Journal of Economics, Finance and Business Analytics
2024; 2(1): 22 -33
https://quantresearchpublishing.com/index.php/jefba
ISSN:3006-0745 (Online)
Effect of SPS and TBT Non-tariff barriers and other
covariates on Trade among EAC member countries in the
pre and during COVID-19 pandemic period
Flossie Nkatha Kithinji and Tabitha Kiriti Nganga
Economics Department, University of Nairobi, Nairobi, Kenya
Suggested Citation to this article:
Kithinji, F. & Nganga, T. (2024). Effect of SPS and TBT Non-tariff barriers and other covariates on Trade among EAC member countries in
the pre and during COVID-19 pandemic period. Journal of Economics, Finance and Business Analytics, 2 (1), 22- 33
Received: MM DD, 2023; Accepted: MM DD, 2023; Published: MM DD, 2023
Abstract: The study sought to examine the effect of SPS and TBT Non-tariff barriers and other covariates on Trade among
EAC member countries in the pre and during COVID-19 pandemic period. The study adopted panel regression models based
on augmented gravity models. Secondary data was sourced from EAC states Central banks and the World Trade Organization
(WTO). In the pre-COVID-19 period, Sanitary and Phytosanitary (SPS) measures and technical barriers to trade (TBT)
inversely and positively impacted on trade value respectively. In the COVID-19 period, TBT had hindered intra EAC trade
while SPS policies directly and strongly EAC trade. The recommends that EAC member countries should identify alternative
mechanism to elimination of NTBs and adopt more integration policies rather than nationalistic policies.
Keywords: Non-Tariff Barriers, Sanitary and Phytosanitary Barriers, Technical Barriers, Economic Integration,
Trade, Trade volumes, Intra EAC trade.
1. Introduction
The treaty for the establishment of EAC, in articles sets forth goals of the trading block (EAC, 2007). One of the critical
goals was increase the depth and breadth of economic partnership among member nations. The common market protocols and
custom union were the vehicles through which economic integration aspirations of the region was achieved in early 2005 and
2010 respectively. The purpose of the common market protocols was to achieve open trade in the region via eliminations of
obstacles to trade (Eastern African Sub-Regional Support Initiative (EASSI), 2020). In the protocol, member states made a
commitment to each other to eliminate all tariffs in the intra region trade among themselves. The EAC Common Market
enables free trade, movement of people, labour services and capital and accords right of residence and establishment (EAC,
2009). Even though the integration of the region has improved intra region trade, the volumes and value of trade is still dismal
when compared to global trade with the region (East African Business Council (EABC), 2018). In 2016, the level of Intra EAC
exports was 5.9% of EAC trade with the rest of the world (EABC, 2016), while in 2018, intra EAC exports was 22.4% of the
regions total export trade (EABC 2018). Trade among member has been negatively impacted by Non-Tariff Barriers (NTBs)
that are still in place among EAC members states (EABC, 2020). NTBs account for most of the blockages to intra-regional
trade (Oiro, Owino & Mendez-Parra, 2017). NTBs are broad policy intervention beyond tariffs the border that interferes with
free trade and movement of factors of production among countries (Kinzius, Sandkamp & Yalcin, 2019). Among the kinds of
NTBs, technical barriers to trade (TBT) and sanitary and phytosanitary (SPS) policies are key. (EABC, 2020). The TBT seeks
put in place regulations that are technical in nature, ensure conformity to assessment processes are not discriminatory and
acting as unwarranted blockage to trade. The SPS recognizes the right of WTO member to put in place mechanisms to ensure
environmental and human health protection from harmful products (Calabrese & Eberhard-Ruiz, 2016).
Even though, NTBs have been hampering intra EAC countries trade, the outbreak of Coronavirus of 2019 (COVID-19),
made the trade situation worse for certain aspects of trade. The virus thereafter spread to over 188 nations with over 153
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 23
million and 3.2 million people being infected and succumbing to it respectively (World Health Organization (WHO), 2021).
The WHO later on declared it a pandemic on 11th March 2020 (WHO, 2020). The first case of infection with COVID-19 was
reported in EAC on 13th March, 2020 in Kenya with other countries following after. The outbreak necessitated response
strategies and programs that oscillated between partial to total lockdowns (East African Community, 2020).
The East African states began implementing COVID-19 related NTBs such as closing airports to passenger planes,
restricting movement across borders for people and goods to only essential products and services. Further, within the countries,
public gathering was stopped completely or restricted to a given predetermined number. There was also quarantine in hot spot
areas within the states as well as curfews (EAC, 2020). However, while the other EAC member nation adopted most the
measures, Tanzania adopted a different approach that led to strained relationship with other EAC member states and
international community. Tanzania stopped the reporting of numbers of COVID-19 infection and death cases. The strained
relationship resulted to complete closure of borders between Kenya and Tanzania on May 16th, 2020 (Ogunleye et al., 2020).
Globally, NTB’s hampers the enjoyment or potentials of free trade as they hinder the free movement of people, goods and
factors of production across national boundaries. Free trade is beneficial to nations in terms of market for products, factors of
production and financial assets (Rojas & Pineda, 2020). In EAC member countries, NTBs act as obstructions towards smooth
exchange of goods and service among countries even with the existence of custom union (Hangi, 2010). Several NTBs exist in
EAC, which includes, unstandardized weight bridges, TBT, SPS, road blocks among others (Okumu, 2010). The Covid
situation just happened to aggravate issues concerning impact of NTBs on intra EAC trade (Albertoni & Wise, 2021). The
control measures for the disease made the situation of NTBs worse with some NTBs that had initially been eliminated being
reintroduced and new controls being put in place hence collapse of informal cross-border trade (East African Sub-Regional
Support Initiative, 2020). Most of the empirical literature were carried before the COVID 19 pandemic (Baya, 2019; Ghodsi,
Grübler, Reiter, Stehrer, 2017). Additionally, research on impact of COVID 19 on trade have tended to ignore the marginal
impact of SPS and TBT NTBs on intra EAC countries trade in the COVID-19 period (UNECA, 2020, 2021). This research
sought to examine the difference in the effect of SPS and TBT Non-tariff on Trade within EAC member states pre and during
COVID 19 pandemic.
Hence the general objective of this study was to investigate the effect of SPS and TBT Non-tariff barriers and other
covariates on Trade among EAC member countries in the pre and during COVID-19 pandemic period.
