak after the third policy implementation in 1991, and continued to rise at a high speed. The export trend was almost at a constant level, until the 1991 policy implementation. Afterwards, the trend peaked, gained a momentum after 2000. However, the export line still falls below the import trend.
As reported by the export promotion bureau, in 2005-2006 FY, export earnings have grown by 21.3%, which is due to high demand of, chiefly manufactured goods, led by garments industry and knitwear. During this period, import payments have grown by 9.5%.
In contrast to real export, manufacturing exports exhibit of higher exports than imports which are in percentage of merchandise exports. This is indicative of higher productivity.
When the country adopts rigorous trade policy towards openness and promoting exports, large scale capital owners and labor unions are in oppose to liberalization. The cause of opposition may lie in the fact that, increased competition, effect the workers, in protected public and private sector, who fail to adjust to the exposed economy, with increased productivity. The politicians tend to protect large scale owners, who severely discourage competition and create bias towards exports (Sattar, 2004).
3. Literature Review
3.1. Concepts and Empirics on Export-Led Growth Phenomenon
Export led growth phenomenon has been central to the trade and development literature for many years. The bulk of researches that has taken place on this issue are therefore, not small in number and range. The focuses of these studies were bi-directional. Some of the studies attempted to find whether expanding export would improve the growth performance. Others tried to find the paths through which the expansion of export will affect the growth performance. Economic theory confirms that export expansion leads to increase the growth performance, as that efficiently allocates productive resources and also with the high volume of productive resources accumulated as a result of higher capital earned through export growth, (Bardhan, 1970; Cheneray, 1966; Basu, 1991; Romer P. , 1989 ; McKinnon, 1964; Grossman, 1991).
Export expansion makes the home country to concentrate on comparative advantages and to earn economies of scale. The home country continues to invest on its economies of scale and achieve better efficiency. The increased efficiency creates external competitive pressure on the home country, along with improved internal competition. In the face of more competition, the monopolistic and oligopolostic behavior of the market is eliminated. Efficiency is also created as learning by doing. Knowledge is transferred to other sectors and growth is enhanced. The external competition, aids the small economy to realise and emphasize on removing limitations, through economies of scale, and by reaping the true advantage of globalisation, which is acquired by increasing export.
The theoitical literature also argues that, export expansion increases the investment and capital accumulation in a country. The two-gap model explains , that growing export reduces the constraints that prevail in foreign exchange. Such reduction of constraints lead to better accumulation of productive resources, capital goods and intermediate goods (McKinnon, 1964; Bacha, 1984; Cheneray, 1966). Export expansion also increases investement opportunity of a country. Modern economists suggest that, savings by domestic and banking system, government savings and foreign exchange savings cannot only induce investment. Investment opportunities determine investment rather than savings. The growth of export will provide investment opportunities to home country, (Sandrum, 1994).
Theories suggest that, the relationship of export and economic growth is bi-directional. Economic growth may also increase export for a country. The effect of better learning and technological development, give rise to output. The growth of output ensures that domestic demand is met and export will expand. However, this technological process development or learning process development is not directly related to the export promoting policies (Jung, 1985). As the home country realizes economies of scale, expansion of export takes place. Investemnt (Grossman, 1991). Therefore, the bi-directional theory suggests that, GDP is a function of investment, that enhances export performance, and export increases investment opportunity, that is directed to GDP growth.
In the designing process of development economics, entailed was dominant export passimistic theories, rather than export promoting views. After the end of world war II, import substitution strategies were to be followed by many countries. However, the initial phase of implementing import substitution strategies seemed ideal during that period, but the results of taking attempts to implement the import substitution strategies were not favorable for all economies. Economists found that, that export passimistic views were not justified for many economies that have reached a certain level of development, and industrialization. On the other hand, the import substitution trade policies laid undesirable effcts on balance of payment. As a result, the growing economies did not accept import substitution strategies, rather accept export promoting views (Adelman, 1984).
The growth led export is also suggested by theories. According to Bhagwati (1988), growth led export hypothesis is dominant when supply and demand is induced by growth. In such cases, anti-trade bias is turned down.The possibility of bi-directional causality prevails in many major theoritical literture (Grossman, 1991; Bhagwati J. , 1988).
On the other hand, Irma Adelman (1984) argues that, export-led growth is not the only open development strategy for a least developed country. The open development strategy that ensures the allocation of agriculturally driven resources may prove superior than the strategy for allocation of capital for investment resources. An alternative to the import substituion strategy, for a closed development approach, maybe an ADLI ( agricultural-demand-led-industrialization) strategy. It is important to identify the phase of closed development ideology, which is dominant after the era of Pro-agricultural strategy is over. Another theoritical literature similarly argued that, ADLI as a balanced-gowth-approach, can only be a mean, to attain the goal for developing countries need for higher growth, and the need for growing industrialization by expanding the demand for domestically produecd consumer goods and intermediate goods. The linkage effect to agriculture with industrialization is also examined (Singer, 1979). Singer (1979) defines this target as Lime target and ADLI is the solution. However, Irma Adelman(1984), also puts out the constraints to ADLI strategy as it requires the improvement of infrastructure in Agriculture which is difficult to attain in South Asian Sub-continent. The physical capital needs to be perfectly infrustructred for ADLI yield expected results. Therefore, ADLI strategy is a solution to allow time to the developing countries to bring abouth changes structurally, in 1980s to 1990s. This strategy cannot wire out the importance of export-promoting strategies as an alternative for import substituion strategies at all (Adelman, 1984).
Alongside the theoritical literatures on export-led growth hypothesis, the number empirical literature from the first of its kind by Maizels(1963) is many. Jung and Marsall(1985) scanned 11 empirical studies that were carried out from 1967 to 1982 timeframe, and all of the studies found supportive relationships to export and growth. Greenaway and Sapsford(1994), reviewed 14 empirical studies that were conducted on the export-led growth hypothesis, and 12 empirical studies vividly shown relationships between export and growth. One of a major study conducted by Giles and Williams (2000), which conducted 150 cross country analysis from 1963 to 1999. Out of the 57 countries that were analysed , only 4 countries failed to show significant relationships between export and growth, and only 10 out of the 102 time series analysis didn't show significant relationships between export and growth.
It has interested the empirical literature to examine export-led growth hypothesis prior and after the oil shock in 1973-74 timeframe. Among the studies,that took place prior to the oil shock, Michalopoulos and Jay (1973) conducted a study in a 1960-73 timeframe, by estimating export and gowth into a poduction function, signifcant relationships were found. Tyler (1981) conducted study on a group of middle income countries, putting export and growth into similar production function framework, and found similar relationships.Therefore, export orientation into the framework is supposed to effect growth therough economies of scale, allcation of productive resources and utilization of capital, optimally. The study of Feder (1983) found similar results. Balassa (1983) were dubious about the relationship afther oil shock, as in the face of economic recession 1974-75 that took place after the quadrupling of oil price in 1973-74, may have effected the relationship because of the orientation of external shock in the production function. The study he conducted had taken th period of 1973-79, after the shcok, on 43 developing countries, who were directly affected by subsequent recession. The result did show export affecting growth positively and the numerical magnitude of the effect did grow compared to early results. The changes in intercountry growth rate before and after the oil shock is rather a result of different trade policies introduced.
An important theoritical implication is increasing export also paves the way for imported capital goods to be entered into the country (Islam M. , 1998). As productivity is increased, investment along with profit grows and the economy enjoys higher growth (Edward, 1993; Levine, 1992). In last two decades, exports of newly industrialized countries grew by 20%. Manufacturing exports entailed 70% of total exports. As a third factor, import of manufactured and productive capital goods increased. The demand for these capital goods indicates the increasing rate of growth. Therefore, the plethora of studies on export and growth make this issue important enough to review.
3.2. Cross-country empirical analyses
In light of previous section, many empirical literature also focused on perticular countries or a category of countries to examine the export-led growth hypothesis. In theoritical literature, it is defined that, if export growth coefficients and and output growth coefficients are significantly positive in regression, the country follows export promoting strategies. If output growth causes export growth in regression than the country is labelled IGE or internally generated exports. On the other hand if a export growth coeiicient and output growth coefficient is negetively correlated in the regression for growth, the country follows ERG or export reducing growth strategy (Jung, 1985). Such countries are following inward-oriented strategies rather than outward-oriented policies. Inward-oriented countries may also follow IS (import substitution) trade policy.
Many empirical studies on cross-country did confirm the existence of export-led growth for different countries, and in some countries results otherwise is found. Hatemi J. and Irandsout (2000) continued analysis on Ireland, Portugal and Mexico, and significant relationship was confirmed. In the same study, they failed to confirm causal relationship for Greece and Turkey (Hatemi-J.A., 2000). The study by Ghirmay et al. (2001) did find positive relationship of export and growth for a number of developing countries. Just after one year, another study by Greenaway et al. (2002) conducted analysis on a number of selected developing countries and found that the growth rate for these countries dropped immediately after trade reform, for a constant rate of export, but gained momentum following a J-curve response after the affect of trade refor wires out.
