Growth and structural transformation: Options for Pakistan

Tasneem, Farah; Aamir Khan, Muhammad Article Growth and structural transformation: Options for Pakistan Research in Globalization Provided in Cooperation with: Elsevier Suggested Citation: Tasneem, Farah; Aamir Khan, Muhammad (2024) : Growth and structural transformation: Options for Pakistan, Research in Globalization, ISSN 2590-051X, Elsevier, Amsterdam, Vol. 8, pp. 1-11, https://doi.org/10.1016/j.resglo.2023.100190 This Version is available at: https://hdl.handle.net/10419/331114 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Research in Globalization 8 (2024) 100190 Available online 19 December 2023 2590-051X/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Growth and structural transformation – Options for Pakistan Farah Tasneem , Muhammad Aamir Khan * Department of Economics, COMSATS University Islamabad, Pakistan ARTICLE INFO Keywords: Production economics Trade liberalization Structural transformation Cross-sectoral linkages Pakistan ABSTRACT Structural transformation, which signifies a shift towards high value-added production and an elevated standard of living, is a vital ingredient for inclusive growth. Pakistan’s pace of structural transformation has been amongst the slowest and lowest compared to other countries in Asia. Against this backdrop, the purpose of this paper is to quantify the economic benefits of accelerating the rate of transformation (trade and productivity-oriented) in Pakistan. This research explores this relationship using an Input-Output (IO) framework in conjunction with a global Computable General Equilibrium (CGE) model. This allows for the evaluation of the economy-wide impact of exogenous changes in the trade and productivity of critical sectors. The IO framework helps in identifying these key Pakistani sectors by providing backward, forward, and inter-sectoral linkages for each sector. The results of the IO framework show that the sectors with strong backward linkages in Pakistan are “Food, Beverages and Tobacco”, “Textiles and Textile Products”, “Leather and Footwear”, and “Construction”. Whereas “Agriculture”, “Mining and Quarrying”, and “Wholesale and Commission Trade” are the sectors with strong forward linkages. The estimated results from a global CGE Model demonstrate that structural transformation towards promising sectors (manufacturing and services) has a positive impact on the macro-economic variables as well as at the household level in Pakistan. 1. Introduction Structural transformation is termed as a panacea for economic development, especially in the case of the developed world (Felipe, 2012). Structural transformation encompasses the reallocation of economic activities within three sectors of the economy; Agriculture, Manufacturing, and Services to endeavor economic well-being (Herrendorf et al., 2014). However, understanding the mechanism of how economies structurally transform as they grow is crucial for developing sensible growth strategies (Anderson and Ponnusamy, 2023). In many developing economies, the agriculture sector has lower productivity, and a considerable part of the labor force allocated in the agriculture sector hinders getting away from the poverty trap (Caselli, 2005; Dorosh and Thurlow, 2018). Lower productivity in the agriculture sector accounts for huge differences in income across the countries. Furthermore, structural differences also provide a brief insight into income differences (Ghosh et al., 2023; Gollin et al., 2007). The growth in the agriculture sector pushes capital and labor to manufacture and services, this shift from the non-capitalist (low wage and production) to capitalist (high wage and production) sector increases the total production, income, and overall economic efficiency. Innovation-based growth can also help stimulate the productivity of enterprises, thus bolstering exports (Huang et al., 2022; Vasin, 2022). Lahiri and Hnatkovska (2014) discussed that structural transformation in African, Latin American, and Asian developing economies has led to constant economic growth. The overall difference in the growth output of labor among these developing economies is because of the difference in the pattern of structural transformation. Many countries consider export-oriented structural transformation as a major driving force for economic development (Khan, 2020). Tariff liberalization broadened industrialization and triggered the trade of agricultural imports. Furthermore, liberalization can offer a set of new export opportunities to firms, spurring them to expand their production facilities (Jung, 2021; Ahmad et al., 2022; Gopalakrishnan et al., 2022). In South Korea, free trade policies have encouraged the structural transformation process more than trade restrictions (Teignier, 2018). Betts et al. (2017) find that Korean trade policies of subsidies and tariff liberalizations are key for possible structural transformation. Garrige et al. (2023) find that Urbanization and structural transformation emerge as key drivers of China’s house price boom. On the other hand, a drop in trade with a decrease in exports and a fall in industrial output was accompanied by subsidy liberalization. Both these trade reforms * Corresponding author. E-mail addresses: [email protected] (F. Tasneem), [email protected] (M. Aamir Khan). Contents lists available at ScienceDirect Research in Globalization journal homepage: www.sciencedirect.com/journal/research-in-globalization https://doi.org/10.1016/j.resglo.2023.100190 Received 6 September 2023; Received in revised form 14 December 2023; Accepted 14 December 2023 Research in Globalization 8 (2024) 100190 2 offset each other and leave unchanged structural transformation. Dragging trade costs and change in patterns of comparative advantage promotes skill-based structural transformation that leads to a gradual decline in manufacturing share in the world’s output (Cravino and Sotelo, 2019). Many economies of East Asia have attained the “miracle” of “growth with equity” but have also encountered worsened income inequality (Kanbur et al., 2014; Jain-Chandra et al., 2016; Piketty, 2015; Hillbom and Bolt, 2015; Khan et al., 2021). Kuznets (1955) explained the correlation between economic growth and income inequality with an inverted U-shaped curve. According to Kuznets, as economies grow, income inequality increases as capital will get more of that growth, and this leads to pro-rich growth. Hence, while the income distribution may initially worsen, absolute poverty tends to fall, as indicated by the case of East Asia (Lopes, 2019). Moreover, this pulls low-skilled labor into the high-productivity and income sector, and this leads to pro-poor growth (Ravallion and Chen, 2009). 