2. Literature Review
2.1 Theoretical Review
This research on the nexus between NTBs and Trade is supported by the Heckscher-Ohlin (H-O) theory. The theory was
postulated by Hecksher and Ohlin in 1920s to examine how nations can benefit from trade by utilizing abundant factors. The
H-O expanded on the work of Ricardo by introducing capital as second factor of production for finished goods in addition to
labor (Ohlin, 1933). The theory further assumes that their exist differences in factors proportions across industries and
endowment across nations (Gandolfo, 2014). Capital intensive firms have low labour to capital ratio while labour intensive
firms have high labour to. Further, developing counties are often endowed with labor relative to physical capital while
developing countries often endowed with labor relative to physical capital stock.
Capital to labor endowment describes the comparative factor abundancy between countries and that countries enjoy
comparative advantage when they produce commodities in line with their relative factor abundance. Further, the theory
assumes that differences among trading counties is explained by relative factor endowment and that trade will be advantageous
and will occur among countries when the countries have differing comparative factor endowments and different firms and
industries adopt different capital labour ratios (Feenstra and Taylor, 2017). Further, countries tend to benefit from trade when
they extract, manufacture and sell products that are made available from relatively abundant factors of production (Suranovic,
2012; Obuya & Olweny, 2017). Even though H-O theorem explains reasons why countries trade, it has some gaps regarding its
predictive power given that trade among countries cannot only be explained by relative factor endowment. Further, empirical
literature has shown technology endowment may outweigh labor endowment many times (Gandolfo, 2014). The H-O theorem
has posited that if a country cannot benefit from trade where they produce and export commodities in line with their abundant
factor due to distortions of factor prices by other countries, then trade protection through NTBs and other measures is justified.
Further, countries may impose restrictions to free trade to protect infant industries especially in line of production where they
possess comparative advantage. The protection is thus justified to give the infant industries opportunity to grow and compete in
the global arena.
2.2 Empirical Review
The nexus between NTBs and trade globally, regional and locally exist. Kinzius, Sandkamp and Yalcin (2019) investigated
the effect of NTBs on imports. The study adopted gravity model and data was extracted from Global Trade Alert database. The
24 Kithinji & Nganga: Effect of SPS and TBT Non-tariff barriers on Trade in EAC member countries
research showed that NTBs hampers exports. The study further revealed different NTBs impact on trade to a different
magnitude. Kinzius, Sandkamp and Yalcin (2019) was restricted to measures restricting imports into the implementing country
only. Hence, there is need for another study that specifically focuses on SPS and TBT Non-tariff barriers and how they
influence trade both export and import trade. Further there is need to examine the effect of NTBs in the COVID-19 period.
Ghodsi (2020) examined the effect TBTs reported China on its imports for manufacturing from 2002–2015. The study used
gravity model that controls for sample selection bias, multilateral resistance, endogeneity bias and heterogeneity of exporting
firms. The study establishes major effect of TBT on imports. The impact was however differentiated across exporting countries
implying that some exporting countries were affected more with the TBT being selectively prohibitive. Ghodsi (2020) was
carried out in China with different economic environment.”
Vickers et al. (2020) evaluated whether the COVID-19 impacted on trade in products of food in nature among countries in
Commonwealth. The study revealed programs to contain COVID-19 spread abruptly disrupted international merchandise trade.
The study revealed that national and regional lockdowns resulted in the interruption of trade in foods products. Vickers et al.
(2020) was limited to food products in Commonwealth countries and was limited to few data points at the onset of COVID-19
pandemic, there is therefore need for another study that has more data points and carried specifically within EAC focusing on
all export and import types. Okute (2017) investigated the impact of NTBs on volume of trade among exporters from Kenya to
EAC region. The research was based on explorative research type with the study targeting nine thousand five hundred and
eighty-five exporters from Kenya of which one hundred and twenty-one were sampled. The study administered questionnaires
to collect primary data that was further analyzed. The findings revealed that various NTBs to hampered trade in EAC region.
Okute (2017) was based on descriptive analysis hence could not establish causal effect relationship between NTBs and trade.
Further, the study was carried out before COVID-19 hence does not capture effect of associated shocks on trade.
Khaliqi, Rifin, and Adhi (2018) evaluated whether SPS and TBT impacted on Indonesian export of shrimp. The gravity
model was used with results revealing that the GDP of exporting nations as well as forex rate inversely effected shrimp exports.
Additionally, the trade cost and GDP of the importers have direct effect exports. Further, the study revealed that SPS and TBT
policy do impact exports of shrimp from Indonesia to the world. Akin-Olagunju and Yusuf and Okoruwa (2018) examined the
influence of the SPS standards on coco exporters’ competitiveness. The study revealed that increasingly stringent SPS global
standards enhanced trade. UNECA (2020) examined whether COVID-19 had adverse social and economic impacts in Eastern
Africa. Secondary data was extracted from UNCTAD, CB of Kenya, Rwanda, Uganda Tanzania, Burundi, International
Monetary Fund (IMF) among other institutions. Trend and structural break analysis showed extensive impacts of the pandemic
on the region. The study revealed that formal merchandise trade (imports and exports) values have recovered sharply from the
declines in the second quarter of 2020. Likewise, services trade performance has remained dismal for countries in the region,
mainly due to the sharp drop in tourism. UNECA (2020) was based on data points at the beginning of pandemic and another
study needs to be done that has more data points. Fernandes, Lefebvre and Rocha (2018) evaluated whether SPS and TBT
polices via PTA impact on exports of from Columbia, Chile and Peru. The study adopted data extracted from World Bank Deep
from 1996 to 2015. The study adopted gravity equation. The study revealed that existence of PTA agreements increased firm’s
exports in destination markets. The study also showed that SPS and TBT were major for smaller firms.
Santeramo (2020) evaluated how trade in agricultural products was impacted by SPS measures. The study specifically
compared the impact of SPS on volume of trade for members and nonmembers of regional trade agreements (RTA). The study
findings revealed that befits received from being a signatory to RTAs declined with implementation of nondiscriminatory SPS
measures. Further, both SPS and RTAs catalyze trade. RTAs offer opportunity for members to renegotiate SPS members thus
improving trade among members of RTA. Santeramo (2020) was not based in EAC countries hence a study in EAC states is
necessary. Ghodsi (2021) examined the effect of different TBTs on imports of ICT goods globally. The goods were evaluated
based on HS from 1996-2018. Pseudo Poisson Maximum Likelihood (PPML) technique was adopted in parameter estimation.