M. Michaely (1976) analysed 41 developing countries for a significantly large period. The resulting conclusion implicated that while Greece, Taiwan, Portugal, Spain, Israel, Yoguslavia and Koria had rapid growth with increasing export, Portugal did not show significant export growth, while GDP was growing in same pace. On the other hand, when Ethiopia incresed its export performance considerably, but failed to increase its growth to the pace with other countries. On the basis of rank correlations, M. Michaely (1976) concluded that export performance will positively effect growth of a country, only when a country achieves development of a perticular level. Countries below this level will fail to exhibit good export-growth relationships. Bela Balassa(1977) followed similar study conducted by M. Michaely (1976), running rank correlations on a sample of countries that established industrial base for a timeframe of 1960-73. Among these countries, Korea, Singapore and Taiwan adopted EP (export promoting) strategies at a very early stage and provided incentive to the exporting sector by subsidizing the sector in many ways. On the contrary, though, Israel and Yoguslavia promoted export during the same period, but their efforts seemed to dim in the later periods. On the other extreme, Argentina, Brazil, Columbia and Mexico, continued the existing trade policy, supporting import substitution (IS). During this period, Chile and India continued their inward-oriented policies and was in the phase of weakly introducing export promotion policies. The resulting conclusion estimated that, while Korea and Taiwan would have less growth with more export, Chile, India, Mexico,Brazil and almost all other countries would have better levels of growth with higher levels of export. The countries that moved to opposite direction is due to the unfavorable internal conditions and policy constraints, the countries have. Similarly, for Phillipines and Srilanka, opposite direction of relationship is found (Islam M. , 1998).
3.3. Controversial Theories and Evidence
Due to the debt crisis and continued recession that prevailed during 1980s, after many countries adopted export promoting strategies, theorists and economicsts were dubious about export-led growth hypothesis. It became a necessity to re-examine the export promoting strategies (Bhagwati J. , 1988).
The revived passimistic school of thought was dominated by old and new school of thoughts. The most influencial school of thoughts were suggested by two great contemporay development economists, that were Raul Prebisch (Prebisch, 1952) and Ragner Nurkse (Nurkse, 1953). Prebisch (1952) recommended that, chief exports for newly industrialized countries will decline following a natural cycle, regardless of the trade policy implemented by the home country. Producers of home economy will respond by rapid industrialization and the economy will respond by employing more protection and higher level of restrictions. All these attempts will make export promotion unjustified. The other dominant export passimism was realized by Nurkse(1953), who stressed more on “balanced growth”. The “balanced growth” theory suggests that the accelerated pace of growth and exports of developing countries, make foreign markets unable to accommodate imports on sufficient level. Developing economies shift from raw materials to synthetic materials as inputs, and damage exports for developing countries longer run. Riedel (1984) suggests that, demand dominates export performance. It is a biased view towards export growth relationship if that is explained by export performance of selected countries. Export performance maybe more reliant on domestic incentives of a country, rather than the external conditions (Riedel, 1984).
The new literatures focus more on successful Asian exporters, which maybe a mistake to be implemented as a general strategy for all economies. The sources of worry advocates that, markets are shifted to export promotion, markets will fail to absorb all exports. The earlier wave of export passimism was afflicted by this idea. Economists have always supported the idea of intra-industry specialization which leads to adoption of inward-oriented policies, and terms an economy as closed economy.The countries that adopt outward-oriented policies to a greater extent, or publicly promotes export promoting strategies, associates some level of government intervention. The government intervention makes sure that exports are promoted, subsidized and invested into. The countries that are empirically supportive of export-led growth hypothesis, mostly follow government intervention. This practice is generalized except countries like Hong-Kong. However, these are exceptional cases and should not be generalized (Bhagwati J. , 1988).
Economists also view that, export promoting trade policies, make the domestic market less sheltered and susceptible to world economic condition, outside pressure, world competition as well as innovation. This view has also been critisized and the opposite direction is supported by Schumpeterian arguments (Bhagwati J. , 1984). The theory of market imperfections addressed by Fields(1984) suggests that, in presence of excessively high wages, countries may do poorly. An example of this theory is set to Jamaica. Another interesting theory is the satisfaction theory of import substitution suggests that, the export promotion strategies are not suitable for many newly industrialzed countries due to their lack of flexibility for movemet of capital resources. Countries also lack the political capabilities to implement this flexibility (Ruggie,1983). Similar argument is also suggested by Adelman (1984) who argues that, agricultural-demand-led-industrialization strategy should be applied to allow a country enough time, for it to develop a structural base, before the country can successfully implement export promoting strategies. Export-led growth will follow if the country can achieve a minimum level of development as suggested before. While in the face of rapid growth, many countries are doing well with export promoting trade policies, countries like Taiwan, Sri-Lanka, Phillipines, Jamaica, Brazil, Korea. are examples of countries, that was not in the position to implement export-promoting trade strategies,and the desired export-led growth was not achieved.
In some theories, learning by doing or intra-industrial knowledge transfer was an important factor for adopting export promotion strategies. It is believed to be a mean of acquiring economies of scale for industries. The know-how process is a major motivating factor for economies, to support export promotion strategies. However, even, learning by doing effect dims down and may stop completely in absence of newly developed technology (Young, 1991). This makes outward-orientation unjustified and export-led growth, a failure. Import substitution and export promotion strategies do best when they are complementary (Grabowski,1994; Hamilton and Thompson, 1994).
One important alternative suggested by Adelman (1984), is the ADLI (agricultural-demand-led-industrialization) strategy. The ADLI argument can be set into the similar footsteps of IGE (internally gorwn exports) of Jung (1985). Adelman (1984) argues that, when the countries became more reliant on industrial export-led growth, the controversies arise as many of least developed countries experienced lower employment, deteriorated income distribution, high level of food imports as domestic demand is not met, and lower level of growth. Therefore, countries required to implement more basic need oriented strategies. The rising foreign exchange constraints, and the serious liquidity problems, least developed countries faced, as they moved towards more export promoting strategies, following the export-led growth hypothesis, newly renewed export passimism (Adelman, 1984).
As empirical evidence, causality tests between export and groth conducted by Jung and Marshall (1985) on 37 countries should be addressed. In this empirical study, countries as many as South Africa, Korea, Pakistan, Israel, Bolivia and Peru did not show significantly positive relationship between export and growth. Rather these countries showed export reducing growth, which is the other way around. If these countries implements export promoting trade strategy, the countries will experience crippled economy and lower growth. Countries such as Iran, Kenya and Thailand are in favor of internally generated growth, and in the process of successfully implement export promoting growth policy (Jung, 1985).
Looking at the above results, countries therefore, support import substitution as a pro-agricultural trade policy (Adelman, 1984). Countries are also adviced to move towards ADLI strategy in this stage. It may not be favorable for countries to immediately implement export promoting trade strategies, hoping to yield benefits of export-led growth hypthesis. In the primary stage, countries require import substitution policies, to develop intra-indistrial skills, economies of scale, and a trade base to a minimum level. The level in between is a complementary stage between import substitution and export promotion. Countries as India, Malaysia, Bangldesh in South east Asian region, followed import substitution for longer period until they believed to reach the minimum level, before these countries can move towards outward-oriented policies and introduced trade openness, and enjoyed the benefits of export-led-growth. For many countries mentioned above, steps taken in an earlier phase, have backfired. Therefore, the controversies to export led growth is as prevailent as the support toward the hypothesis.
3.4. Empirics on export led-growth and trade liberalization in the context of Bangladesh
As one of Asia's growing power house economy, The export-led growth hypothesis has been examined in the context of Bangladesh, in many empirical literaures. Among the newly conducted researches, conintegration analyses, vector error correction models, explained many important variables such as manufacturing exports, investment capital to the total exports and growth. This part will briefly review the studies conducted in the context of Bangladesh.
Since its independence, Bangladesh embarked in a import substitution trade policy; following the ideology that a pro-agricultural society should be motivated to develop intra-industry to achieve economies of scale (Adelman, 1984). The mounting foreign debt, instable political condition, low productivity and growth, lower national income, did not allow the country to achieve its economic objective. Therefore, the country had to convert its inward looking policies, towards more outward looking policy, and adopted export promoting trade policy in 1982. Many structural adjustments were adviced by world bank and international monetary fund. The country went under further reforms in fiscal, monetary and industrial sectors, in the periods of 1985-86 and 1991-92; that helped the country to reduce its bias against exports or convert to export promoting country from an import substituting country (Hossain M. a., 2002; Rahman, 1995). During this period the de novo hypothesis, that states, that “manufacturing exports becoming a new engine of growth” , replacing total exports (Athukorala, 1998; Helleiner, 1995) were examined against Bangladesh in some empirical literaures (Hossain M. a., 2002; Ahmed N. , 2000; Hossain M. a., 2002).
