A Gravity Model Approach towards Pakistan's Bilateral Trade with SAARC Countries

Jan, Waheed Ullah; Shah, Mahmood Article A Gravity Model Approach towards Pakistan's Bilateral Trade with SAARC Countries Comparative Economic Research. Central and Eastern Europe Provided in Cooperation with: Institute of Economics, University of Łódź Suggested Citation: Jan, Waheed Ullah; Shah, Mahmood (2019) : A Gravity Model Approach towards Pakistan's Bilateral Trade with SAARC Countries, Comparative Economic Research. Central and Eastern Europe, ISSN 2082-6737, De Gruyter, Warsaw, Vol. 22, Iss. 4, pp. 23-38, https://doi.org/10.2478/cer-2019-0030 This Version is available at: https://hdl.handle.net/10419/259215 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-nc-nd/4.0 Waheed Ullah Jan, Mahmood Shah Comparative Economic Research. Central and Eastern Europe Volume 22, Number 4, 2019 http://doi.org/10.2478/cer‑2019‑0030 A Gravity Model Approach towards Pakistan’s Bilateral Trade with SAARC Countries Waheed Ullah Jan Ph.D. Research Scholar, Department of Economics, Gomal University Dera Ismail Khan, Pakistan, e‑mail: [email protected] Mahmood Shah Associate Professor, Department of Economics, Gomal University Dera Ismail Khan, Pakistan, e‑mail: [email protected] Abstract This research paper attempts to estimate the bilateral trade of Pakistan with SAARC countries using a gravity model of trade. This panel study covers the period from 2003 to 2016. The empirical results are obtained through pooled OLS, fixed‑effects, and random‑effects estimators. On the basis of Hausman test results, the paper con‑ centrates only on the findings of the fixed‑effects model. The empirical findings re‑ veal that the GDPs of both Pakistan and the partner country have a positive impact on bilateral trade. Market size has a negative impact on trade and this is justified on the basis of the absorption effect. Similarly, distance and exchange rate also have a negative correlation with bilateral trade. The study finds that Pakistan has very low trade with India and Afghanistan, despite the common border. A common language has a positive but insignificant impact on Pakistan’s bilateral trade. The Paper also attempts to calculate the trade potential of Pakistan. The findings reveal that Pakistan has high trade potential with all SAARC member countries except the Maldives and Afghanistan. Keywords: bilateral trade, common language, exchange rate, gravity model, population JEL: F14, F15, F31, F53 24 Waheed Ullah Jan, Mahmood Shah Introduction The exchange ofgoods for the purpose oftrade between two countries istermed bi‑ lateral trade. Inbilateral trade, the partner countries try toeliminate tariffs and oth‑ er trade barriers tofacilitate and encourage bilateral activities. Additionally, bilater‑ al trade agreements and their implications, mobility oflabor and increasing access toforeign markets are afocus inbilateral trade. The main objective ofbilateral trade isto achieve persistent economic growth and development along with poverty allevi‑ ation and jobs creation. Pakistan has established bilateral trade relations with anum‑ ber ofcountries around the world. The South Asian Association for Regional Cooperation (SAARC) was formed in1985. Atthe initial stage itwas aset ofseven nations (India, Pakistan, Bangladesh, Sri Lanka, Nepal, Bhutan, and the Maldives), but later on, Afghanistan joined this group. Now, SAARC comprises eight countries. Since the birth ofSAARC, the mem‑ ber countries have struggled towards regional cooperation and economic assimilation. InApril 1993, the SAARC countries signed atrade agreement called the South Asian Preferential Trading Agreement (SAPTA). This agreement was amilestone for the eco‑ nomic assimilation ofthe SAARC member countries. The agreement was made oper‑ ational inDecember 1995. Furthermore, in2004, SAARC member countries signed another agreement, called the South Asian Free Trade Area (SAFTA). This agreement was made applicable in2006. The main objective ofthis agreement was todeclare South Asia afree trade area bythe end of2016 (Hassan and Rehman 2015). Trade relations between Pakistan and South SAARC countries are not new phenom‑ ena. After independence, Pakistan established trade relations with neighboring coun‑ tries aswell asother countries ofthe region. These relations accelerated inthe 1990s when the global trade scenario was changing. Since