The findings showed that TBTs in general have a major direct impact on the value of imports. However, a few TBTs acted as
barriers to trade hence impacted on value and volume of trade negatively. Ghodsi (2021) study was not localized in EAC
member states and moreover it is focused on ICT goods only. Escaith and Khorana (2021) examined whether COVID-19
pandemic impacted trade in Commonwealth member states. The research utilized secondary data to estimate loses in trade
arising from COVID-19 pandemic. The study was a simulation of possible future impacts of COVID-19 on flow of trade based
on 3 conditions including a consensus, pessimism and optimism. The results showed that trade is inversely impacted with
advanced nations being affected more and the severity of effect is based on magnitude and duration the pandemic will last.
Escaith and Khorana (2021) was on the commonwealth countries hence there is need for another study that focuses on EAC
countries. Further, the study did not examine the indirect effect of COVID-19 on trade through NTBs. Erokhin and Gao (2020)
examined the interactions between food trade, food prices and COVID-19 cases. The study happened among forty-five
countries with autoregressive method, causality test, and variance decomposition being adopted. COVID-19 impacts were
greater among developed countries compared to developing countries with regards to food inflation. The less developed
countries were affected more in terms of food security risk. Erokhin and Gao (2020) was based on food trade hence there is
need for another study that extends the breath of the study in all merchandise traded in the context of EAC states.
Ali, Fugazza and Vickers (2020) examined the nexus between exports in primary commodities and the COVID-19 in China,
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 25
UK, USA, EU and Australia. The study utilized export scenarios and historical trends. The research showed that supply and
demand shocks emanating from COVID-19 was hurting trade. Ali, Fugazza and Vickers (2020) was based tend analysis where
the standard deviation of import values was exclusively associated with COVID-19 shocks. Kahenu (2014) examined the
impact on NTBs on Kenyan exports to EAC. The study collected primary data from respondents and secondary data from
reported NTBs between the periods 2007-2013. The results revealed that the fall in exports from Kenya between 2012 and
2013 was explained by NTBs introduced by trading partners. Anyona (2018) was based on exploratory design hence could not
establish causation effect between trade and NTBs. Ghodsi et al. (2017) examined whether varied kinds of NTBs impacted on
international from 1995 and 2014. Data was sourced from WTO. The study used gravity model with the findings showing that
NTBs impeded trade.
3. Methodology
3.1 Theoretical Framework
Analytical research design was adopted to examine the causation between NTBs and trade. The study adopted panel
regression models in examining the nexus between NTBs and trade in the pre and during COVID-19 pandemic period.
(Gujarati & Porter, 2003). The study adopted gravity model to quantify EAC member countries trade in the pre and during
COVID-19 pandemic. The simplest gravity model is given in equation [1].
Tij = A * )………………………………………….……….[Equation 1]
Where Tij is the export and import of country i &j, Yi * Yj is the product of GDP of country i and i and dij is the geographic
distance between trading pair capital cities. In accordance to the model, ceteris paribus, volume of trade between two countries
is direct function of the product of the GDPs of the two countries trading with each other. Further, the volume of trade between
trading pairs is an inverse function of the distance between the trading pairs capital cities. Therefore, higher GDPs product and
shorter distance should lead to high trade volume between the two trading countries and vice versa (Krugman & Obstfeld,
2006). Dummy variables are often included to capture the effect of the qualitative variables such as sharing of national
language, sharing of a border and trade agreements (Stay & Kulkarni, 2016). The effect of SPS and TBT NTB can be added to
the model. The effect of COVID-19 pandemic shocks being qualitative factors affecting trade were added in the augmented
model as dummy variables.
3.2 Empirical Framework/Model specification
Even though there exist different variants of gravity model applied in various empirical studies, the current study adopted
model suggested by Krugman and Maurice (2005) as shown in equation [1]. To model the effect of SPS and TBT Non-tariff
barriers and other covariates on Trade among EAC member countries in the pre and during COVID-19 pandemic period, the
model in equation [1] was augmented by adding SPS and TBT NTBs. The resulting gravity model in the form presented in
equation [2] was used to estimate the effect of SPS and TBT NTBs and other covariates on trade among EAC member
countries in the pre COVID-19 pandemic period.
ln (Tijt) = α0 +α1ln (GDPit * GDPjt) +α2 ln(SPSit*SPSjt) + α3 ln(TBTit*TBTjt)+ α4ln(PoPit *PoPjt) + α5Exrijt +α 6lnDij+α7La
+α8 Bo+ ɛijt.............................................................................[Equation 2]
Where:
Tijt : Value of Trade
GDPit and GDPit: Gross Domestic Product of trading pairs respectively.
PoPit and PoPjt : Population of trading pairs respectively.
Exrijt: Exchange rate between trading pairs.
Dij: Distance in kilometers between capital cities of trading pairs.
La: common national official language.
Bo: Sharing of border dummy variable.
αi = is parameter estimates where i is 1,2......,7
ɛijt: Error term, ln= Natural Logarithm
i and j are reporting country and trading partner of reporting country respectively.
26 Kithinji & Nganga: Effect of SPS and TBT Non-tariff barriers on Trade in EAC member countries
NTBit: Non-tariff barriers imposed by country i in time t
NTBjt: Non-tariff barriers imposed by country j in time t
t=2014, 2015,…, 2019.
The model in equation [equation 2] was modified further by adding dummy variable COV to capture COVID-19 shocks.
The study further adopted monthly data from April 2019 to March 2021. This translates to 12 months before COVID-19 struck
and 12 months after COVID-19 struck. The resulting gravity model in the form presented in equation [4] was used to estimate
nexus between SPS & TBT NTBs and other covariates on Trade among EAC member countries in the COVID-19 pandemic
period.
ln (Tijt) = α0 +α1ln (GDPit *GDPjt) +α2 ln(SPSit *SPSjt) + α3 ln(TBTit *TBTjt) α4ln(PoPit *PoPjt) + α5Exrijt +α 6lnDij+α7La
+α8 Bo+ α9COV+ ɛijt..............................................................[Equation 3]
Where:
COV: COVID-19 dummy variable
t= 2019April, 2019 May… 2021 March.
αi = is parameter estimates where i is 1,2......,8
3.3 Variables Definition and Measurement
The Table 1 has the variable notation, measurement and the expected sign of the nexus between the regressand and
regressors.