The study of Hossain M. A. and Karunaratna (2004) conducts analysis to examine whether manufacturing exports has replaced total exports as a new engine of growth for a period of 1974-1999, applying vector error correction and cointegration models. Significant relationships supporting export led growth is found, when the results weaken the de-novo claim and concludes that total exports is still a dominant factor to GDP growth. This is due to the fact that, agricultural exports may dominate the manufacturing exports in total exports, as a pro-agricultural society, that follows the ideology subscribed by Adelman (1984). In a study of cointegration on export and growth by Hossain M.A. and Karunaratna (2003) , empirical findings suggested that the policy reforms in Bangladesh have not been able bring about significant structural changes in the economy, instantly. However. The structural changes have brought about competition, that has fuelled the rigorous implementation of export promoting policies with significant government intervention, and export-led growth is therefore, supported (Hossain M. a., 2004).
The article prepared by Alauddin M. A. and Hossain (2005) examined the long term effects of trade liberalization to the structure of export, import, growth and other relevent variables for a period of 1974-1999. Trade openness was measured with trade openness bias index, restructuring of tariff rates, the degree of openness and positive effects were found. However, the structure of effects change its dominance along with the transitional phase, between trade liberalization, export and growth. The early periods were dominated by agricultural exports and, the newer periods are converting towards industrial outputs; while in Bangladesh, exports are dominated by a few commodities (Alauddin M. A., 2005). A similar study was conducted by Nasiruddin Ahmed (2000) with a aggregate export-supply function for a period of 1974-1995, for Bangladesh. The applied Cointegration and Vector error correction model examines varibles of real quantity of export, real exchange rate, relative exports, trade liberalization proxied by a dummy variable and capacity to prduction proxied by real GDP. The resulting output did find significant relationships between export, growth and trade liberalization. The recommendation for this study was, to better export promotion, trade liberalization should be associated with devaluation,that supports the theory of Dornbusch (1974). For higher export promotion, existing bottlenecks should be removed, and that will reduce anti-export bias (Ahmed N. , 2000). Among the Asia's high power economy countries, in the study of Muhammad N. Islam (1998), using cointegration methodology, evidence of cointegration between export and growth was found in Bangladesh, Sri Lanka, India, Nepal, Fiji; while other countries showed evidence for causality but no cointegration was found among the 15 countries surveyed.
One of the major studies on Bangladesh, was conducted by P.J.dawson(2005), running cointegration analysis on export, import and GDP with the help of cointegration analysis for a period of 1973-2003; where significantly positive reltionships between export and growth is found. However, constraints toward successful implementation of export promoting trade policy still existant, restrains to achieve true export promoting policy. The results therefore suggested, that export-led growth hypothesis is supportive to the context of Bangladesh and trade liberalization should be continue in current pace, due to the fact, that it contributes to economic growth, in contrast to the ‘populist view', that it hinders growth. The continued improvement of trade liberalization will reduce the anti-export bias in the long run (P.J.Dawson, 2005).
3.5. Significance of further study
The pleothra of studies conducted to export-led growth phenomenon, may have applied methodology and range of data, that have been modified in recent times. The studies by Michaely M. (1976), Balassa (1977) and Maizels (1963) applied rank coeffcients to identify relationships between export and growth hypothesis, and did not apply statistical causation. Other studies by Bela Balassa, (1985); Singh and Kohli (1989) applied simple linear regression model, and did not emphasize on the time series data being unit root or not.
The empirical literature have used a wide range of definitions of “economic growth” and “export”. As export being a component of growth, many of these empirical findings were spurious, and researches tried to circumvent these problems at a later stage. This has been defined as ‘accounting identity' problem. By taking export-gdp ratio (Michaely M. , 1976; Islam M. , 1998) or gdp net of exports (Heller P. a., 1978; Islam M. , 1998), the occurances of spurious regressions were attempted to be avoided.
While econometric methodology and trade policies in the world have changed, highly in the last few decades, it is indeed, important to study the affect of export to growth is required to be tested for recent times, applying state of art econometric techniques and taking in consideration the ‘accounting identity' problems for the many important literatures still being reviewd.
While many countries have converted their policies, as one of Asia's high power economies, Bangladesh have converted their policies from inward-looking towards more outward-looking policies. The post-agricultural Bangladesh had import substituting trade policies for many years, and than converted to export-promoting trade policies. The early empirical literatures found support for export-growth hypothesis for Bangladesh. Major empirical studies such as, the one conducted by Islam, M(1998), did apply econometric methodogy, taking in consideration, the ‘accounting identity' and unit root analysis, conducted study for the period till 1991. Among the latest, the study by Hossain and Karunaratna (2004) applied vector error correction and cointegration methods, but up until 1999. Similarly, the stuy of Muhammad A. Hossain and Alauddin (2005), find support for export-led growth, for the period until 1999, applying cointegration methods. One of the major study conducted by Nasiruddin Ahmed (2000), who established cointegration relationship between trade liberalization, growth and export, did estimate for upto 1995.
Among the old theoritical and empirical litratures, conducted by Balassa (1977), Michealy M.(1976), Irma Adelman (1984), Jung and Marshall (1985), Jagdish N. Bhagwati (1978; 1988) contributed highly in the empirical literature. However, in the phase of recent changes in the methods of estimation and global circumstances, may have rendered many of the important literatures insignificant, empirically. The major drawback for the studies carried out for Bangladesh, which have applied econometric methods to draw conclusion to the export-led growth phenomenon, is the timeframe of study. Many years have passed afterwads. In the last 10 years, the world economy has faced many obstacles, and Bangladeshi economy is affected by the ups and downs of world economy. In addition, the changes from inward-looking to outward-looking orientations, may not bring consistency in the pattern of growth. To adjust to immediate impediments, Bangladesh have converted and changed the trade policies and approaches. Aftter the incident of 9/11, the export sector of Bangladesh faced hindrance and exports reduced. The dramatic political changes in the recent years may have also affected the export-growth relationship. In such situation, it is important to analyse export-led growth hypothesis with significant emphasize on trade liberalization for recent years to see, whether the current indicators follow the same pattern as before, or their exists a necessity to bring about further changes to the adopted export promoting policies.
4. Methodology to Analysis
4.1. Empirical Framework
Empirical literature in past has conducted co-integration test, often in two-variable framework, neglecting the probability of misspecification. To avoid the occurrence of such errors, study will not exclude the variables that are assumed to have significant impact on growth. This study will therefore, include investment that is assumed to influence GDP. In the study conducted by Hossain M. a. and Karunaratna (2004), two different multivariate models were suggsted to analyse impact of manufacturing export and real export on GDP. In this study, real export and manufacturing export will be suggested in a single multivariate model, to analyse their impacts in a cointegration framework. The direction and the level of impact is necessary to observe. Other empirical investigations were carried out on trade openness and export promotion. Ahmed N. (2000) , Alauddin and Hossain M.A.(2005) have conducted important empirical studies on trade liberalization and growth. Therefore, in a single multivariate framework, trade openness is added to suggest its impact, in presence of other important variables. The presence of investment variable is important because, if export coeifficient is significant, that implies an improvement in efficiency. If increasing exports cause capital accumulation or capital formation to increase, this will have significant effect on investment (Ghirmay, 2001). As government increases spending and import of intermediary goods, and capital goods increases,that is proxied by capital formation, as a result of increased export and openness to trade, suggested by many literatures, is a good indicator of growth. The multivariate model, that combines the growth, export and trade variables, and as natural logarithms are denoted by lower case letters, and t denoting the time subscripts, we have
Model: U_t(〖lgdpnet〗_t,〖lOP〗_t ,〖lx〗_t, 〖lcapital〗_t, 〖lmanu〗_t)
Can be written as ,
〖lgdpnet〗_t = α_0+ α_1 〖lOP〗_t + α_2 〖lx〗_t + α_3 〖lcapital〗_t + α_(4 ) 〖lmanu〗_t + ε_t
〖lgdpnet〗_t = natural logarithm of real gdp net of exports
〖lOP〗_t = natural of trade openness proxied by total GDP
〖lx〗_t= natural logarithm of real exports
〖lcapital〗_t= natural logarithm of real investment proxied by capital formation
〖lmanu〗_t = natural logarithm of real manufcturing exports
If we denote 〖lgdpnet〗_t by y_t, following the estimation procedure of Engle and Granger (1987), the long run static equation is
y_t = α_0+ α_1 〖lOP〗_t + α_2 〖lx〗_t + α_3 〖lcapital〗_t + α_(4 ) 〖lmanu〗_t + ε_t …………………………….... (1)
Equation (1) is estimated using ordinary least square estimator and the short run dynamic equation is established as
y_t = αy_(t-1) + β_0 〖lOP〗_t +〖 β〗_1 〖lOP〗_(t-1)+ 〖 γ〗_0 〖lx〗_t + 〖 γ〗_0 〖lx〗_(t-1) + ϑ_0 〖lcapital〗_t + ϑ_1 〖lcapital〗_(t-1)+ θ_(0 ) 〖lmanu〗_t+θ_(1 ) 〖lmanu〗_(t-1)+ ε_t ……………..................................................................(2)
And this equation can be written as
y_t = δ_1 〖∆y〗_t + β〖lOP〗_t +〖 δ〗_2 〖∆lOP〗_t+ γ〖lx〗_t + 〖 δ〗_3 〖∆lx〗_t + ϑ〖lcapital〗_t + δ_4 〖∆lcapital〗_t+ θ〖lmanu〗_t+δ_5 〖 ∆lmanu〗_t+ u_t
Estimating this model with short run terms 〖∆y〗_t , 〖∆lOP〗_t ,〖∆lx〗_t,〖∆lcapital〗_t ,〖 ∆lmanu〗_t is just equivalent to estimating the long run parameters of β,γ,ϑ,θ.