then, Pakistan has been very keen toexpand its trade with countries inthe region and has signed various trade agree‑ ments with the regional and neighboring countries. The outcome ofthese agreements isthat tremendous enhancement has been seen inthe trade pattern ofthe region and asignificant increase inexports has been observed. Pakistan has been following ex‑ port‑based policies since 2000/01. Obviously, the achievements ofthese policies de‑ pend upon the access ofPakistan’s products tothe worldwide markets. Pakistan has made serious efforts for the improvement ofglobal trade, but still, its import and ex‑ port volume isnot remarkable with SAARC countries. Politics and the interference ofthe armed forces ingovernment policies are the key hurdles inthe way ofregional trade (Gul and Yasin 2011). SAARC countries have faced many developmental obstacles. Large fiscal deficits were observed inPakistan, India, and Sri Lanka. Similarly, ahigh degree ofcorrup‑ tion inBangladesh, the civil war inSri Lanka, macroeconomic volatility inNepal, the Maldives, and Bhutan, and the lack oftolerance and political hostility between thetwo neighboring countries ofIndia and Pakistan have significantly slowed down their eco‑ nomic development and regional collaboration. However, these challenges were suc‑ 25 A Gravity Model Approach towards Pakistan’s Bilateral Trade with SAARC Countries cessfully overcome. Sri Lanka brought liberalization totheir trade policies. India and Sri Lanka took initiatives toliberalize trade and deregulate interest rates, and the same path was followed byPakistan and Bangladesh. Despite these reforms, however, Paki‑ stan’s bilateral trade with the SAARC region isnot encouraging. Itwas about 8percent ofits overall trade in2010–11. This trading level isextremely disappointing compared tothe trade activities ofother regional groups like ASEAN. In2010–2011, Pakistan’s bilateral trade with India, Bangladesh, and Sri Lanka was 2.7 percent, 1.6percent and 0.61 percent respectively (Akram 2013). Pakistan and the other SAARC countries have realized the importance ofintrar‑ egional trade. Thus, they have adopted open trade policies toachieve positive conse‑ quences ofbilateral trade. The recent scenario ofinternational trade shows that bilateral orregional grouping trade onapreferential basis can play asignificant part inimports and exports ofgoods and services. However, SAARC countries are facing some dif‑ ficulties inthe regional integration process. First, there isashortage oftransparent policies for the betterment ofupcoming economic integration and social wellbeing. Second, some problems happen due totariff‑related constraints that limit awareness about economic integration. Finally, structural backwardness and economic deficien‑ cies inthe region aggravate the situation. The foremost objective ofthis research paper isto highlight the major determinants ofPakistan’s bilateral trade with the SAARC region. The other main objectives ofthe study are: 1) Toestimate the bilateral trade flow between Pakistan and SAARC coun‑ tries.2) Tofind the degree oftrade integration via aGravity Model with neighbor‑ ing countries.3) Todetermine the trade potential ofPakistan with the other SAARC countries. The rest ofthe paper isarranged asfollows. Section two contains relevant literature tothe study. The third section consists ofresearch methodology. The fourth section contains the empirical results and inthe last section, conclusions are drawn. Review of the literature This section consists ofempirical studies based onthe gravity model conducted bypre‑ vious researchers. Itprovides aroadmap for the application ofthe gravity model inbi‑ lateral trade for the current study. Kaur and Nanda (2010) examined the trade relations between India and SAARC countries. Inthis regard, they took exports ofIndia asthe dependent variable. All seven member countries ofSAARC were observed and Panel data were collected for the period 1981–2005. They ran the gravity model and estimated the results byap‑ plying three methods: random effects, fixed effects, and pool estimation. From the empirical results, they concluded that India has excellent trade opportunities with SAARC countries, especially Pakistan, Bhutan, and Nepal.Moreover, India’s exports can beextended toSAARC markets ifthey remove mutual barriers ontrade because 26 Waheed Ullah Jan, Mahmood Shah India borders four