1.1. Structural transformation in Pakistan Pakistan is among the few developing economies that achieved impressive economic growth and reduced poverty in the first 40 years of its existence. The high growth in the 1980 s due to a stupendous increase in the manufacturing sector resulted in a growth rate of 6 % per annum, the poverty rate was lowered from 46 % to 18 % while the inflation rate was low, and per capita income was almost doubled. The growth rate dropped to an average of 3 to 4 % in the 1990 s and poverty surged to 33 % of the total population. Table 1 illustrates an overview of Pakistan’s macroeconomic variables over time. Export’s share in GDP was around 16 % in the 1990s and 2000s, but during the last decade, it declined to only 8 %. However, when compared to Asian competitors, Pakistan’s pace of structural transformation has been amongst the slowest and lowest compared to other countries in Asia. The service sector is the largest growing sector whose contribution to employment is higher compared to other sectors. The service sector contributes 54 % to the GDP and about one-third of total employment in the case of Pakistan. This ratio has increased from 34.25 % in 1999–2000 to 37.6 % in 2017–2018. The alarming thing for Pakistan is the declining share of manufacturing value added in the percentage of GDP. Pakistan’s share of manufacturing in value-added, as a percentage of GDP in 2021 is 11 %, which is the lowest when compared to the South Asian average of 15 % and a lower middle-income average of 17 % (Fig. 1). Manufacturing plays a vital part in economic growth, this sector’s percentage share in GDP has not changed in the last 60 years in Pakistan. This decline is somewhat linked to the low productivity of the agricultural sector, the main provider of inputs, to the industrial sector coupled with a continuous rise in energy costs, higher taxes, and currency devaluation. It is pertinent to mention that the gain under the manufacturing sector is heavily concentrated in the textiles and food and beverages sectors, both accounting for almost 45 % of the manufacturing sector’s output in the case of Pakistan. Overviewing Pakistan’s exports basket, where textile and clothing, light manufacturing including leather, rubber and plastics, fruits and vegetables, and minerals are mainly those in which Pakistan has a comparative advantage. While in more sophisticated capital products Table 1 Pakistan’s macroeconomic variables over time. Pakistan 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 GDP Growth 11.35 4.2 10.21 7.5 4.4 4.9 4.2 7.6 1.60 4.7 4.2 GDP Per Capita (USD) 169 166 296 328 360 478 514 693 1024 1428 1547 Exports (% GDP) 7.7 10.85 12.48 10.42 15.5 16.70 13.44 15.68 13.51 10.60 8.2 Imports (% GDP) 14.66 22.3 24 22.8 23.3 19.4 14.6 19.5 19.3 17.0 17.5 Gross Capital Formation (% GDP) 15.63 16.22 18.48 18.32 18.91 18.54 17.2 19 15.8 15.7 16 Source: WDI. Pakistan is experiencing a slow rate of structural transformation. Fig. 1. Share of manufacturing in value added (% of GDP in Pakistan). Source: World Bank. F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 3 such as transport, fuels, metals, heavy manufacturing, machinery, etc., Pakistan has a relatively weak comparative advantage. Pakistan is struggling to accelerate its manufactured products, but performance is not up to the mark. Since 1970, the industrial production structure has mostly remained unchanged. Lower productivity affects Pakistan’s manufacturing sector, and consequently, exports. As a result, it ends up being less competitive. Pakistan must anticipate the global supply chains. Rising East Asian economies are focusing on finished goods, while Pakistan, like many developing countries, is stuck with exports of raw goods. Moreover, the path followed to implement this structural shift from agricultural sectors directly to sectors concerning services while overlooking the development of the manufacturing sector in Pakistan is akin to that taken by most developing countries (Khan, 2020). Pakistan’s pace of structural transformation has been amongst the slowest and lowest compared to other countries in Asia. Against this backdrop, the purpose of this paper is to quantify the economic benefits of accelerating the rate of transformation (trade and productivityoriented) in Pakistan. Using an Input-Output (IO) and global Computable General Equilibrium (CGE) framework, this paper assesses the economy-wide impact of exogenous increases in the productive and exportable capacity of strongly backwardly and forwardly linked manufacturing and services sectors. With an emphasis on trade and productivity-centered growth, a global CGE model is developed to quantify the macroeconomic and household-level impact of structural transformation. The rest of the study is organized as follows. First, we present a literature summary on structural transformation and its different aspects, followed by a detailed section on the methodological framework, data sets, IO, and CGE frameworks used in this study. The results are then discussed in Section 3, followed by concluding remarks in Section 4. 