Then the issue of a good power property arise. In an Engle-Granger framework, the null of no- cointegration is tested. We assume the variables are integrated in their first difference , having an order I(1) integration and there exists one cointegrating vector, the Vector Error Correction model is thereby,
〖∆y〗_t = δ_0+∑_(i=1)^n▒〖 δ〗_1i 〖∆lOP〗_(t-1)+ ∑_(i=1)^n▒〖 δ〗_2i 〖∆lx〗_(t-1) + ∑_(i=1)^n▒〖 δ〗_3i 〖∆lcapital〗_(t-1)+∑_(i=1)^n▒〖 δ〗_4i 〖 ∆lmanu〗_(t-1)+ δ_5 〖 ECM〗_(t-1) +ε_t ……………………………………………….......................................(3)
Here, ECM is the error correction term, or in other words, residuals from the first model.
The error correction term or equilibrium correction term is therefore,
〖 ECM〗_(t-1)= y_(t-1) - β〖lOP〗_(t-1)- γ〖lx〗_(t-1)-ϑ〖lcapital〗_(t-1)- θ〖lmanu〗_(t-1)
The long run error correction model for this analysis, can be written as,
y_t = δ_0+〖 〖 δ〗_1 lOP〗_t+ 〖〖 δ〗_2 lx〗_t + 〖 δ〗_3 〖lcapital〗_t+〖 δ〗_4 〖 lmanu〗_t+ δ_5 〖 ECM〗_(t-1) +ε_t ………………….(4)
The short run dynamic error corretion model , that dependes on the previous lags of the series and estimates the current value, accounting for only a few periods change, that may change its behavior in the long run, but worth looking at is
〖∆y〗_t = δ_0+〖 δ〗_1 〖∆lOP〗_(t-1)+ 〖 δ〗_2 〖∆lx〗_(t-1) + 〖 δ〗_3 〖∆lcapital〗_(t-1)+〖 δ〗_4 〖 ∆lmanu〗_(t-1)+ δ_5 〖 ECM〗_(t-1) +ε_t…(5)
4.2. The Data
The study has covered sample period from 1980 to 2007. The data are obtained mainly from ‘World development indicator, 2009'. The source files are located into Economics and Social data service (www.esds.ac.uk). The data has been collected for Bangladesh from world bank. Besides World Bank, data is collected from International Monetary Fund, Balance of Payments data. Some data has also been collected using Datastream.
The following section consists the definition of the variables that has been used in empirical analysis.
GDP net of exports is the indicator of growth variable. GDP net of exports is the dependent variable. The data for this variable is in constant local currency unit, for the period of 1980 to 2007. The net value added is used rather than the gross value added output, representing growth. Gross value added output is better than gross output, as it entails the inclusion of raw materials as well as it accounts for differences [Kalirajan and Salim, 1998]. Gdp net of exports is more appropriate than gross value added, as it provides superior strcture of data allowing for better time series analyses.
TGDP or total gross domestic product is a proxy variable for trade openness or trade liberalization. The reason for inclusion of tgdp as proxying for liberalization is the fact that, the growth of total gross domestic product indicates the reduction of anti-trade bias for a country. More liberalization of trade is parallal to the total growth of output, with higher levels of input into production, with increasing import of raw materials and intermediary capital goods. This has been confirmed by a plethora of empirical and theoritical literatures. The direction of movement of these variables therefore ensure that one can be a proxy of the other.
Gross Capital formation is the variable that is representing investement as the increasing rate of capital growth indicates growth of efficiency, growth of productivity, production of more raw material and intermediary goods. The aggregate book value of land, buildings, machinary, transport, offices etc., which is estimated as gross fixed assets is representative of Capital formation. As defined by Salim and Kalirajan(1999), in countries like Bangladesh, Capital is utilized at constant level of efficiency, beyond the accounting life defined by depreciation, until capital is completely discareded. In other empirical literatures, replacement cost of existing capital is also a good measure investment. In this analysis, data on fixed capital assets is used. The fixed assets are in constant local currency units.
Real Export variable is indicating the exports of goods and services. The exports are in constant local currency units. Real export is indicative of the growing exports of the country, due to policy changes in favor of liberalization, perticularly after 1984-85. Real export comprises agricultural output, industrial output as well as services.
Manufacturing Export is the variable that represents, the total export of manufactured output or capital exports only. This is an important variable as suggested by many empirical literature. While the de novo claim establishes “manufacturing exports as engine of growth”, some empirical findings also suggest that “manufacturing exports is the new engine of growth for Bangladesh” (Hossain M. a., 2004). On the other hand, in theoritical literature by Irma Adelman (1984), it is discussed that, as pro-agricultural society, agricultural exports may have better influence on growth than manufacturing exports, and otherwise there may exist negetive relationship, as observed in many developing countries (Adelman, 1984). This discussion magnifies the importance of the inclusion of real export and manufacturing export variable together, to observe the effect of manufacturing export on growth and real export on growth in a distinct manner. This will let us observe, whether the other factors of agricultural exports or services are still playing a dominant role, rather than the manufacturing exports , on growth, and the after affect of trade liberalization.
The annaul data is used for the range from 1980 to 2007. Though data is avalibale from 1967 to 2007, the data of 13 years is excluded for various reasons. The primary reason is , Bangladesh was in a war uptil 1972. The post-war era experienced highly instable political situations and conflicts of many level, that put the economic growth in a standstill. During this period, the infrustructure to export was being built as the country adopted an inward looking policy. However, in absence of significant infrustructure, there was no significant export data available, to be included in the observations. Furthermore, the inclusion of data from this period may cuse arbitrariness in the entire group of sample. The infrustructure in favor of exports and imports were in the process of development, but became active from 1980s, and hence the sample data is collected from this period.
It is important to mention that, in estimating financial time series, the natural logarithm is taken for the data avilable. It is important for financial time series, because it renders the non-linear data to linear data and thus enables the time series analyes to be applied on the series (Brooks, 2008).
4.3. Econometric Methodology
While in the past literatures, a major issue was the absence of testing for stationarity, that renders many old analyses invalid. The presence of non-stationay data may inflict spurious regression, the test statistics will be invalid and shocks on the data will persist forever. Therefore, checking for whether the error term is serially uncorrelated and homoscedastic is very important, before going into further analyses. If the error term is autocorrelated and heteroscadastic, the regressions will be spurious and the results will be unacceptable. To estimate the long run relationships among the variables, the following three step procedure is inherited.
In the first step, the staionarity of the variables is checked. As Brooks (2008) state it “ the stationarity or otherewise of a series can strongly influence its behaviour and properties”. To check for stationarity, the existence of unit root is traced in null and alternative hypothesis. In presence of a unit root, on the level data or its first difference, than it can be stated that the series is stationary in its first difference or second difference. In order to do that, tests for order of integration will be run on the level data, and in their 1st difference. The tests, to check for unit root will be the Augmented Dickey-Fuler test (ADF), which is highly used in recent empirical literatures. The other two tests will be Phillips-Perron tests (PP) and the suggested test by Kwaitkowski et al. (1992), known as KPSS test. The criticisms of these tests are is on their power properties, and thus we will consider all the tests.
In the second step, which is the major part of the analysis, is that we check for co-integration among the variables in the system. The objective of the analysis is to look for the long run dynamic relationship among the varibles, and the direction of movement, as well as significance of the cointegrating relationships are checked. There exists two important estimation criteria suggested by Engle, R.F. and Granger,C.(1987) and by Johansen, S. and Juselius,K.(1990). These two estimation criteria are known Engle-Granger two step procedure based on the estimation of residuals and the Johansen's full information maximum likelihood method. The problem with the former one is that, the Engle-Granger approach is based completely on the residuals of the data, and therefore, may mislead the result. In addition, the Engle-Granger method fails to provide the rank of cointegrating vectors, and the cointegrating vectors. The Johansen's approach overcomes all of these and thus is considered a better estimation method for the analysis. I, therefore, will base my results on Johansen's full information maximum likelihood estimation procedure. This will give us, the rank of cointegration as well as the adjustment coefficients.