Table 1: Operationalization of Study Variables
Variable
Notation
Measurement
Expected sign
Dependent
Trade
T
Trade value between reporting and trading partner in USD
Explanatory
GDP_1
GDPi
GDP of reporting country in USD
+
GDP_2
GDPj
GDP of trading partner in USD
+
SPS_1
SPSi
The number of SPS notified as imposed by reporting country
-
SPS_2
SPSj
The number of SPS notified as imposed by trading partner.
-
TBT_1
TBTi
The number of TBT notified as imposed by reporting country
-
TBT-2
TBTj
The cumulative number of TBT measures notified as imposed by trading partner.
-
PoP_1
PoPi
Population of reporting country in numbers
+
PoP_2
PoPj
Population of trading partner state in numbers
+
Exchange rate Exchange rate
Exrij
Monetary Currency exchange rate between reporting country and trading partner
+/-
Distance
Dij
Distance between capital cities of trading pairs
_
Language
La
Official national language sharing where 1= Yes 0=No
+
Boundary
Bo
Sharing national boundary where 1= Yes 0=No
+
COVID-19
COV
COVID-19 shocks where 1= Pre-covid months 0= covid months.
-
Table 1 presents the variables definition and measurement. The expected relationship between SPS and TBT Non-tariff
barriers and trade value was negative. The negative relationship results from trade inhibition caused by NTBs. Kinzius,
Sandkamp and Yalcin (2019) revealed that NTBs inversely impact on imports. The expected relationship between economic
growth and value of trade was positive given that increasing GDP means increasing National Income (NI) as well as
purchasing power for goods and services. The expected relationship between population size and value of trade was positive.
The size of the population represents the demand for products and increasing population implies increased demand for
products hence more trade value. Krugman and Obstfeld (2006) version of Gravity model of international trade showed that
population of countries as additional mass for bilateral trade just like GDP of the trading countries. The expected relationship
between exchange rate and value of trade is not defined hence it could be negative or positive. If the change in exchange rate
leads to a net reduction in value of trade between trading pairs, then then nexus between forex rate and value of trade becomes
negative and vice versa (Iscan, 2016). The expected relationship between distance and value of trade was negative. The
increasing distance between capital cities of reporting country and trading partner implies increasing cost of transportation that
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 27
inhibits trade (Krugman & Obstfeld, 2006). Countries that have a common national official language tends to trade more given
that the traders between the two countries can easily understand each other while trading. Stay and Kulkarni (2015) revealed a
positive effect of sharing a common national official language and trade between the UK and its pairs in trade. The expected
relationship between COVID-19 shocks and value of trade was negative. COVID-19 pandemic led to restricted movement
across borders, reduced industrial output hence reduced value of trade. UNECA (2020) revealed that formal merchandise trade
(imports and exports) values declined sharply in the second quarter of 2020 when COVID-19 struck EAC countries.
3.4 Diagnostic Testing
The assumptions tested included homoscedasticity, collinearity, autocorrelation and cross-sectional dependence. In
econometrics, model is said to be homoscedastic when error terms have constant and finite variance however when the
variance is not constant, then the model is said to be heteroskedastic (Gujarati, 2008). Modified Wald test was utilised to
examine the existence of group heteroscedasticity. The research would conclude presence of homoscedasticity if p-value
generated was greater than the 0.05. In the presence of heteroscedasticity alone, the study can use robust standard errors.
However, in the presence of heteroskedasticity and serial correlation, Panel correlated standard errors (PCSEs) or Feasible
Generalized least squares (FGLS) should be adopted. The regression is based on the assumption of no multicollinearity.
Multicollinearity exist in the estimation model when regressors are highly correlated among themselves (Goldberger, 1964).
The presence of multicollinearity results to inflated coefficients. The pairwise Pearson correlation coefficient was adopted
where multicollinearity is said not to exist when bivariate correlation coefficient is above 0.8 (Gujarati, 2003). Linear
regression model is based on nonexistence of serial correlation. Serial exist in a model errors terms in the present time is highly
correlated with its previous period values (Gujarati (2003). The regressand in the present period is explained by its own values
in the previous period as well as regressors both in the present and previous period. Serial correlation is common for long time
series of 10 years and above. Wooldridge Drukker test was adopted to examine the presence of serial correlation of error terms
such that p-value greater than 0.05 implies absence of autocorrelation (Wooldridge, 2012). In the presence of serial correlation
only, the study would adopt clustered standard errors, however, if both serial correlation and group heteroskedasticity exist in
the model, the study would adopt either PCSEs or FGLS model for parameter estimation. Heterogeneity is a characteristic of
panel data where there are significant differences across entities to be studied which in this case are the countries within EAC.
Heterogeneity exists when something within the individual entities affects the explanatory or dependent variables and therefore
needs to be controlled (Wooldridge, 2012). Fixed effects (FE) models can be used to control unobserved time-invariant
variables that might affect regressand and regressors to able study examine net effect of regressors of concern variables on the
regressand. In cases where the unobserved time invariant characteristic is non-correlated with regressors of concern, then, fixed
effect model adoption to isolate the effect of unobserved time invariant characteristic results in inefficient estimators
(Wooldridge, 2012). FE model is thus not suitable and there is need for the relationship to be modelled differently using other
models like random-effects (RE) model. The research employed Hausman test to choose between FE and RE models. If p-
value > greater than .05, RE ought to be adopted.
3.5 Estimation Technique
Based on the result of Hausman test, the study would decide whether to adopt fixed effect or Random effect model. In cases
of violation of classical least squares assumption such that panel data exhibit serial correlation, heteroscedasticity and cross-
sectional dependence, then the research would choose FGLS model.
3.6 Sources of Data
Data was sourced from UNCTAD, Central Bank of Kenya (CBK), National Bank of Rwanda (BNR), Bank of Uganda
(BoU), Bank of Tanzania (BoT), Banque de la République du Burundi (BrB), IMF, WTO, East African Community Secretariat
(EACS) among other institutions. South Sudan which joined EAC recently was omitted from the study given that it has no
complete data for the study period. Data for estimating the impact of TBPE of identified NTBs on intra EAC trade was
collected for a period of 20 years beginning 2000-2020. Data on SPS and TBT NTBs on trade among EAC member countries
in the pre COVID-19 pandemic period was sourced for seven years from 2014-2020 because of scarcity of secondary data on
SPS and TBT NTBs. WTO provided secondary data on SPS and TBT NTBs from 2014-2020. Data for estimating the effect of
SPS and TBT NTBs on Trade among EAC member countries during COVID-19 pandemic period was collected monthly for a
period of 24 months beginning April 2019 to March 2021. The first 12 months was period before COVID-19 struck and the
next 12 months was period after COVID-19 was first reported in Kenya (The first among EAC countries).