2. Summary of literature on structural transformation and its different aspects The literature related to structural transformation and economic growth can be broadly categorized into three types of studies, i.e., one type is based on decomposition methodology, the second is regressionbased/econometrics studies, and the last one uses the general equilibrium framework. The first method is a purely descriptive accounting technique that decomposes the changes in aggregate productivity into changes in productivity within sectors (called within effects or intra effects) and changes in productivity due to the reallocation of labor across the sectors (Ahson et al., 2017). Decomposition studies have shown mixed results. Fagerberg (2000), Timmer et al. (2015), and Ahson et al. (2017) found intra-effects stronger than the structural transformation effect in bringing change in aggregate productivity. Hasan et al. (2013) examined different contributions from within effects and structural transformation effects in 15 different Indian states. Nguyen et al. (2018) investigated the effect of structural transformation on overall labor productivity in Vietnam and found that nearly half of the contribution was from structural transformation. Some researchers used econometric techniques (McMillan and Harttgen, 2014). Silva and Teixeira (2011) used one-step GMM and found a positive influence of structural changes on productivity growth in Japan and 20 OECD economies. Structural transformation does not always lead to a positive impact on income distribution and increasing inequality. Using the average Gini coefficient, a higher level of income inequality was noticed in Latin American and African regions from 2002 to 2009. Particularly sub-Saharan Africa has a higher level of inequalities globally, even after the early stages of growth. The higher level of inequalities has proved to be less poverty alleviating instead of enhanced development. Hence, poverty alleviation in this region is slow instead of a decisive growth decade (Baek, 2018). Andersson and Chaverra (2016) examined the structural change and income inequality in 27 growing economies for the period of 1960–2010. The authors used the inter-sectoral productivity gap to examine structural changes and to analyze their effect on income inequality. The results show a positive correlation between the intersectoral gap and income inequality. Along with this, the nature of growth in most of these economies is characterized by the development of the services sector rather than the industrial sector, which is labor intensive. Some subsectors have intensive high-skilled labor and increase income inequality, while others cause low productivity with less skilled labor. International trade plays a significant role in structural transformation. The speed of trade-induced structural transformation depends on the comparative advantage of a developing country relative to a developed one. (Sen et al., 2020) studied the Long-term Impact of Trade Wars and ‘Make in India’ policy on the Indian Economy using a dynamic CGE model and concluded that the combined impact of both policies has somewhat negative ramifications for exports, jobs, and investment growth in India. Teignier (2018) analyzed the neoclassical two-sector growth model with a closed and open economy. In autarky, reallocating factors from agriculture (because of low-income elasticity) to manufacturing and services in a country increases growth. International trade may boost structural transformation or hinder it, depending on the relationship between domestic and international prices. In South Korea and the US, trade positively affected the structural transformation process as the price of agricultural goods was lower in the international market than in the domestic market. Countries imported agricultural commodities and somehow saved their agriculture sector. Yi (2010) used a two-country and three-sector model to examine the impact of structural transformation in an open economy and concluded that trade could trigger a “hump” in industrial employment share. Khan (2020) studied cross-sectoral linkages to explain the structural transformation in Nepal using Input-Output and a computable general equilibrium model. The results show that sectoral and labor productivity in manufacturing and services have significant positive macro–micro effects on Nepal’s economy. Chen et al. (2023) evaluated the impact of the US-China Trade War and found that this adversely impacts innovation in the Chinese ICT industry. Norbu et al. (2021) studied the structural transformation and production linkages in Asia-Pacific countries using the Input-Output models. They found that most of the growth in the Asia-Pacific less developed countries has come about by increasing the output of existing products, with limited diversification. Roberts (2022) explores the case of a “prematurely deindustrialized” post-apartheid South Africa. Like Pakistan, South Africa underwent a structural shift featuring a declining share of the manufacturing sector in GDP accompanied by a rising share of the services sector in GDP, with statistics ranging from 21 % in 1994 to 13.3 % in 2016 and 60 % in 1994 to 68 % in 2016, respectively. Parallels can also be drawn between the cases of Indonesia and Pakistan. Sulistiawati and Lestari (2022) employ a Spearman rank correlation test to rank shifts in sectoral GDP and employment caused by the Industrial Revolution 4.0 and the COVID-19 pandemic in Indonesia. They identify education and trade services sectors as key sectors that are not only absorbing large proportions of labor but also contributing significantly towards GDP. Gollin (2021) investigates the skewed patterns of development in the rural and urban areas of sub-Saharan Africa and finds that despite 60 % of the rural population being engaged in agricultural activities, production seems to lag severely behind that of other areas. 