SAARC member countries. The geographical location ofIndia will favor their trade. Sherif and Fantazy (2013) analyzed the trade among the Gulf countries (Kuwait, Qatar, UAE, Oman, Bahrain and Saudi Arabia) fitting the Gravity Model tothe Pan‑ el data. The results ofthe gravity model assigned expected signs toall the variables ofthe gravity model. Economic size (GDP), market size (population), and GDP per capita have asignificant and positive impact onthe export pattern ofSaudi Arabia. Distance has anormal effect onSaudi Arabia’s exports, assuggested bygravity theo‑ ry. Its impact isnegative onSaudi Arabia’s exports. Hence, itis proved that the factors ofthe gravity model truly explain the volume ofbilateral trade. Shujaat (2015) examined Pakistan’s bilateral trade with 140 countries byusing the augmented gravity model and random effects methodology. Along with basic varia‑ bles, the researcher incorporated inflation rate, common language, free trade agree‑ ments, supply capability, and demand potential asindependent variables inthe gravi‑ ty model. The results exposed the fact that distance (transportation cost) has noeffect onPakistan’s bilateral trade. Similarly, Pakistan’s inflation rate and supply capabilities were also found inthe critical form. The basic variable (GDP) was found tobe areli‑ able factor inPakistan’s bilateral trade. Estimates ofthe gravity model suggested that free trade agreements are not infavor ofPakistan. Their impact isnegative onPaki‑ stan’s bilateral trade. Panda and Kumaran (2016) attempted toinvestigate the trade volume between China and India byapplying the gravity model. They estimated the results through arandom‑effects model (Panel regression model). Their results show that the trade level between the two countries will behigher than between countries that lie far away from each other. According totheir research, the trade volume between China and India will flourish because ofthe small distance between them. India’s bilateral trade ishighly predisposed byChina’s economic size (GDP) and language similarities, while this trade decomposed with low‑income countries. Wang (2016) researched eighty countries and used panel data for the period 2000– 2013. For analysis, the gravity model oftrade was fitted tothe data through the PPML technique. The results fully supported the assumptions ofthe gravity model and sug‑ gested that economic mass and bilateral trade have apositive relationship, while dis‑ tance creates negative shocks onmutual trade between two countries. Hussain (2017) examined the factors affecting Pakistan’s exports with the help ofthe gravity model. The researcher used panel data and selected the period from 1993–2013. For analytical estimation, the author used arelatively new technique, called the PPML estimator, which isconsistently used with agravity model totackle the problems ofpanel data. The empirical findings ofthe study confirmed the theoretical structure ofthe gravity model. Itwas established that distance has anegative impact onPakistan’s bilateral trade with its partners. Economic size was found tobe posi‑ tively related tobilateral trade when Pakistan’s GDP increases, its total trade with its trading partners will increase. 27 A Gravity Model Approach towards Pakistan’s Bilateral Trade with SAARC Countries Though avery rich literature isavailable onthe estimation ofbilateral trade through the gravity model, very little work has been done onPakistan’s bilateral trade with SAARC countries. Bearing inmind the above literature, our research paper can con‑ tribute inseveral ways tothe existing literature. First, weanalyzed Pakistan’s bilateral trade with all SAARC member countries. Previous researchers took into account only the major countries ofthe region and ignored small the countries. Second, weapplied three different techniques for our estimation. Most researchers used asingle estima‑ tion technique for analysis. Research Methodology The research methodology isastrategic plan through which wecan reach our speci‑ fied objectives. This plan isexplained step bystep inthis chapter. The universe of the study The study has selected the SAARC region toestimate Pakistan’s bilateral with the member countries. The area consists ofeight countries: Afghanistan, Sri Lanka, Bhu‑ tan, Nepal, Bangladesh, India, the Maldives, and Pakistan. This area was chosen be‑ cause the world largest market (India) lies next toPakistan, and they share along border. Similarly, Pakistan and Afghanistan share along border and have historical trade relations. Other countries inthe region also have close trade terms with Paki‑ stan. Though Nepal, the Maldives, and Bhutan are small economies inthe region, Pakistan has the chance toimprove trade relations and extend the range ofexports tothese countries. Data sources For our estimation, apanel data set isused ranging from 2003 to2016. All the data are taken onayearly basis. Countries’ individual and bilateral imports and exports data are taken from the International Trade Center (ITC), based onUN Comtrade statistics (2017). Data onmacroeconomics variables (population, RGDP, and exchange rate) are obtained from the World Development Indicators (WDI 2017). Data onthe distance between trade centers (normally capital cities) are taken and calculated from online Great Circle Distance. All the data are converted into US$ million. Data onthe pop‑ ulations for all countries are also presented inmillions. 28 Model specification and theoretical framework Under this heading, the general framework ofthe gravity model isdeveloped toana‑ lyze Pakistan’s bilateral trade. Itwill facilitate the researcher inexplaining the perfor‑ mance ofN cross‑section units (iand j= 1, 2, 3, …, N) inT years (t= 1, 2, 3…, T). The fundamental structure ofthe regression model isoutlined below: it it i it YXa be=+ + (3.1) Where Yit=dependent variable orregressand. i= cross‑section measurement for each individual country. t =time series measurement ofthe data. a =intercept (in‑ dicating countries’ fixed effects). i b = slope orcoefficient. Xit=independent varia‑ ble (showing variation for country iintime period t). it e = error term. The values ofthe fixed intercept, time variation and the coefficients orslope ( i b ) remain dif‑ ferent for each country. The insertion oftime trends and fixed‑effects inthe mod‑ el enable the researcher tofind out the role ofthe omitted variables inthe long run (Sakyi 2011). The gravity model oftrade has become more important inrecent years ininterna‑ tional trade. Itwas used for the first time byTinbergen (1962) and Pöyhönen (1963) for empirical analysis. Inaccordance with Newton’s law ofuniversal gravitation i.e., two items attract each other inproportion totheir masses and inversely proportion‑ al totheir distance, two countries will trade with each other according totheir GDP sizes and proximity. Krugman etal. (2012) are ofthe opinion that the gravity model isapplicable intwo‑sided trade because high‑income countries spend ahuge share oftheir income onimports and attract other countries topurchase goods from them because they have alarge variety ofgoods and have avast home market. So, the larger the econo‑ my, the larger the trade. Krugman etal. (2012) also mentioned other factors that cause bilateral trade, but these factors fail tooperate because ofdistance. Hence, when two countries are located far away from one another, their transportation cost will begin toincrease, and trade volume will decrease. Insuch acase, both countries will lose the gains from bilateral trade. The customized structure ofNewton’s law ofuniversal gravitation ispresented inthe following functional shape. 1 2 3 it jt ijt ij GDP GDP TT D bb b g æö ´÷ ç÷ ç =÷ ç÷ ç÷ ç èø (3.2) Where g = gravitational constant, ijt TT = volume oftotal bilateral trade (the summa‑ tion ofimports and exports), ij = respective countries, and t = time period. GDPs=eco‑ nomic sizes ofcountry iand country j. Dij=distance between two trading countries (normally between capital cities). βi= coefficients (β1, β2 and β1) tobe estimated. Waheed Ullah Jan, Mahmood Shah 29 A Gravity Model Approach towards Pakistan’s Bilateral Trade with SAARC Countries Taking the natural log ofboth sides ofequation (3.2), the gravity equation looks like: 12 3 ijt it jt ij ijt lnTT lnGDP lnGDP lnDab b b e=+ + - + (3.3) Where ln = natural logarithm, log ,ag= and ijt e  = the disturbance term orwhite‑noise error term After applying natural log tothe variables, the coefficients represent elasticities ofindependent variables inbilateral trade flows. From the previously reviewed literature, itcan beconcluded that there are many other elements that are responsible for bilateral trade flows, but they are not plotted inthe above equation. For the current study, the basic gravity model isenlarged