Finally, It is not only desirable to look for the long run realtionships among the vrible series, but also desirable to look for shocks in the series and the length of time, the shocks will persist, in the relationship. If suppose, the series of variables are in an equilibrium steady state level, than the deviations should be estimated. It is desirable to look for, when the shocks would converge on the equilibrium condition once again. For this purpose, the Vector error correction model or Equilibrium correction model will be estiamted. The testing of significance of the error correction term is very important for acceptability of the error correction model. The error correction term will give the speed of adjustment coefficient, which will further strengthen the long run relationships of the equation, with considering any further deviations form the series.
5. Empirical results
5.1. Time Series Properties of the Data
The concept of stationarity is very important for time series analyses for various reasons. A stationary series have a constant mean, constant variance and a constant autocovariance for a given lag. The stionarity can influence the behavior of a series in many ways.
One important consideration is the affect of a ‘shock'. A ‘shock' is an unexpected or unpredictable change in a variable, which is expected to gradually die away, only if the series is a stationary one. The effect of a sock in time t will have smaller effect in time t+1, and a smaller effect in time t+2. However, for non-stationary series, the shocks will be infinite and in some instances will be increasing. Another consideration is maybe the existence of a ‘spurious regression'. If there are independent random series in the system, which may have no relationship, running a regression might show that, according to the t statistic and F statistic , the variables are insignificant. The goodness of fit may also give a low R^2. However, if the data are not stationary, the resgressions may look good , with high R^2 and significant outcome, but the standard assumptions of asymptotic analysis will be valueless, due to the fact that t-ratios and F-ratios will not follow t-districution and F-distribution. In addition, a non-stationary series will produce enormous large values if regression is run. Therefore, to conduct this study, stationary must exist in variables.
A series that is differenced d-times to be stationary is said to be integrated of order d and can be written as y_t~ I(d).
In an autoregressive process, the y_t depends on its previous value, that is y_(t-1) and the error term u_t. The equation can be expressed as
y_t= 〖 φy〗_(t-1)+u_t
And with repeated substitutions, can be written as,
y_t= φ^(T+1) y_(t-1) + 〖φu〗_(t-1) + φ^2 u_(t-2) + φ^3 u_(t-3)+……………………+〖 φ〗^T u_(t-T) + u_t
And with further substitutions,
y_t= φ^T y_0 + ∑_(j=0)^T▒φu_(t-j)
Now, from this equation, if φ<1, the shock in the system will gradually die away, and this is a stationary process. If φ=1, the shock in the system will persist and this is called a unit root process.
Now, to test for the existence of unit root, Augmented Dickey-Fuler test (ADF), Phillips-Perron tests (PP) and Kwaitkowski et al. (KPSS) test is run. The data are in level and test regression includes intercept.
For ADF to be valid, the error term must be a white noise. ADF includes p lags of dependent variables to make the distribution white noise porcess. In the following Augmented-Dickey Fuller equation, y_t will represent, 〖lgdpnet〗_t ,〖lOP〗_t , 〖lx〗_t , 〖lcapital〗_t ,〖lmanu〗_t .
〖∆y〗_t= ψy_(t-1) + ∑_(i=1)^p▒〖α_i 〖∆y〗_(t-i) 〗 + ε_t
Where the H_(0 )= y_t ~ I(1) , unit root
Against 〖 H〗_(1 )= y_t ~ I(0), stationary.
The first step is to check if the residuals are abiding by the condition of error term to be a white noise.
Table 1: ADF test on level data
Series Test statistics Critical Value *MacKinnon (1996) one-sided p-values
Prob.* H_(0 )= y_t ~ I(1) , unit root
H_(1 )= y_t ~ I(0), stationary.
LGDPNET 5.313523 1% level -3.699871
5% level -2.976263
10% level -2.627420 1.0000 Test stat> critical value Do not Reject H_(0 ).
LOP 0.427760 1% level -3.699871
5% level -2.976263
10% level -2.627420 0.9805 Test stat> critical value Do not Reject H_(0 ).
LX -1.572761 1% level -3.699871
5% level -2.976263
10% level -2.627420 0.4823 Test stat> critical value Do not Reject H_(0 ).
LCAPITAL 1.319921 1% level -3.711457
5% level -2.981038
10% level -2.629906 0.9980 Test stat> critical value Do not Reject H_(0 ).
LMANU 0.781027 1% level -3.699871
5% level -2.976263
10% level -2.627420 0.9917 Test stat> critical value Do not Reject H_(0 ).
From th ADF test tale, all the variables have been checked for stationarity at level. Unit root is found in Log of Gdp net of exports series at 1%, 5% and 10%, at level data, and it can be said that it is a I(1) process and the series is stationary at first difference, and with very good level of significance. However, the series is not I(2) and non-stationary at 2nd difference (appendix). The log of trade openness is stationary at first difference as unit root is found in level data as well, and the null of unit root is not rejected. Unit root is rejected convincingly in first difference and therefore, cannot be concluded that, the series is stationary at second difference ( appendix). Similarly for, log of real exports, log of capital formation and log of manufacturing exports are all stationary at first difference, as unit root in level is found. However, a distinctive feture is that, among all the series, only real exports series show lower power, as significance is lower. The rest of the series follows the trend of the former and are not stationary in their second difference (appendix).
In the following Phillips-Perron equation, y_t will represent, 〖lgdpnet〗_t ,〖lOP〗_t , 〖lx〗_t , 〖lcapital〗_t ,〖lmanu〗_t .
〖∆y〗_t= f(t)+γy_(t-1) + μ_t
Where the H_(0 ): γ_t =0 , (unit root)
Against 〖 H〗_(1 ): γ_t ≠0 (stationary)
f(t)= α_0+ 〖βD〗_t where α_0= constant and 〖βD〗_t = trend.
The Phillips-Perron test gives similar result to the ADF test. Unit root is found in level data, and can be concluded that the all the series are stationary at their first difference, but not at their second difference. The test results are included in the appendix.
The ADF and PP test has been criticized for having low power, where the root of stationary process is very close to the non-stationary boundary. KPSS test is run on the variables, with opposite null and alteative hypothesis, that checks for stationarity directly.
KPSS (1992) is testing for non-stationary component , if that is really zero. As y_t = α+ β_t + μ_t is trend-stationary process, The KPSS test equation can be expressed as
y_t= f(t)+x_t + ε_t where f(t)= α+ β_t and x_t= x_(t-1 )+ ϑ_t
Here, ε_t is IID and ϑ_t is stationary.
In KPSS test the
Where the H_(0 ): σ^2=0 , (stationary)
Against 〖 H〗_(1 ): σ^2 ≠0 (unit root)
Series LM Test statistics Critical Value H_(0 )= y_t ~ I(0) ,
H_(1 )= y_t ~ I(1)
LGDPNET 0.674246 1% level 0.739000
5% level 0.463000
10% level 0.347000 Test stat< critical value Reject H_(0 ).
LOP 0.592119 1% level 0.739000
5% level 0.463000
10% level 0.347000 Test stat<critical value Reject H_(0 ).
LX 0.530114 1% level 0.739000
5% level 0.463000
10% level 0.347000 Test stat<critical value Reject H_(0 ).
LCAPITAL 0.674674 1% level 0.739000
5% level 0.463000
10% level 0.347000 Test stat<critical value Reject H_(0 ).
LMANU 0.666900 1% level 0.739000
5% level 0.463000
10% level 0.347000 Test stat<critical value Reject H_(0 )
From the KPSS test, for all the variables, unit root at first differenceis found. The alternative hypothesis has been accepted for all the variables, and can be concluded that, all the series are stationary at first difference, as non-stationarity is rejected.
Finally, from all the three tests in looking for unit root, all the series of data are stationary, and the are fit to run the cointegration on the data, as the possibility of spurious regression has been removed and the series will provide desired and valued results.
5.2. Cointegration in Multivariate Sytstem – Johansen Approach
As the variables contain a unit root, than the components or the variables are integrated at order 1, expressed as I(1). A set of variables will be cointegrated only if the linear combination of these variables are stationary. If not stationary, the variables might still move together, due to some influence that is bound on the variables, in the long run. A cointegration is a long run phenomenon, since it is possible that, the variables in a cointegration may deviate from their mean in the short run, but will return into the association in the long run. This association is the desirable property that needs to be estimated. If the variables of interest are non-stationary and not-cointegrated, than there would be no long-run relationship that will bind the variables together. In absence of a constant mean, there will not be any mean reversion,and will not be any association among the variables (Brooks, 2008).