3. Methodology 3.1. Input-Output model This research first incorporates the Input-Output framework to identify backward and forward linkages in Pakistan’s economy over time. The Input-Output table has a distinctive feature that gives a F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 4 mechanism for specifying direct and indirect linkages between total production and trade (Mercer-Blackman et al., 2017). Backward linkages show demand-driven interlinking with the upstream sector, while forward linkages indicate the relationship of supply-providing downstream sectors. Whereas total linkages contain both backward and forward linkages. The sector has strong backward and forward linkages considered a critical sector. The reason behind classifying sectors with the highest linkages is that any increase in investment or productivity of these critical sectors would bring a much more positive spread over the rest of the economic sectors having low linkage values (Khan, 2020). In policymaking, sectors with higher linkage values possess more importance. These IO Tables are for the year 2010–2020 and are provided by the Asian Development Bank (Asian Development Bank 2022 Asian Development Bank. 2022. Pakistan: Input-Output Economic Indicators, 2022) These tables are of dimension 35 ×35 sectors and the detailed sectoral list is shown in Appendix A. To find cross-sectoral linkages between the industries, the following formulae are applied to the Pakistan IO Tables of the years 2010 to 2020. (1) Direct and Total Backward linkages (Chenery and Watanabe, 1958) BL(d)j=∑ n i=1 aij (1) BL(t)j=∑ n i=1 lij (2) (2) Direct and Total Forward linkages (Chenery and Watanabe, 1958) FL(d)j=∑ n j=1 bij (3) FL(t)j=∑ n j=1 gij (4) Reference (Beyers, 1976; Jones, 1976; Chenery and Watanabe, 1958). where: BL: Backward linkages FL: Forward Linkages d:Depicts ‘Direct’ linkages t:Depicts ‘Total’ linkages a ij : Technical coefficient b ij : Elements of allocation. The Chenery–Watanabe linkage portrays effects irrespective of the economic sector’s total production. This approach provides direct economic impact by incorporating Input-Output matrices’ rows and column multipliers. Rasmussen’s approach considers indirect economic effect and overcomes the drawback of the Chenery–Watanabe approach (Miller and Blair, 2009). The sectors with strong backward and forward linkage values impact domestic demand for production (Cai and Leung, 2004). The linkage value of more than one represents that the sector draws more than average, and the linkage value of less than one indicates that the sector draws less than average from the economic system (Parra and Wodon, 2008). The interpretation of linkages is as follows: Key sector: BL and FL >1. Backward oriented sector: BL >1 and FL <1. Forward oriented sector: BL <1 and FL >1. Weak sectors: BL <1 and FL <1. 3.2. Global computable general equilibrium (CGE) model This research used a multiregional, multisector, computable general equilibrium model with perfect competition and constant returns to scale. The Computable General Equilibrium (CGE) model is an economic model that uses real-world economic data to examine changes in government policy, technology, and the environment. It is a multi-sectoral model that expresses explicit data on economic agent behavior. Consumers are utility-maximizing agents, while producers are modelled as cost-cutting and profit-maximizing actors in the economy. Agents are assumed to make production and consumption decisions based on prices set by demand and supply equilibrium conditions. The CGE model is a commonly used and acceptable tool in the investigation of welfare, particularly poverty, and inequality (Savard, 2003). 3.2.1. Global trade analysis project (GTAP) GTAP is a globally consistent sectoral and regional model of broad categories of the CGE family. The GTAP model is broadly based on accounting and behavioral equations. Accounting equations maintain the revenues and expenses of all agents in an economy. Whereas behavioral equations explain the response of the economic agent. This paper uses standard GTAP model’s nested CES production functions, in which primary factors of production exhibit constant elasticity of substitution ( σ VA), impacting the economy’s ability to modify its output mix in response to relative prices and factor endowments. These parameters also have an impact on sectoral supply responses, particularly in those with specialized or slowly changing production components. On the other hand, composite value added and intermediates are used in a fixed proportion. However, users, if required, can define the elasticity of substitution (ESUBT) between composite intermediates and composite value-added at the top production nest. 3.2.2. Producer behavior The GTAP model contains two types of commodities that are differentiated as produced for either the domestic market or the international market. These commodities are produced under constant elasticity of transformation and are considered imperfect substitutes for each other. Equations (5)–(7) represent the total output of commodity ‘i’ in region ‘r’, and the supply of a particular commodity for the domestic and export markets. Yir =[ α Y irD1+1/n ir +βY irX1+1/n ir ]1/(1+1/n)(5) Dir =Yir α D ir (pD ir ,pX ir)(6) Xir =Yir α X ir(pD ir ,pX ir)(7) The demand for primary factors (capital, land, labor, and intermediate inputs) is considered to be directly proportional to the total level of production and can be expressed as in Equation (8); IDir =∑ j Yjr α ijr (8) Here, α ijr is a coefficient of the demand for the intermediate input and taken inelastic to price in the core model. In the