with some other variables that hamper orpromote bilateral trade. The augmented gravity model developed for the current study isof the form: ( ) ( ) 12 345 6 ln * ln * ijt it JT it jt ijt ij ij ij ijt lnTT RGDP RGDP POP POP lnEXR lnDIST lnCBOR lnCLANG U ab b bbb b =+ + + ++ + + + (3.4) The same equation (3.4) can bewritten as: 1 23 45 6 ijt ijt ijt ijt ij ij ij ijt lnTT lnRGDP lnPOP lnEXR lnDIST lnCBOR lnCLANG U ab b b bb b =+ + + + ++ + + (3.5) where, TT ijt = total trade volume between country iand country j intime period t. RGDP ijt = the product ofReal Gross Domestic Product ofcountry iand country j inperiod t. POPijt= the product ofthe population ofcountry iand country j intime period t. EXRijt= bilateral exchange rate between country iand country j intime pe‑ riod t. DISTij= distance between country iand country j. The dummy variables ofthe study are: CBOR= Common Border (which takes the value of‘1’ ifthe border iscom‑ mon, ‘0’ otherwise) CLANG= Common Language (which takes the value of‘1’ ifthere isacommon language, ‘0’ otherwise) α= intercept, Uijt=omitted variables orunob‑ served factors that influence bilateral trade, and βi=coefficients (β1 , β2 ,…, β6 represent elasticities ofvariables). Analytical techniques The Analytical Techniques are the methods through which weenumerate the level ofbilateral trade between Pakistan and the SAARC countries. These techniques in‑ clude pooled OLS, fixed effects, and random effects. All these techniques are explained inthe coming sections. 30 Waheed Ullah Jan, Mahmood Shah Pooled Ordinary Least Squares (OLS) The easiest technique for panel data estimation isthe pooled ordinary least square (OLS) technique. Itignores the panel format (time and space dimensions) ofthe data and merely applies the typical OLS regression technique. The pooled OLS estimator can bewritten as: it it i it YXa bm=+ + (3.6) Where it Y =the dependent variable for country iintime period t. α=intercept. Xit =1 × K vector ofindependent variables for country iintime period t. βi= K ×1 vector ofcoefficients, and it m = the error ordisturbance term ofcountry iintime period t. This technique isbased onthe assumption that the intercept ( a ) and all the parameters (βi) are equal for all countries individually across time, and that 2 ~ (0, it iidms ) for all iand t. Itimplies that there isno autocorrelation, and the er‑ ror terms are homogenous for each individual country iduring time period t. Fixed Effects Estimator Itis awell‑known reality that every individual cross‑sectional component has sever‑ al unique properties. The intercept ofafixed‑effects equation varies among different individual units, but the same intercept remains constant (novariations) over time. While estimating the FEM, the disturbance term it m isdivided into two parts, one iscomponent‑specific and the other isthe time‑specific element. The summation ofin‑ tercept a and specific disturbance term it e constitutes: . it it m ae=+ Therefore the estimated fixed effect model (FEM) can bewritten as: it it i i it YXbae= ++ ( ) 2 ~ 0, it iides (3.7) The term i a = the fixed parameter, that isto beestimated. Dummy variables are in‑ corporated for every cross‑sectional component during the estimation process. This pro‑ cedure isknown asthe Least Squares Dummy Variables (LSDV) method. The estimation ofthe fixed effect model (estimators) helps toremove the endogeneity problem from the OLS regression model. Otherwise, itwill give spurious results (Roy &Rayhan 2011). Onthe other hand, the random‑effects model (REM) isbased onthe assumption that individual effects (heterogeneity) are separately divided between the disturbance term ( it e ) and the intercept (αi). Itmeans that the disturbance term isnot linked tothe explanatory variables. The REM can beshaped as: it it i i it YXb ma e= ++ + ( ) 2 ~ 0, it iides ( ) 2 ~ 0, iiidas (3.8) 37 A Gravity Model Approach towards Pakistan’s Bilateral Trade with SAARC Countries References Achakzai, J.K.(2006), Intra‑ECO Trade: APotential Region for Pakistan’s Future Trade, “The Pakistan Development Review”, 45(3), pp.425–437. Akram,A. (2013), Pak‑SAARC Intra‑industry Trade, PIDE Working Papers, 93. Buch, C.M., Kleinert,J., Touba, F.(2003), The Distance puzzle: Onthe interpretation ofthe distance coefficient inthe Gravity equations. Kiel Working Paper No.1159. Chan‑Hyun S.