As mentioned before, due to some constraints and limitations in Engle-Granger 2 steop method, Johansen-Jesulius full rank cointegration method has been applied. The null hypothesis in this system implies of no-cointegration, where the alternative hypothesis implies the existence of cointegrating equation. The result is organizaed in the following table for cointegartion rank test based on λ_trace test. Here, r denotes rank (number) of cointegration.
Table 3 : Cointegartion rank test based on〖 λ〗_trace
H_0 hypothesis H_1 hypothesis λ_trace Test statistics 95% Critical Value Decision
r = 0 r = 1 81.61887 69.81889 Reject H_(0 ).
r≤ 1 r = 2 42.99245 47.85613 Do not Reject H_(0 ).
r≤ 2 r = 3 17.06495 29.79707 Do not Reject H_(0 ).
r≤ 3 r = 4 5.803932 15.49471 Do not Reject H_(0 ).
r≤ 4 r = 5 0.296372 3.841466 Do not Reject H_(0 ).
Table 4 : Cointegartion rank test based on〖 λ〗_max
H_0 hypothesis H_1 hypothesis λ_trace Test statistics 95% Critical Value Decision
r = 0 r = 1 38.62642 33.87687 Reject H_(0 ).
r≤ 1 r = 2 25.92750 27.58434 Do not Reject H_(0 ).
r≤ 2 r = 3 11.26101 21.13162 Do not Reject H_(0 ).
r≤ 3 r = 4 5.507559 14.26460 Do not Reject H_(0 ).
r≤ 4 r = 5 0.296372 3.841466 Do not Reject H_(0 ).
From both the trace test and maximum eigenvalue test, the null hypthesis of no cointergration is rejected once. Therefore, the rank of cointegration vector is 1 and there exists one cointegrating equation. The implication of this decision is that the model is integrated at order I(1). In addition, there exists significant long run association among the variables.
5.3. Vector Error correction procedure
“ One definition of the long run that is employed in econometrics implies that the variables have converged upon some long term values and are no longer changing,…..hence all the difference terms will be zero……the class of model that can overcome this problem is error correction models” (Brooks, 2008).
The acceptence of the Vector error correction model, depends on the significance of the error correction term and the equilibrium correction term with the correct sign. Binding β_1=1 restriction into the , rank 1 cointergrated system, the derived error correction term is statistically significant, have the correct sign, with the value of -0.353633, showing a high speed of convergence to the equilibrium. The long run error correction model (ECM) ,which fits the model best, experimenting the general form of equation (4), is therefore,
y_t = 0.48 -0.32〖lnlOP〗_t+ 0.53〖lnlx〗_t + 0.17〖lncapital〗_t+0.49〖 lnmanu〗_(t-1)-0.35 〖 ECM〗_(t-1) …(6)
Where y_t denotes log of gdp net of exports. The desired long run relationship, of the model states that, Bangladesh's real export, manufacturing export, investment is positively related to real gdp or growth, in the long run. From the model it is confirmed, that real export, real manufacturing export, real capital formation have significantly positive relationship with growth, while trade openness has negetive realtionship with growth. For 1% change in real export, real manufacturing export and real capital formation or investment, gdp will grow consecutively by 0.53%, 0.49% and 0.17%, which are all statistically significant. In contrast, for 1% change in trade openness, real gdp will fall by 0.32%, which is also significant. The speed of adjustment coefficient determines that, it is expected for the model to converge into equilibrium within 3 quarters.
The shor run error correction model, will look into the current situtation of the model, which is run on the differential log of the variables. Satisfying the previously applied conditions of unit root, one cointegrating vector, and similar restrictions applied, the error correction model examines the general form of equation (5),
〖∆y〗_t =-0.14 -1.23 〖∆lOP〗_(t-1)+0.82〖∆lx〗_(t-1)+0.62〖∆lcapital〗_(t-1)+1.62〖∆lmanu〗_(t-1)- 0.143 〖 ECM〗_(t-1) ….(7)
The short run error correction model, also have statistically significant variables.Here, it can be noticed that, for 1% change in manufacturing exports, gdp will increase by 1.62%, which is very high, compared to real gdp, which causes the gdp to grow by 0.82%. Manufacturing exports are therefore, showing a very good impac even in short run, as well as long run.On the contrary, 1% increase in trade openness, gdp will fall by 1.23%, which is also very significant. The error correction term has the correct sign, and a good speed of adjustment. In short run, the rate of convergance should be around 1 quareter.
6. Concluding remarks
The study seeks to investigate, if export-led growth hypothesis is applicable to Bangladesh, as well as if “manufacturing exports becoming a new engine of growth” (Hossain M. a., 2002). The study also seeks to investigate the relationship between growth and trade openness. The objective of the study was to seek for associations between indicators of growth, trade openness, export, and manufacturing exports. This would substantially suggest on the validity of the policy framework adopted and should be followed. In order to suggest on these, dynamic time-series econometric methods have been applied on the time frame from 1981-2007. Johansen's cointegration method has been applied and vector error correction model (VECM) was drawn. These models were conditional on the data sets being stationary. The statistical significance of both long run and short run model is checked and relationships drawn.
The empirical results imply that
The data for GDP net of exports, or growth indicator, trade openness indicator total gdp, real export data, manufacturing export data and real capital formation are all stationary in the long run. This would explain that, Bangladesh had a stable growth, with distortions that are not highly significant, and converges into equilibrium in the long run. Therefore, the data are valid to be estimated and have drawn reliable conclusions.
Growth has found to be positively associated with real export, that would imply, increasing exports will yield higher productivity for the country. This association has proved that Bangladesh continues to follow, Export-led growth hypothesis.
Growth is significantly associated with manufacturing exports, which carries a higher coefficient. While growth is reliant on exports, the sector of export is highly comprise of manufacturing exports, making growth to be affected highly by manufacturing exports only. This has verified that, manufacturing exports are becoming the new engine of growth.
Growth is negatively associated with trade openness indicator. This is justified by the fact that higher trade openness has caused Bangladesh to import more than export. As a result, the country is being reliant on imports. This association is significantly alarming, as if the process gains a momentum, Bangladesh might experience a negative growth in the long run. However, higher imports are bringing in more capital goods which is increasing productivity, ensured by the association capital allocation and growth variable.
The study has specified the duration to adjustment of distortions, which is very important prediction.
The study has found to be supportive of the study conducted by Ahmed (2002), where it was found, that Bangladesh follows export led growth significantly. As Asia's power house economy, among a few countries Bangladesh is in line to the theory (Ahmed N. , 2000). The study also advocates the previous study conducted by Hossain and Karunaratna (2002), explaining that manufacturing exports is becoming engine of growth following de novo hypotheses. It can be concluded that, de novo hypothesis is applicable in the case of Bangladesh. The analyses states that increased imports is affecting the balance of payment, and productivity negatively (Dornbusch, 1974).
The study is different to the studies conducted before, because the study primarily, uses the current level of data, which has been updated in recent dates. Secondarily, the method of econometrics applied is far superior, from methods applied to many studies, which has significantly derived conclusions that can be followed closely. The benefit of recent methods have given a contrasting fact , that implies, Bangladesh still needs to monitor its openness very closely, as the study finds a negative relationship, that is caused by increased imports; and if continued, may reduce productivity.
The study suggests, that the government should continue to encourage exports. Increased imports are bringing in intermediate goods, and capital goods, which are desired, but after a specific structural formation, imports should be discouraged gradually, as this process is hurting growth. The data implies that policies seem to overlook this fact. It is high time, for this to be considered.
Future studies should continue to look into the associations, with different endogenous indicators, and look into specific elements of export sector, such as Garments and knitwear for Bangladesh. Future studies should also look into agricultural sector and its association to growth. In conclusion, Bangladesh has emerged to be successfully implementing hypothesis, in short run and long run policy implementation, and may emerge as a model nation for developing countries, that can be compared to many different growth models, and this pace should be maintained.
Bangladesh country profile. (n.d.). Retrieved September 2, 2009, from Export promotion bureau, Bangladesh: http://www.epb.gov.bd/bangladesh_Country_Profile.php
Adelman, I. (1984). Beyond Export-Led Growth. World Development , 12, 937-949.
Agion, P. ,. (1992). A Model of Growth with Creative Destruction. Econometrica , 60, 323-351.
Agosin, M. (1991). Trade policy reform and economic performance: a review of the issues and some preliminary evidence. UNCTAD discussion papers no.41 .
Ahmed, N. (2000). Export response to trade liberalization in Bangladesh: a cointegration analysis. Applied Economics , 32, 1077-1084.
Alauddin, M. a. (1995). Labour Absorption and Agricultural development in Bangladesh: Prospects and Predicaments. World Development , 23 (2), 281-97.
Alauddin, M. A. (2005). Trade liberalization in Bangladesh: the process and its impact on macro variables perticularly export expansion. The journal of developing areas , 39 (1), 127-150.