production of specific goods ‘i’, region ‘r’ consumes both domestic and imported intermediate commodities and following Armington (1969) both types of inputs are taken as imperfect substitutes. Thus, demand for the compound intermediate goods can be; IDir =[ α I irDI ρ ir +βI irMI ρ ir]1/ ρ (9) where DIir is demand for domestic intermediate goods and MIir represents imported intermediate goods. The Cobb–Douglas production function is used to represent a functional form of relationship between the level of inputs and outputs. By assuming Yjr as given and the condition of linear homogeneity holds, then this production function can be rewritten as a compensated demand function, depending on factor prices and taxes levied on those F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 5 factor inputs as; FDfir =Yir α F fir(pF r,tF ir)(10) The total output of the public sector is assumed to be exogenous and hence is fixed for a given time, and input used for production is elastic to prices and applicable taxes, hence; GDir =Gr α G ir (pD ir ,pM ir ,tG ir )(11) 3.2.3. Consumer behavior In this model, the final demand is determined by the demand of commodity ‘i’ by a representative agent in the region ‘r’. A consumer is supposed to be rational and wants maximum utility and functional form is assumed to be a Cobb-Douglas utility function as follows; Ur=∑ i θC irlog(CDir )(12) CDir =[ α C irDC ρ ir +βC irMC ρ ir]1/ ρ (13) CDir represents demand for commodity by an economic agent and the equation can be restated to obtain aggregate final demand as follows; CDir =θC irMr pC ir(1+tC ir )(14) pir represents the unit price of the composite commodity (aggregate of domestic and imported goods) and tir is a notation for tax. Where Mr is the total expenditures of region ‘r’. Mr=∑fpF frFfr +∑itY ir(pD ir Dir +pX irXir)+∑ijtID ijr pID ir Yjr α ijr +∑fitF firpF frFDfir +∑itG ir pGD ir GDir +∑itC ir pCD ir CDir +∑istX irspX irsMirs +∑istM isr(pX isMisr(1+tX isr)+pTTisr )−∑ipG ir Iir −∑ipG ir (1+tG ir )GDir −pC nBr (15) In the core GTAP model, capital flow is also supposed to be exogenous and hence held fixed. The GTAP data and model assume that every region/country has a representative household. This household is a combination of both government and private households. This assumption is a bit unrealistic; hence, one cannot find the impact of any policy scenario on private household income and government revenue separately. Against this backdrop, this research used an updated extension of the model, known as the MYGTAP model (Minor and Walmsley, 2012). 3.3. MyGTAP model The MyGTAP model was developed by Minor and Walmsley (2012) and is an extended version of the GTAP model (Hertel and Tsigas, 1997). The GTAP model is based on a common global database, the GTAP database (Aguiar et al., 2016). In the standard GTAP model, there is only one private representative household, which has characteristics of both government and household income making it difficult to differentiate the impact on the individual entity. The core of this research is to find out the impact of structural transformation on a household with different income levels. In the MyGTAP model, income sources of government are taxes (TTAX), foreign aid in (AIDI), less foreign aid out (AIDO), and other transfers (TRNG) to a private household from the government (Equation (16)). GOVINC(r) = AIDI(r) − AIDO(r) + TTAX(r) − sum(h,HHLD,TRANG(h,r) (16) This income is further used for government expenditure and government savings are shown in Equations (17) and (18), respectively. Furthermore, the gap between government income and consumption can be either deficit or saving, Equation (19)) it can be assumed that government expenditure can remain constant or the government can fix the deficit (Minor and Walmsley, 2012). yg(r) − gincome(r) = dpgov(r) − dpgav(r)(17) psave(r) + qgsave(r) − gincome(r) = dpgsave(r) − dpgav(r)(18) ovdef (r) = gincome(r) − yg(r)(19) Private households obtain income from factors (EVOAH) less depreciation (VDEPH), plus net foreign labor remittances (REMIH and REMOH) and foreign capital income (FYIH and FYOH), transfers between households (TRNH), and from the government (TRNG) (Equation (20)). HHLDINC(h,r) = sum[i,ENDWCOMM,EVOAH(i,h,r) ] − VDEPH(h,r) +REMIH(h,r) − REMOH(h,r) + FYIH(h,r) +sum[k,HHLD,TRNH(k,h,r) − TRNH(h,k,r) ] +TRNG(h,r) (20) The private household income is divided into private consumption and saving by using Cobb–Douglas. Regional savings are the sum of savings of government and private households. Savings are adjusted to Fig. 2. Direct Backward Linkages (2010–2020). F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 6 create an equilibrium between income and expenditures (Minor and Walmsley, 2012). SAVE(r)*qsave(r) = SAVGOV(r)*qgsave(r) +sum(h,HHLD,SAVHHLD(h,r)*qhsave(h,r)(21) 3.3.1. Database and aggregations This study uses the latest global GTAP 10 database (Aguiar et al., 2016) and Pakistan’s Social Accounting Matrix (SAM) for 2010–2011. GTAP 10 is a globally consistent multi-sectoral and regional data tailored for international policy issues such as trade, production economics, migration, development, energy, the environment, etc. The model is known as the GTAP model and is coded using GEMPACK, see (Hertel and Tsigas, 1997). There is a single regional household in standard GTAP models, and all factor income accumulates in this aggregated household type. Moreover, this single household uses all regional consumption and savings. However, this research includes multiple HH types and factors of production taken from the latest Pakistani Social Accounting Matrix (SAM). The Social Accounting Matrix (SAM) of Pakistan contains detailed information for multiple household (HH) types based on geographical zones and at the province level. These 16 multiple HHs derive their income from 12 factors of production (Appendix B) and