4. Results and Discussion
The study also evaluated whether SPS and TBT NTBs and other covariates impacted on intra EAC trade between member
28 Kithinji & Nganga: Effect of SPS and TBT Non-tariff barriers on Trade in EAC member countries
countries in the pre and during COVID-19 challenge. The effect was examined in two levels. The first level involved the
impact of SPS & TBT NTBs and other covariates EAC trade in the pre COVID-19 pandemic period based on annual data. The
second level involved examining impact of SPS & TBT NTBs and other covariates on intra EAC trade among member
countries during COVID-19 pandemic period based on monthly data.
4.1 Effect of SPS and TBT Non-tariff barriers and other covariates on Trade among EAC member countries in the pre
COVID-19 pandemic period.
The study first examined the impact of SPS & TBT NTBs and other covariates on EAC member countries trade in the pre
COVID-19 period. The examination of the objective was based on annual panel data between 2014 and 2019.
4.1.1 Diagnostic Tests
The cross-sectional dependence (contemporaneous correlation) was not tested given that the data had time series component
of only six years. The Wooldridge Drukker test for serial correlation revealed the presence of autocorrelation (p=.0006 < 0.05).
Further, Modified Wald showed that there was presence of group heteroscedasticity (p-value = .000< 0.05). Given that the
panel data had the twin problem of serial corelation and group heteroscedasticity without the problem of contemporaneous
correlation, the study could adopt either FEM or REM with clustered robust standard errors to eliminate the problem of
heteroscedasticity and serial correlation. The FEM was further eliminated given that the observed variables including
boundary, language and distance were omitted by the model because they were constant within the group. The study thus
utilised REM with clustered S.E.
4.1.2 Random Effect Model
The study adopted REM with clustered S.E to eliminate the problem of heteroscedasticity and serial correlation. The
findings based on REM with clustered robust standard errors were shown in Table [2].
Table 2: Random Effect Model with Clustered Robust Standard Errors
lnT
St.Err.
t-value
p-value
[95% Conf
Interval]
Sig
lnGDP1GDP2
0.186
9.43
0.000
1.391
2.121
***
lnExr
0.132
1.66
0.098
-0.040
0.476
*
D
0.001
-4.63
0.000
-0.006
-0.002
***
La
0.537
0.31
0.760
-0.888
1.216
Bo
0.557
-1.42
0.156
-1.881
0.302
lnSPS1SPS2
0.030
-2.33
0.020
-0.128
-0.011
**
lnTBT1TBT2
0.051
0.33
0.741
-0.083
0.117
Constant
7.818
-7.45
0.000
-73.582
-42.937
***
Mean dependent var
21.463
SD dependent var
2.882
Overall r-squared
0.869
Number of obs
60.000
Chi-square
1452.331
Prob > chi2
0.000
R-squared within
0.002
R-squared Overall
0.8685
*** p<0.01, ** p<0.05, * p<0.1
Note: product of Sanitary and Phytosanitary of reporting country and its trading partner (SPS1SPS2), product of Technical Barrier to trade of reporting
country and its trading partner (TBT1TBT2).
In Table 2, the coefficient of determination (R2) revealed that the model explained 86.85% of the total variation in trade
within EAC region. The remaining variation of 13.15% is explained by unobserved (time variant and time invariant) variables.
The study thus concluded that the model had satisfactory goodness of fit. Further, the F-test showed that overall, the
explanatory variables including product of SPS notified by reporting country and its trading partners (SPS1*SPS2), product of
TBT to trade notified by reporting country and its trading partner (TBT1*TBT2), distance, national language sharing (La),
border sharing (Bo), Exchange rate (Exr), Product of GDP of trading pairs (GDP1*GDP2) had significant effects on intra EAC
trade. This is evidenced by .05 level of significance higher than overall p-value.
Further, the effect of the product of SPS notified by reporting country and its trading partners strongly and inversely
impacted on value of trade. A 1% improvement in the number of notified SPS by reporting country and its trading partner leads
to reduced value of trade by 6.9%. This finding implies that SPS measures act as inhibitors to intra EAC trade by reducing the
trade values between reporting EAC nation and its trading partners within EAC. The finding agrees with Kinzius, Sandkamp
and Yalcin (2019) who revealed that the NTBs inversely impacted on imports. Chen (2013 established that TBT increased trade
volumes and duration of trade relationships. Further, Akin-Olagunju, Yusuf and Okoruwa (2018) revealed that SPS and TBT
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 29
policy do not affect exports of shrimp from Indonesia to the world.
Regarding the other covariates, the product of GDP of the trading pair countries and Distance had significant effects on the
value of intra EAC trade. A one percent increase in distance between trading pairs capital cities within EAC led to reduced
trade values. Distance was thus an inhibitor to trade given that longer distance between economic centres of countries implies
increased transportation cost that makes the product more expensive in the market. Mortazavi and Do (2006) revealed an
inverse effect of distance on value of trade between counties in EU. The effect of the product of the GDP of trading pairs
strongly impacted on trade within EAC region. A 1% improvement in the product of the GDP of the trading pairs leads to
increased trade value by 175.6%. Larger GDP means that the countries have more purchasing power in terms of national
income hence their demand for products also increases. Finally, the effects of boundary sharing, national language and
exchange rate were not significant in explaining trade value among EAC member countries. Krugman and Obstfeld (2006)
study showed that value of trade is directly proportional to the product of the GDPs of the reporting country and its trading
pairs.
4.2 Effect of SPS and TBT Non-tariff barriers and other covariates on Trade among EAC member countries during
COVID-19 pandemic period
The study also examined the effect of SPS & TBT NTBs and other covariates EAC member trade during COVID-19
pandemic period. The study adopted monthly secondary data from April 2019 to March 2021 which was 12 months before and
during COVID-19. The study relied on 8 EAC trading pairs of countries after two pairs (Kenya- Rwanda pair and Kenya-
Tanzania pair) were omitted because they lacked sufficient monthly bilateral trade figures. The study carried out test of
robustness of the estimation model and chose the appropriate panel data model for estimation.
4.2.1 Descriptive Analysis
Trend analysis of bilateral trade between some EAC countries. The trend is based on monthly data running from April 2019
to March 2021. The trends are presented in Figures [1-3].