Arslan, I. a. (1993). Export incentives,exchange rate policy and export growth in Turkey. Review of Economics and Statistics , 75 (1), 128-33.
Athukorala, P. (1998). Export Response to liberalization : the Sri-Lankan Experience. Hitotsubashi Journal of Economics , 39, 49-65.
Bacha, E. (1984). Growth with limited supplies of foreign exchange: a reappraisal of two-gap model. New York: Academic Press.
Bahmani and Janardhanan, A. (1993). Export growth and economic growth: an application of cointegration and error correction model. Journal of Development Areas , 27 (4), 532-42.
Bahmani, O. a. (1991). Exports growth and causality in LDC. Journal of development economics , 36, 405-15.
Balassa, B. (1982). Development Strategies in Semi-industrialized Economies. New York: Oxford University Press.
Balassa, B. (1985). Exports, Policy Choices, and Economic Growth in Developing Countries After the 1973 Oil Shock. ournal of Development Economics , 18 (1), 23-35.
Balassa, B. (1988). The lessons of East Asian Development: an Overview. Economic Development and Cultural Change , 36, 273-290.
Balassa, B. (1971). The structure of protection in Developing countries. Baltimore: John Hopkins University.
Banerjee, A. D. (1993). Cointegration, Error Correction and the Econometric Analysis of Non-Stationary Data. Oxford, England: Oxford University Press.
Bardhan, P. a. (1970). Models of Growth with imported inputs. Economica , 37 (148), 373-85.
Basu, P. a. (1991). Terms of trade fluctuations and economic growth in developing economics. Journal of Development Economics , 37 (12), 89-110.
Bayes, A. H. (1995). Trends in the external sector: trade and aid. Experience with economic reform: a review of Bangladesh's development , 243-97.
Bhagwati, J. (1978). Anatomy and consequence of Exchange Control regimes. Cambridge, M.A.: Balinger Publishsing Co.
Bhagwati, J. (1988). Export Promoting Trade Strategies : Issues and Evidence. The World Bank Research Observer , 3, 27-57.
Bhagwati, J. (1978). Foriegn trade regimes and Economic Development: Anatomy and Consequences of Exchange Control Regimes . Cambridge: Ballinger Publishing Company.
Bhargava, A. (1986). On the theory of Testing for Unit Roots in Observed Time Series. Review of Economic Studies , 53, 369-384.
Brooks, C. (2008). Introductory Econometrics for Finance. Cambridge: Cambridge University Press.
Caves, R. (1971). Export led growth and the new economic history. Trade Balance of Payments and growth , 495-534.
Chang, T. F. (2000). Exports, Imports and Income in Taiwan: An examination of Export-led Growth Hypothesis. International Economic Journal , 14 (2), 151-160.
Cheneray, H. a. (1966). Foreign assistence and economic development. American Economic Review , 56 (4), 679-733.
Choi, K. (1987). Theories of Comparative Economic Growth. Amestardam: Iowa state University.
Chow, P. (1987). Causality between Export Growth and industrial performance: empirical evidence from the NICs. Journal of development Economics , 26, 53-63.
Clarke, R. a. (1992). Trade policy reform and economic performance in developing countries. Manchester University Press.
Cline, W. (1982). Can the East Asian Model of Development be Geeralized? World Development , 81-90.
Cline, W. (1984). Exports of Manufactures from Developing Countries. Washington, D.C.: Brooking Institution.
Cooley, T. a. (1985). Atheoritical macroeconomics: a critique. Journal of Monetary Economics , 16, 283-308.
Copeland, L. S. (2007). Exchange rates and International Finance. Delhi: Pearson Education.
Corden, W. (1974). Trade Policy and Economic Welfare. Oxford: Clarendon Press,.
Corden, W. (1997). Trade policy and Economic Welfare. Oxford: Clarendon Press.
Domer, E. (1957). Essays in the theory of Economic Growth. Oxford: Oxford University Press.
Dornbusch, R. (1974). Tariffs and nontraded goods. Journal of International Economics , 28, 221-54.
Edward, S. (1993). Openness,trade liberalization and growth in developing countries. Journal of Economic literature , XXXI (3), 1358-93.
Edwards, S. (1993). Openness, trade liberalization and growth in developing countries. Journal of Economic Literature , 31, 1358-93.
Edwards, S. (1992). Trade Orientation, distortion, and growth in developing countries. Journal of Development Economics , 39 (1), 31-57.
Emery, R. F. (1967). The realtion of exports and economic growth. Kyklos , 470-486.
Engle, R. a. (1987). Co-integration and Error Correction: Representation, Estimation and Testing. Econometrica , 55, 251-276.
Esfahani, H. (1991). Exports,imports and economic growth in semi-industrialized countries. Journal of development Economics , 35, 93-116.
Fajana, O. (1979). Trade and growth: the nigerian experience. World Development , 7 (1), 73-78.
Feder, D. a. (2002). On Export and Economic Growth: The Case of Italy. Journal of International Trade and Economic Development , 11 (3), 323-340.
Ghatak, S. M. (1997). Exports, export composition and growth:Cointegration and causality evidence from Malaysia. Journal of Applied Economics , 29, 213-23.
Ghirmay, T. G. (2001). Exports, Investment, Efficiency and Economic Growth in LDC: An Empirical Investigation. Applied Economics , 33, 689-700.
Gonclaves, R. a. (1987). Intercountry comparison of export performance and output growth. The Developing Economics , 25, 3-18.
Granger, C. (1980). Testing for causality. Journal of Economic Dynamic and Control , 2, 329-52.
Greenaway, D. M. (2002). Trade Liberalization and Growth in Developing Countries. Journal of Development Economics , 67, 229-244.
Grenaway, D. a. (1994). What Does Liberalization Do for Exports and Growth? Weltwirtschaftliches , 130 (1), 152-174.
Grossman, G. a. (1991). Innovation and Growth in the Global Economy. Cambridge, M.A.: The MIT Press.
Harris, R. (1994). Using Cointegration analysis in econometric modelling. London: Harvester Wheatsheaf.
Harrison, A. (1994). Productivity, Imperfect Competition and trade reform. Journal of International Economics , 36 (1/2), 53-73.
Hatemi-J.A., a. I. (2000). Time-Series Evidence for Balassa's Export-Led Growth Hypothesis. Journal of International Trade and Economic Development , 9 (3), 355-365.
Hecksher, E. (1949). The effect of foreign trade on the distribution of income. In E. Hecksher, Readings in the Theory of International Trade. Blackston.
Helleiner, G. (1995). Manufacturing for Exports in the Developing World: Problems and Possibilities. London and Newyork: Routledge.
Heller, P. a. (1978). Export and growth: an empirical re-investigation. Journal of Development Economics , 5, 191-3.
Hossain, M. a. (2002). Export Response to the Reduction of Anti-Expot Bias: Empirics from Bangladesh. Brisbane: University of Queensland.
Hossain, M. a. (2004). Exports and Economic Growth inBangladesh Has manufacturing exports become a New Engine of Expory-Led growth? The international trade journal , 40 (3), 303-325.
Hossain, M. (2003). Trade Liberalization and Economic Performance: Analytics and Empirics from Bangladesh. Brisbne: School of Economics, University of Queensland.
Hrishman, A. (1981). A generalized linkage approach to development with special referee to staples. Cambridge: Cambridge University Press.
Hsiao, M. (1987). Tests of causality and exogeniety between export and economic growth: the case of Asian NICs. Journal of Economic Development , 12, 143-59.
Islam, M. a. (n.d.). Export Growth Nexus in a Small Open Economy: The Case of Bangladesh. (C. a. Weis, Ed.) The Economy of Bangladesh: Problems and Prospects .
Islam, M. (1998). Export Expansion and Economic Growth Testing for Cointegration and Causality. Applied Economics , 30, 415-425.
Jenkins, R. (1996). Trade liberalization and export performance in Bolivia. Development and Change , 27 (4), 693-716.
Jin, J. (2000). Openness and Growth: An Interpretation of Empirical Evidence from East Asian Countries. Journal of International Trade and Economic Development , 9 (1), 5-17.
Johansen, S. a. (1990). Maximum Likelihood Estinmation and Inference on Cointegration-with Application to the demand of Money. Oxford Bulletin of Economics and Statistics , 52, 169-210.
Johnson, H. G. (1958). Comparative Costs and Commercial Policy. Pakistan Economic Journal , 2, VIII.
Johnson, H. G. (1950). Factor Endowments, International Trade and Factor Prices. Review of Economic Studies , 1 (48), XIX.
Johnson, H. G. (1950). Optimum Welfare and Maximum Revenue Tariffs. Review of Economic Studies , 1 (48), XIX.
Jung, S. a. (1985). Exports, Growth and Cusality in Developing Countries. Journal of Development Economics , 18, 1-12.