foreign remittances. Out of these 12 factors, there are five categories of labor, i.e., small farmer, medium farmer, farmworker, non-farm skilled, and nonfarm unskilled workers. Three types of land, i.e., land small, medium, and large, while capital includes agricultural capital, formal capital, and informal capital. Most of the factors of production are used in the production of three agricultural commodities (grain crop, vegetables, and fruit and livestock meat). In contrast, non-agriculture factors non-farm formal capital, informal capital skilled, and non-farm un-skilled workers are used across all other sectors except grain crops (Appendix C). Household types are divided into land proprietorship, farm size, and non-farm activity. Data from SAM, especially the Household and Factor types, are fed into the latest GTAP database keeping in mind the data and share homogeneity. 3.4. Policy Experiment/Simulation This research aims to quantitatively assess the trade and productivity-oriented structural transformation on the macro as well as at the household level in Pakistan. The textile industry is one of Pakistan’s essential export sectors, contributing 8.5 % of the total GDP and accounting for almost 60 % of Pakistan’s exports. The Pakistan Ministry of Commerce has aimed to rationalize the tariff structure for duty-free access to import raw materials for the manufacturing and export sector in the National Tariff Policy of 2019–2024. With this backdrop, this research quantifies the impact of export-oriented structural transformation that includes up-gradation of the institution and labor reallocation in the export potential sector on the macro as well as at the household level in Pakistan. Secondly, this study examines the role of increased textile productivity and the effects of trade liberalization and protectionism (trade war). The initial tariffs between Pakistan and ROW are shown in Appendix D. The simulation/experiment design is given below. Simulations Code Description SIM-I Manufacturing A 10 % increase in Sectoral and Labor productivity in the Light and Heavy manufacturing sectors for Pakistan. SIM-II Services A 10 % increase in Sectoral and Labor productivity of the Services sector for Pakistan. SIM-III Textile A 10 % increase in Sectoral and Labor productivity of textile and clothing productivity for Pakistan. SIM-IV Trade Lib A 5 % decrease in import tariffs across the board (All countries/regions and in all tradable commodities). SIM-V Trade War A 5 % increase in import tariffs across the board. (All countries/regions and in all tradable commodities). 4. Results 4.1. Sectors with higher backward and forward linkages in Pakistan and value-added multiplier Identifying sectors with strong backward and forward linkage values is important because any increase in productivity or investment in these sectors translates into a multiplier effect on the other allied sectors. The figure demonstrates the backward and forward linkages of Pakistan’s economy for the 2010 and 2020 years, respectively. In Pakistan, the “Wood and Products of Wood and Cork”, “Pulp, Paper, Paper Printing and Publishing”, “Coke, Refined Petroleum and Nuclear Fuel”, “Chemicals and Chemical Products and Electricity,” “Gas and Water Supply” sectors belong to the light manufacturing sector and are generally a dependent sector. They have substantial backward and forward linkages value. An economic sector with strong backward and forward linkage value is more connected with the manufacturing sector. The sectors with strong backward linkages in Pakistan are “Food, Beverages and Tobacco”, “Utilities” “Textiles and Textile Products”, “Leather, Leather and Footwear”, “Construction”, “Hotels and Restaurants” and “Transport and Communication”. This implies that the input of these industries comes from other sectors of the economy and these Fig. 3. Direct Forward Linkages (2010–2020). F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 7 sectors depend more on intermediate goods, which are mostly capitalintensive. Whereas, “Mining and Quarrying,” Wood and its product”, “Wholesale and Commission Trade,” “Retail Trade, Sale, Maintenance and Repair”, “Other Supporting and Auxiliary Transport Activities”, and “Financial Business and Services” have strong forward linkages. This means the output of these sectors is used in other sectors as an input on a larger scale. Figs. 2 and 3 illustrate the top 3 sectors with the highest Backward and forward linkages for the year 2010–2020 respectively. 4.2. Impact on Macroeconomic variables of Pakistan Table 2 demonstrates the impact of all experimental designs in this research on macroeconomic variables, real GDP, terms of trade, welfare, government income, and volume of real imports and exports. All changes are measured at base prices of 2014. We find that an increase of 10 % in the manufacturing sector’s output (export-oriented structural transformation) is expected to have a positive impact of 3.48 % on Pakistan’s real GDP, equal to a dollar value of 7441.5 million USD. Whereas transformation (shift of labor productivity towards export-oriented subsectors) towards the services sector with an increase in output of 10 % is expected to have a positive and significant impact on real GDP by 7.11 %, which has a monetary worth of 15185.03 million USD. Textile and wearing apparel are among the top exportable commodities of Pakistan, and an increase of 10 % production of textile predicts an increase of 1.02 % in real GDP, which have a monetary value of 2177.83 million USD. Trade liberalization (reduction in tariff) affects an economy in two ways, in the first one, reducing the tariff on the final commodity leads to a decrease in the price of imported commodities and an increase in the price of the domestically produced commodity, which affects the domestic industry. On the other hand, a