(2005), Does the gravity model explain South Korea’s trade flows?, “The Japanese Economic Review”, 56(4), pp.417–430. Cho,G., Sheldon,I., McCorriston, S.(2002), Exchange Rate Uncertainity and Agri‑ cultural Trade. “American Journal ofAgricultural Economics”, 84, pp.934–942. Eichengreen,B., Irwin, D.(1995), Trade blocs, currency blocs and the reorientation ofworld trade inthe 1930s, “Journal ofInternational Economics”, 38, pp. l–24. Gul,N., Yasin, M.(2011), The Trade Potential ofPakistan: AnApplication ofthe Grav‑ ity Model. “The Lahore Journal ofEconomics”, 16, pp.23–62. Hassan,R., Rehman, S.(2015), Economic Integration: AnAnalysis ofMajor SAARC Countries, “A Research Journal ofSouth Asian Studies”, 30, pp.95–105. Hussain, H.(2017), Globalization and Gravity Model ofTrade ofPakistan‑ APPML Es ‑ timator Analysis, “Management and Administrative Sciences Review”, 6, pp.15–27. Kandilov, I.T.(2008), The Effects ofExchange rate Volatility onAgricultural Trade, “American Journal ofAgricultural Economics”, 90, pp.1028–1043. Kaur,S., Nanda, P.(2010), India’s Export Potential toOther SAARC Countries: AGrav‑ ity Model Analysis, “Journal ofGlobal Economy”, 6(3), pp.167–184. Oguledo, V.I., MacPhee, C.R.(1994), Gravity Models: areformulation and anappli‑ cation todiscriminatory trade arrangements, Applied Economics, 26, pp.107–120. Panda,R., Sethi,M., Kumaran, M.(2016), AStudy ofBilateral Trade Flows ofChina and India, “Indian Journal ofScience and Technology”, pp.1–7. Poyhonen, P.(1963), ATentative Model for the Volume ofTrade between Countries, Weltwirtschaftliches Archiv, 90, pp.93–100. Rahman, M.(2005), The Determinants ofBangladesh’s Trade: Evidence from the Gen‑ eralized Gravity Model, Working Paper No.3, University ofSydney, School ofEco‑ nomics. Roy,M., Rayhan, I. (2011), Trade Flows ofBangladesh: AGravity Model Approach, “Economics Bulletin”, 31(1), pp.950–959. Sakyi, D.(2011), Economic Globalisation, Democracy and Income inSub‑Saharan Afri‑ ca: APanel Cointegration Analysis, Proceedings ofthe German Development Eco‑ nomics Conference, Berlin 2011, No.72. Sherif,S., Fantazy, K.(2013), Factors Influencing Export inBilateral Trade, “Interna‑ tional Journal ofManagement, Economics and Social Sciences”, 2, pp.12–27. Shujaat, A.(2015), Economic Survey ofPakistan, (2015–16) Islamabad, Ministry ofFi‑ nance, Government ofPakistan. Sokchea, K.(2006), Ananalysis ofCambodia’s Trade Flows: AGravity Model, Working Paper Series, 21464, pp.1–24. Tinbergen, J.(1962), Shaping the World Economy: Suggestions for anInternational Eco‑ nomic Policy. New York: Twentieth Century Fund, 1962. 38 Waheed Ullah Jan, Mahmood Shah Turkcan, K.(2005), Determinants ofIntra‑industry Trade inFinal goods and Interme‑ diate Goods between Turkey and Selected OECD Countries, Ekonometri veIstatis‑ tik Say, 1, pp.20–40. Wang, J.(2016), Analysis and Comparison ofthe Factors Influencing Worldwide Four Kinds ofVegetable Oil Trade: Based onGravity Model, “Modern Economy”, 7(2), pp.173–182. Streszczenie Zastosowanie modelu grawitacyjnego do oszacowania bilateralnej wymiany handlowej Pakistanu z krajami SAARC W artykule podjęto próbę oszacowania wielkości bilateralnej wymiany handlowej Pa‑ kistanu z krajami SAARC przy użyciu grawitacyjnego modelu handlu. Niniejsze ba‑ danie panelowe obejmuje okres od 2003 do 2016 r. Wyniki empiryczne uzyskano za pomocą metody najmniejszych kwadratów (pooled OLS), metody efektów stałych i estymatorów efektów losowych. Z uwagi na wyniki testu Hausmana w pracy skon‑ centrowano się wyłącznie na ustaleniach modelu efektów stałych. Badania empirycz‑ ne wskazują, że zarówno PKB Pakistanu, jak i państwa partnerskiego, mają pozytyw‑ ny wpływ na wielkość wymiany handlowej. Wielkość rynku ma negatywny wpływ na handel i jest to uzasadnione z uwagi na występowanie efektu absorpcji. Podobnie odległość i kurs wymiany są również ujemnie skorelowane z wielkością wymiany han‑ dlowej. Badanie wykazało, że pomimo wspólnej granicy wielkość wymiany handlowej Pakistanu z Indiami i Afganistanem jest bardzo niska. Wspólny język ma pozytywny, ale nieznaczny wpływ na wielkość wymiany handlowej Pakistanu. W artykule podję‑ to również próbę obliczenia potencjału handlowego Pakistanu. Wyniki tego badania wskazują, że Pakistan ma duży potencjał handlowy w relacjach ze wszystkimi krajami członkowskimi SAARC, z wyjątkiem Malediwów i Afganistanu. Słowa kluczowe: bilateralna wymiana handlowa, wspólny język, kurs walutowy, model grawitacyjny, liczba ludności