Kaldor, N. (1940). A Note on Tariffs and the Terms of Trade”,. Economica , 28, VII.
Karunaratne, M. A. (2004). Exports and Economic Growth in Bangladesh: Has Manufacturing Exports Become a new Engine of Export-Led Growth. The International Trade Journal , 18 (4), 303-334.
Kavoussi, R. (1984). Export expansion and economic growth:further empirical evidence. Journal of Development Economics , 18, 1-12.
Kohli, I. a. (1989). Exports and Growth: Critical Minimum Effort and Diminishing Returns. Journal of Development Economics , 30, 391-400.
Kormendi, R. a. (1985). Macroeconomic determinants of Growth: a cross-country evidence. Journal of Monetary Economics , 16, 141-63.
Krishna, K. a. (1991). Technical Efficiency in Bangladesh manufacturing Industries. Bangladesh development studies , 19, 89-106.
Krueger, A. (1974). The Political Economy of the Rent-Seeking Society. American Economic Review , 291–303.
Krueger, A. (1997). Trade Policy and Economic Development: How we learn. American Economic Review , 87 (1), 1-22.
Krueger, A. (1998). Why Trade Liberalization is Good for Growth? The Economic Journal , 108, 513-522.
Krugman, P. (1984). Import Protection as Export Promotion: International Competition in the presence of Oligopoly and Economics of Scale. Monopolistic Competition in International Trade , 180-193.
Krugman, P. (1987). Is Trade Policy Passé?'. Economic Perspectives , 1, 131–44.
Krugman, P. (1986). New Thinking about Trade Policy. In P. Krugman, Strategic Trade Policy and the New International Economics (pp. 1–22). Cambridge, MA: MIT Press.
Kwan, A. a. (1990). Economic growth and the expanding export-sector:China 1952-1985. International Economic Journal , 5, 105-16.
Lee, C. a. (1988). Trade in east Asian development with comparative reference to the southeast Asian experience. Economic development and cultural change , 36, S139-S152.
Levine, R. a. (1992). A sensitivity analysis of cross-country growth regressions. The American Economic Review , 82, 942-63.
Levine, R. a. (1992). A sensitivity analysis of cross-country regressions. American Economic Review , 82, 942-63.
Lewis, W. (1966). Development Planning: the essential of economic policy. London: Allen and Unwin.
Little, I. S. (1970). Industry and trade in some developing countries: A Comparative Study. London: Oxford University Press.
Liu, X. S. (1997). An empirical investigation of the causal relationship between openness and economic growth in China. Applied Economics , 29 (12), 1679-86.
Lucas, R. (1988). On the Mechanics of Economic Development. Journal of Monetary Economics , 22, 30–42.
Maddala, G. (1977). Econometrics. New York: McGraw-Hill.
Mahmud, W. (1995). Recent Macroeconomic developments. Dhaka: University press limited.
Maizels, A. (1968). Export and Economic Growth in Developing Countries. London: Cambridge University Press.
Maizels, A. (1963). Industrial growth and World Trade. Cambridge: Cambridge University Press.
McKinnon, R. (1964). Foriegn Exchange Constraints in Economic Development. Economic Journal , 74, 388-409.
McMillin, W. (1988). Money growth volatility and macroeconomy. Journal of money, credit and banking , 20, 319-35.
Meade, J. (1951). Trade and Welfare: the theoty of International Economic Policy. London: Oxford University Press.
Meier, M. (1995). Leading Issues in Economic Development. New York: Oxford University Press.
Mellor, J. W. (1967). Developing Rural India. New York: Cornwell University press.
Michaely, M. (1976). Export and Economic Growth: An Empirical Investigation. Journal of Development Economics , 4, 49-53.
Michaely, M. P. (1991). Liberalising Foriegn Trade: Lessons of experience in the Developing World. Oxford: Basil Blackwell.
Michalopoulos, C. a. (1973). Growth of export and income in the developing world: a neoclassical view. Washington D.C.: International development discussion papers 28.
Moschos, D. (1989). Export expansion,growth and the level of economic development: an empirical analysis. Journal of Development Economics , 30, 93-102.
Nurkse, R. (1953). Problems of Capital Formation in Underdeveloped countries . Oxford: Basil Blackwell.
Ohlin, B. (1933). Inter-regional and International Trade,. Harvard University Press.
P.J.Dawson. (2005). The export- income relationship and trade liberalization in Bangladesh . Journal of Policy Modeling , 889-896.
Phillips, P. a. (1988). Testing for Unit root in time series regression. Biometrica , 75, 335-46.
Phillips, P. (1982). Understanding Spurious regressions in econometrics. Journal of econometrics , 33, 311-40.
Prebisch, R. (1952). Problemas Teoricos y Precticos del Crecimiento Economico. United Nations: Economic commissions of Latin America.
Rab, A. (1989). Value of Bangladesh's Policies to promote export. Dhaka: Unpublished report to World Bank.
Rahman, S. (1995). Trade and Industrialization in Bangladesh: An Assessment. (G. Helleiner, Ed.) Manufacturing for Exports in the Developing World: Problems and Possibilities .
Riedel, J. (1984, March). Trade as the Engine of Growth in Developing Countries, Revisited. Economic Journal , 56-73.
Rodrik, D. (1994). King Kong Meets Godzilla: The World Bank and The East Asian Miracle.
Romer, P. (1990). Endogenous Technical Change. Journal of Political Economy , 98, 71–102.
Romer, P. (1993). Idea Gaps and Object Gaps in Economic Development. Journal of Monetary Economics , 32 (3), 543-573.
Romer, P. (1989). What determines the rate of growth and technology change? World Bank Working Papers , 279.
Rushdi, A. (1982). Factor substitutability in the Manufacturing Industries of Bangladesh: An Internationl Comparison. Bangladesh Economic Review , 1, 32-46.
Sachs, J. a. (1995). Economic Reform and the process of Global Integration. Brookings papers on Economic Activity , 1, 1-18.
Salim, R. a. (1999). Sources of Output growth in Bangladesh Food Processing Industries: A Decomposition Analysis. The development Economics , XXXVII (3), 355-74.
Salim, R. (1999). Capacity Realization and Productivity growth in a developing country: has economic reform had impact? Aldershot: Ashgate Publishing limited.
Sandrum, R. (1994). Exports and Economic Growth. From Classical Economics to Development Economics , 104-20.
Sattar, S. A. (2004). Trade liberalization, growth and poverty reduction; the case of Bangladesh. Dhaka: The World Bank.
Shafaeddin, S. (1994). The impact of trade liberalization on export and gdp in least developed countries. Geneva: UNCTAD Discussion papers No.85, UNCTAD.
Shand, R. a. (1997). Economic Profiles in South Asia. Canberra: Australian National University.
Sharma, S. a. (1994). Causal Analysis between exports and economic growth in developing countries. Applied Economics , 26 (12), 1145-57.
Sheehay, E. (1995). Trade, efficiency and growth in cross section of countries. Welwritscheftliches Archives , 131 (4), 723-36.
Sims, C. (1980). Macroeconomics and Reality. Econometrica , 48, 1-48.
Singer, H. (1979). Policy Implications of the Lime Target. Industry and Development , 3, 17-32.
Streeten, P. (1982). A cool look at "outward looking" strategies for development. The World Economy , 5 (2), 159-169.
Teame Ghirmay, R. G. (2001). Exports, investment,efficiency and economic growth in LDC: an empirical analysis. Applied Economics , 33, 689-700.
The World Bank. (1996). Bangladesh: Trade Policy Reform for Improving the Incentive Regime,Report No. 15900-BD. Dhaka.
Thomas, V. J. (1991). Bets Practices in Trade policy reform. Oxford: Oxford University Press for the world bank.
Tybout, J. a. (1995). Trade Liberalisation and the dimensions of efficiency change in Mexican Manufacturing Industries. Journal of International Economics , 39 (1/2), 53-78.
Tybout, J. D. (1991). The Effects of Trade Reforms on Scale and Technical Efficiency: New Evidence from Chile. Journal of International Economics , 31, 231–50.
Viner, J. (1937). Studies in the Theory of International Trade,. Harper.
Wallis, K. (1074). Seasonal Adjustments and relations between variables. Journal of the American Statistical Association , 10 (1), 18-31.
Weiss, J. (1992). Export response to trade reforms: recent Mexican experience. Development policy review , 10 (1), 43-60.
Weisskopf, T. E. (1972). The impact of foreign capital inflow on domestic savings in underdeveloped countries. Journal of International Economics , 2, 25-38.
World Bank. (1996). World Development report. Washington, D.C.
Young, A. (1991). Learning by Doing and the Dynamic Effects of International Trade. Quarterly Journal of Economics , 106, 369-405.
Zivot, E. a. (1992). Further Evidence on the Great Crash, the Oil Price Shcok and the Unit Root Hypothesis. Journal of Business and Economic Statistics , 10, 251-270.