decrease in tariff on intermediate commodities results in an increase in domestic production coupled with an increase in export demand because intermediate inputs produce valuable forward linkages to domestic production; hence this moves towards industrialization. As depicted in the table, a 5 % tariff reduction has a small but positive effect of 0.03 % on real GDP with a value of 73.89 million USD. However, an increase of 5 % tariff (trade war) negatively affects real GDP. Thus, Tariff liberalization in the case of Pakistan yields a positive impact on growth. The most widely used measure for the welfare effect is Equivalent Variation (EV). This can be further decomposed in the number of compositions which include ‘allocative efficiency’, ‘term of trade’, and ‘change in capital stock’. In allocative efficiency, marginal production costs are equivalent to the output’s marginal utility. The term of trade (TOT) is also associated with overall welfare as it is the ratio of export pricing to the price paid for imports. The overall value of welfare is increased with the increase in production level in the services sector with the value of 14,773 and the level of welfare in the manufacturing sector is 8270 million USD. Trade war has a small but positive impact of 8.38 million USD and in the case of trade liberalization, the welfare effect is negative. The real export of services and textiles is projected to increase but there is a decline in export in the case of manufacturing. The reason behind this is labor productivity and resources tend to move toward the manufacturing sector, so the output level of exportable agricultural commodities will fall. As Pakistan has a massive fraction of agri-based exportable goods, the shift of resources will cause a fall in the overall export level. After the imposition of a 5 % tariff on import commodities, real imports are expected to go down, and terms of trade will improve. However, liberalization brings more growth opportunities. Pakistan is expected to gain more in terms of allocative efficiency due to liberalization compared to protectionism. We find the same results in the case of Trade liberalization, i.e. allocative efficiency increases. This might be due to a shift of factors from an inefficient sector to an efficient sector resulting in a decline in wastage or increased efficiency in an economy. This will impact domestic prices which in turn help market forces to reduce the gap between unit prices of products and consumer willingness to pay. And hence reduces deadweight loss. Other simulation results positively impact real imports as the import of intermediate inputs for production will rise. Hence, the term trade is negative for textile, services, and trade liberalization but positive for manufacturing productivity. Government income is projected to increase in all simulations except for trade liberalization. In the trade liberalization scenario, we altogether remove the tariffs, so government income is expected to fall substantially. Tariffs are one of the primary income sources for the government. Services and manufacturing have a more positive impact Table 2 Impact on macroeconomic variables of Pakistan (% Changes, Constant 2014 Prices). Indicators Manufacturing Services Textile Trade Lib Trade War % (Million USD) % (Million USD) % (Million USD) % (Million USD) % (Million USD) Real GDP 3.48 (7441.5) 7.11 (15,185.03) 1.02 (2177.83) 0.03 (73.89) −0.04 (−74.84) Welfare 8270 14,773 2468 −11.1 8.38 Real Exports −16.03 (−4959.01) 15.09 (4667.59) 4.05 (1251.59) 1.53 (473.98) −1.49 (−461.45) Real Imports 4.03 (2292.53) 7.66 (4354.66) 4.5 (2560.93) 0.64 (365.43) −0.63 (−356.57) Terms of Trade 3.27 −1.91 −0.22 −0.11 0.10 Government income 4.44 7.79 3.29 −1.4 1.36 Source: Author’s simulations. Table 3 Changes in trade (percentage change from baseline). Commodities Manufacturing Services Textile Trade Lib Trade War Export Import Export Import Export Import Export Import Export Import Grain Crops −12.4 3.98 −34.04 33.45 −7.47 6.38 −0.3 1.40 0.29 −1.36 VegFruit −8.33 8.62 −18.79 16.80 −4.48 3.72 0.74 0.44 −0.57 −0.43 MeatLstk −27.75 15.7 −55.64 44.27 −9.54 4.51 −2.14 0.59 1.03 −0.6 Extraction −32.23 12.55 −24.66 12.25 −8.56 −0.56 0.39 −0.33 −1 0.33 ProcFood −20.84 16.57 −17.73 15.15 −9.66 6.04 0.73 1.29 −1.05 −1.27 TexWap −28.39 11.74 24.80 −0.56 21.01 −5.06 1.4 1.57 −2.34 −1.53 LightMnfc 32.99 −6.1 21.40 16.28 −14.74 8.10 0.9 1.94 −1.26 −1.86 HeavyMnfc 17.74 −0.36 −3.67 5.38 −12.75 4.94 0.29 0.58 −1.41 −0.58 Util_Cons −12.62 14.01 −10.62 13.97 −9.72 6.90 1.2 −0.37 −1.17 0.37 TransComm −19.22 16.18 −12.34 11.47 −9.55 6.47 0.87 −0.37 −0.85 0.36 OthServices −21.97 8.99 132.77 −11.51 −11.07 6.84 0.86 −0.25 −0.84 0.25 Source: Author’s simulations. F. Tasneem and M. Aamir Khan Research in Globalization 8 (2024) 100190 8 on a macro level. To sum up, structural transformation towards services has a more positive and significant impact on macroeconomic variables except in the case of terms of trade. 4.3. Impact on sectoral trade of Pakistan An increase in total productivity positively impacts the balance of trade and enhances trade volume, which leads to an increase in exports and imports. The first simulation of an increase in the manufacturing sector’s total output accompanied by an increase in labor productivity positively impacts the manufacturing sector’s total output and exports. Light manufacturing exports increased by 32.99 % and the export of heavy manufacturing increased to 17.74 %. However, the export of other commodities falls. In the second simulation, the model predicts an increase in the exports of wearing apparel by 24.80 % and the export of light manufacturing by 21.40 %. While the export of other commodities falls. This is due to the reallocation of factors of production in the manufacturing sector due to an increase in the real factor wages. In the