Figure 1: Bilateral Trade for Kenya-Burundi and Kenya-Uganda
The bilateral trade between Kenya and Burundi reveals that export from Kenya to Burundi fell greatly between April 2020
and May 2020. Further, imports from Burundi to Kenya also fell greatly between April 2020 to June 2020 coinciding with the
first reported case of COVID-19 in Kenya after which various restrictions were put in place. However normal trade between
Kenya and Burundi recovered from July 2020 onwards. The bilateral trade between Kenya and Uganda reveals that exports
from Kenya to Uganda fell greatly between April 2020 with slow recovery happening between May 2020 and July 2020.
Further, imports from Uganda to Kenya also fell greatly between March 2020 and May 2020 coinciding with the first reported
case of COVID-19 in Kenya after which various restrictions were put in place. However normal imports from Uganda to
Kenya recovered from June 2020 onwards with full recovery being realized in August 2020 [Figure 1].
30 Kithinji & Nganga: Effect of SPS and TBT Non-tariff barriers on Trade in EAC member countries
Figure 2: Bilateral Trade for Uganda-Tanzania and Uganda-Rwanda
The bilateral trade between Uganda and Tanzania reveals that exports from Uganda to Tanzania fell greatly between March
2020 and April 2020. Further, imports from Tanzania to Uganda also fell greatly in April 2020 coinciding with the first
reported case of COVID-19 in Uganda after which various restrictions were put in place. However normal imports from
Tanzania to Uganda recovered from May 2020 onwards. The bilateral trade between Uganda and Rwanda reveals that exports
from Uganda to Rwanda and imports from Rwanda to Uganda fell greatly in April 2020 and May 2020 due to COVID-19 with
slow recovery thereafter [Figure 2] .
Figure 3: Bilateral Trade for Uganda-Burundi and Tanzania -Burundi
The bilateral trade between Uganda and Burundi reveals that exports from Uganda to Burundi fell greatly in April 2020.
Further, imports from Burundi to Uganda also fell greatly to zero figures between March 2020 and July 2020 coinciding with
the first reported cases of COVID-19 and associated restrictions. However, Exports from Uganda to Burundi recovered from
May 2020 onwards. The imports from Burundi to Uganda is a little intermittent with some monthly reporting zero figures. The
bilateral trade between Tanzania and Burundi reveals that exports from Tanzania to Burundi did not show much of the effect of
COVID-19 with exports figures remaining relatively same between December 2019 and July 2021. This could be explained by
low COVID-19 restrictions in the two nations [Figure 3].
4.2.2 Diagnostic tests
The Wooldridge Drukker test revealed that there was no autocorrelation (p=.904 > 0.05). The study further showed the
presence of group heteroscedasticity based on Modified Wald test (the p = .000< .05. Finally, Breusch -Pagan/LM test
concluded presence of cross-sectional dependence (p-value was less than 0.5. The research thus concluded that estimation
model suffered from contemporaneous correlation.
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 31
4.2.3 Feasible Generalized Least Squares Regression
Given that the model suffered from contemporaneous correlation and group heteroscedasticity, the study adopted the FGLS
model that corrects for contemporaneous correlation and smaller standard errors introduced by heteroscedasticity [Table 3].
Table 3: Feasible Generalized Least Squares Regression
lnT
Coef.
St.Err.
t-value
p-value
[95% Conf
Interval]
Sig
D
-0.003
0.000
-10.78
0.000
-0.004
-0.003
***
La
-1.713
0.403
-4.25
0.000
-2.503
-0.923
***
Bo
4.141
0.870
4.76
0.000
2.435
5.847
***
COV
-0.413
0.093
-4.43
0.000
-0.595
-0.230
***
lnSPS1SPS2
1.603
0.201
7.97
0.000
1.209
1.998
***
lnTBT1TBT2
-4.404
0.348
-12.65
0.000
-5.087
-3.722
***
lnGDP1GDP2
2.910
0.281
10.34
0.000
2.359
3.462
***
lnExr
-0.017
0.077
-0.22
0.823
-0.168
0.134
Constant
-91.034
14.371
-6.34
0.000
-119.200
-62.869
***
Mean dependent var
1.727
SD dependent var
3.988
Number of obs
192.000
Chi-square
13656.078
*** p<0.01, ** p<0.05, * p<0.1
Note: COVID-19 pandemic (COV)
In Table 3, F-test showed that overall, the explanatory variables including product of SPS of reporting country and its
trading partner (SPS1*SPS2), the product of TBT of reporting country and its trading partner (TBT1*TBT2), distance (D),
national language sharing (La), border sharing (Bo), Exchange rate (Exr), Product of GDP of trading pairs (GDP1*GDP2) and
COVID-19 (COV) had a strong impact on intra EAC trade. COVID-19 pandemic inversely and strongly impacted on intra
EAC countries trade and that the occurrence of the pandemic led to fall in intra EAC trade by 41.3% within the first twelve
months after it was first reported in EAC community nations. UNECA (2020) revealed that formal merchandise trade (imports
and exports) values declined sharply in the second quarter of 2020 when COVID-19 struck EAC countries. Technical barriers
to trade (TBT) had hindered trade and resulted to reduced intra EAC countries trade by 440.4%. This implies that more TBT
was imposed during the COVID-19 period with most EAC state reporting additional TBT measures in the year 2020 when
COVID-19 was ravaging EAC countries just like other countries globally. The TBTs introduced by nations were as a result of
efforts to control the spread of Virus where free flow of goods and people was controlled. This finding agrees with Ghodsi
(2020) who established major effects of TBT imposed by China on imports. Sanitary and Phystosanitary (SPS) policies directly
and strongly explained intra EAC trade. A one percent increase in SPS measures led to improved intra EAC trade by 160.3%.
Kinzius, Sandkamp and Yalcin (2019) showed NTBs resulted to falling. Baya (2019) showed that non-tariff barriers strongly
explained EAC countries trade.
Distance between the capital cities trading pairs acted as an inhibitor to intra EAC trade given the significant effect of
distance on trade value. A one percent more distance in kilometres led to reduced intra EAC trade by 3%. This implies that
EAC countries with longer total distance between their capital cities (economic centres) trade less compared to those whose
distance between the capital cities are shorter. Longer distance between capital cities implies more transportation cost.
Ricardian theory and H-O model assumed that trade between two trading countries has zero transportation cost as captured by
distance between trading countries capital cities leads to reduced trade. The product of Gross Domestic Products of reporting
country and trading partner strongly explained trade value. A 1% improvement in the product of reporting country and trading
partner GDP leads to increased trade value by 291.0%. Gross domestic product acts as mass for trade such that countries within
EAC with larger GDP traded more with each other compared with countries with smaller GDPs. Larger GDP means that the
countries have more purchasing power in terms of national income hence their demand for products also increases. Khaliqi,
Rifin and Adhi (2018) showed that GDP of the importers have direct effect on the export of shrimp from Indonesia. The effect
of sharing of national border on trade was direct with trade improving greatly especially in the COVID-19 period. Sharing of
border led to improved intra EAC trade by 414.1%. The finding implies that countries that shared a national border traded
more with each other especially during the COVID-19 period compared to countries that do not share national borders. Baya
(2019) showed that NTBs such as sharing of national language, sharing of border among others explained volume of trade. The
sharing of national languages inversely explained trade and that it resulted to 171.3% reduction in intra EAC trade. This
finding is against expectation where countries that share national language are expected to trade more with each other.
However, the findings agree with Irshad, Xin, Hui and Arshad (2018) where the effect of National language sharing on Trade
was inverse. The findings conflict with the Krugman and Obstfeld (2006) modified gravity model which showed that not
having a common national official language act as trade inhibitor just like distance between capital cities if trading countries.
Stay and Kulkarni (2015) revealed a positive effect of sharing a common national official language and trade between the UK
and its trading pairs.
32 Kithinji & Nganga: Effect of SPS and TBT Non-tariff barriers on Trade in EAC member countries
6. Conclusions and Recommendations
The study examined whether SPS & TBT NTBs and other covariates explained intra EAC member countries trade during
COVID-19 pandemic period. Pre COVID-19, the effect of the SPS notified by reporting country and its trading partners
strongly and inversely explained value of trade. A 1% improvement in the number of notified SPS by reporting country and its
trading partner leads to reduced value of trade by 6.9%. This finding implied that SPS was an inhibitor to intra EAC trade such
that they reduce the trade values between reporting EAC nation and its trading partners within EAC. The effect of TBT was
positive; however, it was not significant hence in can be concluded that between TBT abs SPS, SPS was a trade inhibitor in
EAC while the effect of TBT remained very low. During COVID-19 period, the pandemic inversely and strongly explained
intra EAC countries trade and that the occurrence of the pandemic led to fall in intra EAC trade by 41.3% within the first
twelve months after it was first reported in EAC community nations. Technical barriers to trade (TBT) had hindered trade and
resulted to reduced intra EAC countries trade by 440.4%. This implies that more TBT was imposed during the COVID-19
period with most EAC state reporting additional TBT measures in the year 2020 when COVID-19 was ravaging EAC countries
just like other countries globally. The TBT introduced by nations was as result of efforts to control the spread of Virus where
free flow of goods and people was controlled. Additionally, sanitary and Phystosanitary (SPS) policies directly and strongly
explained intra EAC trade. A one percent increase in SPS measures lead to improved intra EAC trade by 160.3%. Based on
objective three on the direction of causality between SPS and TBT Non-tariff barriers and Trade among EAC member
countries in the pre-COVID-19 pandemic period. The study results revealed that only Gross Domestic Product granger caused
intra EAC trade with both TBT and SPS non-tariff barriers failing to cause intra EAC trade. Additionally, Intra EAC trade does
not granger cause GDP, TBT and SPS. The study thus concluded that only Gross domestic product granger causes Trade with
TBT and SPS non-tariff barriers not causing trade. Further, based on the fact that COVID-19 strongly and inversely explained
intra EAC trade, the EAC communities should adopt deeper regional integration policies rather than nationalistic policies to
combat COVID-19 among its member states. Nationalistic polices of controlling COVID-19 proved to lead to increase in
barriers to trade hence inhibiting intra EAC trade. The countries must allow free flow of commodities and factors across
national borders of EAC states to ensure trade is not inhibited in any way. The countries should also be prepared for future
pandemics that may arise and react more with regionalism policies. Given the significant effect of TBT and SPS barriers to
trade based on both annual and monthly data, the study recommends to EAC countries to find robust mechanism of identifying
and elimination undesirable TBT and SPS. Further, the countries should adopt standardized TBT and SPS measures across the
EAC member states to improve better understanding by exports and importers among member states so as to encourage intra
EAC trade. Data scarcity was the main limitation. SPS and TBT data was only available annually from 2014 to 2020 from the
World Bank statistics. The study therefore limited the model with TBT and SPS to between 2014 and 2020. Further some
trading pairs lacked monthly data on trade amongst themselves. Tanzania and Rwanda did not report monthly intra EAC trade
with individual member countries. They only reported on monthly overall trade without specifying the destination. Therefore,
on the model based on monthly data, the study thus omitted trade pairs involving Kenya and Tanzania and Kenya and Rwanda.
Even with the limitation, the validity of study findings was not affected hence policy recommendations based on findings are
sound. It is critical that further research be carried out in effect of TBT and SPS using other proxies different from the one
used in this study. This would enable policy maker know whether the parameters change significantly with adoption of
different proxies. Another study can also be carried out focusing on specific products given that the effect of COVID-19, TBT
and SPS may different across products.
Conflicts of Interest
“The authors declare no conflicts of interest.”
References
[1] Albertoni, N., & Wise, C. (2021). International trade norms in the age of COVID-19 nationalism on the rise? Fudan Journal of the Humanities and
Social Sciences, 14(1), 41-66.
[2] Ali, S, M Fugazza and B Vickers (2020), Assessing the Impact of the COVID-19 Pandemic on Commodities Exports from Commonwealth Countries,
International Trade Working Paper 2020/14, Commonwealth Secretariat, London
[3] Baya, W. (2019). Effects of tariff and non-tariff barriers on intra-east Africa community trade in agricultural food commodities (Unpublished Masters
Thesis). Egerton University. Nakuru, Kenya.
[4] Calabrese, L., & Eberhard-Ruiz, A. (2016). What types of non-tariff barriers affect the East African Community (ODI Briefing Papers). Overseas
Development Institute. London, UK.
[5] East African Business Council (EABC) (2004). Protocol on the Establishment of the East African Community Customs Union. Arusha, EABC.
[6] EABC (2005). A Study on non-tariff Barriers (NTBs) and development of a business climate index in the East Africa Region. Arusha, EABC.
[7] EABC (2009). EAC Time-Bound Programme for Elimination of Identified Non-Tariffs Barriers. Arusha, EABC.