trade liberalization scenario, those goods in which Pakistan holds a comparative advantage, export rise; for instance, processed food, textile, and wearing apparel, and light manufacturing which includes leather, rubber and plastic. It is pertinent to mention that exports of grain crops and meat livestock are negative, the reason is attributed to the fact that when Pakistan reduces the tariffs on these goods, there will be a surge of imports from highly competitive regional players, especially India. It is notable to mention that Indian exports to Pakistan (Pakistani Imports) in Grain crops and meat livestock have increased substantially. Tariff reduction and lower prices for the import of intermediate goods decrease the domestic cost of production which is followed by an increase in export in those commodities in which we hold a comparative advantage. A trade war will restrict the export and import volume. It is pertinent to mention that Pakistan still has very high tariffs compared to regional South Asia partners. We find that trade liberalization is generally better for Pakistan compared to protectionism. Table 3 explains changes in trade (Table 4). 4.4. Impact on real factor rewards One of the main features of SAM is, that it introduces many factors (land, labor, capital) that contribute to examining the impact of structural transformation and the other three simulations on the economy of Pakistan. There are twelve types of factors of production, SAM has five categories of labor, i.e., small farmer, medium farmer, farmworker, nonfarm skilled, and non-farm un-skilled workers. Along with this, land contains land small, land medium, and land large while capital includes agricultural capital, formal capital, and informal capital. Most of the factors of the production are being used in the production of three agricultural commodities (grain crop, vegetables, and fruit and livestock meat) while non-agriculture factors non-farm formal capital, and informal capital skilled, and non-farm un-skilled workers are being used across all other sectors accept grain crop. Table 5 indicates that all simulation design has different impact across all the types of real factor rewards. Firstly, an increase in manufacturing productivity will reduce the reward of the labor farm worker, land large, land medium, land small, and capital agriculture and along with other factors associated with agriculture productivity such as labor small farm and labor-medium +farmer has low reward as compared to other real factors. Land small, land medium, and land-farm labor are mainly involved in the production of grain crops and vegetable fruits. An increase in manufacturing productivity will hurt the output of grain crops and have a moderate impact on vegetable fruit; hence, the real factor is affected. This is because capital and labor have shifted from the agriculture sector to the manufacturing sector which will diversify the productivity level in agriculture, which ultimately affects real factors of return. The simulation of the rise of productivity of services will have a positive and increasing impact on all real factors of return. Increased productivity of services will increase the total output of all exportable commodities except vegetable fruit and extraction. The increase in export will primarily affect all the factors of production, thereby leading to an increase in real factors of return. The extraction of total output is negative. The main factors involved in extraction are capital formal and labor nonfarm low-skilled labor. This can be shown in the table that these factors have a low real return. Vegetable fruit possesses all types of factors in production, here, the negative impact of the output of vegetables and fruit is distributed in all factors of return. In consequence, it has not affected many factors of return. Increased production of textiles will reduce real factors return of three land categories, labor small farmer, labor farmworker, and labor medium former. This simulation will affect all factors mainly associated with the agriculture sector that ultimately affect the real return of this agriculture sector. While capital formal and capital informal and low-skilled farm and non-farm labor will have a Table 4 Impact on real factor rewards in Pakistan. Factor Types Manufacturing Services Textile Trade Liberalization Trade War flab_s 0.21 12.72 −0.73 0.67 −0.66 flab_m 0.08 13.41 −0.68 0.65 −0.64 flab_w −0.82 12.29 −2.14 0.61 −0.6 flab_l 5.07 5.34 1.9 −0.01 0.01 flab_h 4.45 0.29 1.85 −0.11 0.11 flnd_s −0.28 10.54 −0.43 0.91 −0.89 flnd_m −0.63 11.2 −0.39 0.91 −0.89 flnd_l −1.02 11.92 −0.35 0.92 −0.9 fliv 2.76 19.56 −1.31 −0.12 0.11 fcap_a −1.06 11.82 −0.44 0.92 −0.9 fcap_f 6.07 5 1.85 −0.07 0.07 fcap_i 5.41 3.09 1.79 −0.08 0.08 Source: Author’s simulations. Table 5 Increasing tariffs will impact the household income of Pakistan. HH Types Manufacturing Services Textile Trade Lib Trade War Rural small farmer (quartile 1) 4.53 14.43 1.53 0.43 −0.43 Rural small farmer (quartile 234) 4.45 14.81 1.46 0.45 −0.44 Rural medium + farmer (quartile 1) 2.96 16 1.14 0.66 −0.65 Rural medium + farmer (quartile 234) 3.1 15.93 1.19 0.63 −0.62 Rural landless farmer (quartile 1) 4.25 14.13 1.59 0.51 −0.5 Rural landless farmer (quartile 234) 4.61 13.7 1.71 0.44 −0.43 Rural farm worker (quartile 1) 6.05 12.17 1.64 0.03 −0.03 Rural farm worker (quartile 234) 6.97 10.99 2.14 −0.08 0.08 Rural non-farm (quartile 1) 8.8 7.12 3.48 −0.2 0.2 Rural non-farm (quartile 2) 9.05 6.79 3.51 −0.21 0.21 Rural non-farm (quartile 3) 9.25 6.69 3.54 −0.21 0.21 Rural non-farm (quartile 4) 9.85 7.47 3.59 −0.22 0.21 Urban (quartile 1) 8.54 7.59 3.3 −0.14 0.14 Urban (quartile 2) 8.84 7.11 3.41 −0.18 0.18 Urban (quartile 3) 9.17 7.05 3.47 −0.2 0.2 Urban (quartile 4) 9.83 7.8 3.55 −0.2 0.2 Source: Author’s simulations. F. Tasneem and M. Aamir Khan