[8] EABC (2016). EAC Trade and Investment Reports. EABC, Nairobi, Kenya: East African Business council. Arusha, EABC.
[9] EABC (2020). State of EAC Economies amid COVID-19. Nairobi, EABC
Journal of Economics, Finance and Business Analytics 2024; 2(1): 22-33 33
[10] East African Community Secretariat-East African Business Council (EACS-EABC) (2006). Proposed Mechanism of the Elimination of Non-tariff
Barriers in EAC. Arusha, EACS-EABC.
[11] EACS (2008). Progress of implementation of the EAC council decisions and directives related to common market operations as at end of February 2008.
Arusha, EACS.
[12] EACS (2004). Protocol on the Establishment of the East African Community Customs Union. Arusha, EACS.
[13] EACS (2013). The East African Community trade report. Arusha, EACS.
[14] EACS (2015). The East African Community trade report. Arusha, EACS.
[15] EACS (2018). The East African Community trade report. Arusha, EACS.
[16] Eastern African Sub-regional Support (EASSI, 2020). The escalation of NTBs in the EAC and the impact on women traders. EASSI.
[17] EASSI (2020). Impact of COVID-19 on cross border traders. Kampala, EASSI.
[18] EABC (2020). Impact of COVID-19 on EAC Trade. EAC Policy brief. EABC. Arusha, EABC.
[19] Erokhin, V., & Gao, T. (2020). Impacts of COVID-19 on trade and economic aspects of food security: Evidence from 45 developing countries.
International journal of environmental research and public health, 17(16), 57-75.
[20] Escaith, H and S Khorana (2021), ‘The Impact of the COVID-19 Pandemic on Merchandise Trade in Commonwealth Countries’, International Trade
Working Paper 2021/02, Commonwealth Secretariat, London.
[21] Feenstra, R. & Taylor, A. (2017). International Economics, 4th edition. Worth Publishers.
[22] Gandolfo, G. International Trade Theory and Policy; Second Edition. Rome: Springer
[23] Ghodsi, M., Grübler, J., Reiter, O., & Stehrer, R. (2017). The Evolution of Non-Tariff Measures and their Diverse Effects on Trade, wiiw Research
Report, No. 419, The Vienna Institute for International Economic Studies (wiiw), Vienna.
[24] Ghodsi, M. (2020). The impact of Chinese technical barriers to trade on its manufacturing imports when exporters are heterogeneous. Empirical
Economics, 59(4), 1667-1698.
[25] Ghodsi, M. (2021). Impact of Technical Barriers to Trade on the Trade in Goods in the Information and Communications Technology Sector:
Differentiating by Aim of the Regulatory (No. 208). The Vienna Institute for International Economic Studies.
[26] Gujarati, D., & Porter, D. (2003). Econometría, London: Editorial McGraw-Hill.
[27] Gujarati, D. (2008). Basic Econometrics, New York: MeGraw-Hill.
[28] Kahenu, S. (2014). Non-tariff barriers in EAC Customs Union: implications for trade in manufactured goods in Kenya (Unpublished Masters Thesis).
University of Nairobi, Nairobi, Kenya.
[29] Khaliqi, M., Rifin, A., & Adhi, A. K. (2018). Trade Effect of Sanitary and Phytosanitary (SPS) and Technical Barriers to Trade (TBT) on Indonesia's
Shrimp Export. Indonesian Journal of Agricultural Research, 1(2), 134-141.
[30] Kinzius, L., Sandkamp, A., & Yalcin, E. (2019). Trade protection and the role of non-tariff barriers. Review of World Economics, 155(4), 603-643.
[31] Kothari, C. (2004). An introduction to operational Research. New Delhi: Vikas Publishing House Pvt Ltd.
[32] Krugman, P. R., & Obstfeld, M. (2006). Economía Internacional. Pearson Educación.
[33] Mortazavi, R. & Do, T. (2006). A gravity model for trade between Vietnam and twenty-three European Countries (Unpublished Masters Thesis).
Vietnam National University.
[34] Obuya, M. O., & Olweny, T. (2017). Effect of Bank’s Lending Behaviour on Loan Losses of Listed Commercial Banks in Kenya. International Journal
of Management and Commerce Innovations,5(1),135-144
[35] Ogunleye, O., Basu, D., Mueller, D., Sneddon, J., Seaton, R., Yinka-Ogunleye, A. & Godman, B. (2020). Response to the novel corona virus (COVID-
19) pandemic across Africa: successes, challenges, and implications for the future. Frontiers in pharmacology, 11.
https://doi.org/10.3389/fphar.2020.01205
[36] Ohlin, B. (1933). Interregional and international trade. Cambridge, MA: Harvard University Press (Rev. ed., 1967). (Excerpts are contained in W. R.
Allen, Ed., 1965, chap. 7)
[37] Ohlin, B., Hesselborn, P., & Wijkman, P. M. (Eds.). (1977). The international allocation of economic activity. London: Macmillan
[38] Okumu, L., & Nyankori, O. (2010). Non-Tariff Barriers in EAC Customs Union, Kampala Uganda: Economic Policy Research Center.
[39] Oiro, R., Owino, B., & Mendez-Parra, M. (2017). Non-tariff barriers and ‘complaints’ in the East African Community’s reporting process. CUTS
International Policy Briefing.
[40] Rojas, J. & Pineda, A. (2020). Tariff barriers and non-tariff barriers: appraising Colombia’s protectionism. World Customs Journal, 14(1), 71-94.
[41] Santeramo, F. G., & Lamonaca, E. (2020). The role of non-tariff measures in the agri-food sector: positive or negative instruments for trade? In Positive
Integration-EU and WTO Approaches towards the" Trade and" Debate (pp. 35-59). Springer, Cham.
[42] Stay, K., & Kulkarni, K. G. (2016). The gravity model of international trade, a case study: the United Kingdom and her trading partners. Amity
Global Business Review, 11, 28-39.
[43] Suranovic, S. (2012). Policy and Theory of International Trade. Rome; Springer.
[44] UNCTAD (2013). Economic and policy issues for developing countries. Conference on Trade and Development. Geneva, UNCTAD
[45] Vickers, D. Ali,B., Zhuawu, C., Zimmermann, A. Attaallah, H. and Dervisholli, E.(2020), International Trade Working Paper 2020/15. London:
Commonwealth Secretariat, London.
[46] Wooldridge, J., (2012). Introductory econometrics: A modern approach. Cengage Learning