The Journal of International Trade & Economic

Development
An International and Comparative Review

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rjte20

The upsurge of trade protectionism in sub-

Saharan Africa since the 2008 crisis: Is the twin
deficit guilty?

Guivis Zeufack Nkemgha, Boker Poumie & Hervé Kaffo Fotio

To cite this article: Guivis Zeufack Nkemgha, Boker Poumie & Hervé Kaffo Fotio (2022) The
upsurge of trade protectionism in sub-Saharan Africa since the 2008 crisis: Is the twin deficit
guilty?, The Journal of International Trade & Economic Development, 31:4, 475-492, DOI:
10.1080/09638199.2021.1985160

To link to this article: https://doi.org/10.1080/09638199.2021.1985160

Published online: 11 Oct 2021.

Submit your article to this journal

Article views: 89

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rjte20
THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT
2022, VOL. 31, NO. 4, 475–492
https://doi.org/10.1080/09638199.2021.1985160

The upsurge of trade protectionism in sub-Saharan Africa

since the 2008 crisis: Is the twin deficit guilty?
Guivis Zeufack Nkemghaa , Boker Poumieb and Hervé Kaffo Fotiob
a Economic Department of the Higher Teacher Training College (HTTC) of Bambili, University of

Bamenda, Bambili, Cameroon; b Economic Department at the Faculty of Economics and

Management, University of Maroua, Maroua, Cameroon

ABSTRACT
Despite the growing literature on the political economy of trade protectionism, little is
known about the effect of twin deficit on trade restrictions in general around the world
and particularly in sub-Saharan Africa (SSA). The aim of this study is to fill this gap by
assessing how twin deficit affects trade restrictions in SSA. We used a panel of 16 coun-
tries which recorded simultaneously and successively the current account deficit and
the budget deficit during the period 2008–2018. The empirical evidence is based on the
system Generalized Method of Moment (GMM). Our results show that the twin deficit is
responsible for the upsurge of trade protectionism in sub-Saharan Africa.

KEYWORDS Trade protectionism; twin deficit; SSA; GMM

JEL CLASSIFICATIONS F13, O24, F32

ARTICLE HISTORY Received 26 May 2021; Accepted 20 September 2021

1. Introduction
The year 2016 was the fifth year in a row with merchandise trade growth below 3%,
well below the pre-crisis average of 7%. Despite the cyclical recovery observed in 2017,
the weak post-crisis trade growth reflects a number of factors. On the one hand, weak
global demand is associated with the post-crisis legacy in advanced economies, wors-
ening terms of trade for commodity exporters and the transition to slower growth in
China. On the other hand, longer-term trends have not supported trade growth either.
As a result, the long-run income elasticity of trade has declined (World Bank 2015; Con-
stantinescu, Mattoo, and Ruta 2015), reflecting a shift in demand towards nontradables
and services due to an aging population, a slowdown the expansion of global value chains
and the slowing pace of trade liberalization (IMF 2016).
Not only has the freeze on trade weighed on trade growth, but the post-crisis period
has seen an increase in the number of newly introduced protectionist measures. The
World Trade Organization (WTO) recently warned against this worrying trend, not-
ing that the rate of new restrictive trade measures introduced by G20 countries in 2016

CONTACT Guivis Zeufack Nkemgha nkemghazeufackguivis@yahoo.fr

© 2021 Informa UK Limited, trading as Taylor & Francis Group

476 G. Z. NKEMGHA ET AL.

Figure 1. Comparative evolution of tariffs rate applied between SSA et and the world. Source: Authors through
World Bank statistics.

reached the highest monthly average since 2009 (21 new measures per month), outnum-
bering measures to facilitate trade (WTO 2016). Likewise, based on a broader definition
of protectionist measures, the latest Global Trade Alert (GTA) report reveals that despite
the recent decrease in the number of new protectionist measures, the stock of trade bar-
riers in force does not has stopped increasing (Evenett and Fritz 2017). Among these
measures, import tariff increases account for almost a fifth of the trade barriers imposed
since 2009.
The 2008–2009 financial crisis sparked fears of a potential global protectionist spi-
ral (Baldwin and Evenett 2009) and fears of a shift away from the use of traditional
trade policy instruments such as import tariffs towards ‘murky’ forms of protectionism.
The period of crisis has indeed seen a steady increase in the number of protectionist
measures, which total an annual average of more than 800 new harmful interventions
(Evenett and Fritz 2017). The use of murky forms of protectionism has also increased,
although traditional trade barriers remain the most widely used policy instrument.
Import duty increases represent nearly a quarter of the new trade barriers introduced
since 2009. Similarly, the Sino-US trade war that dates back to January 2018 is lodged in
this register.
Compared to the world average, the average tariff rate in SSA is more than two times
much higher during the period 2008–2017, as shown in Figure 1. However, this graph
tells us that between 2008 and 2017, customs tariffs increased by 2.47%. This is proof
that SSA was not spared by the wind of post-crisis protectionism in 2008. Furthermore,
statistics from the World Bank (2015) reveal that SSA is the second region in the world
where the applied tariff rates are the highest and the first region when referring to the
Most Favored Nation (MFN) clause as shown in Table 1. It is important to remember
that non-tariff measures are relatively less present in Africa. However, border corruption,
cumbersome customs procedures and poor infrastructure are all things that seem to be
an obstacle to trade openness.
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 477

Table 1. Tariffs by region. Source: World Bank WITS database. Note: Trade weighted
averages for 2015.
Applied (%) MFN (Most-Favoured Nation) (%)
European Union 1.5 2.5
United States 1.6 2.7
Other advanced economies 2.6 4.2
China 3.4 3.9
East Asia and Pacific 1.3 2.5
Europe and Central Asia 2.6 4.8
Latin America and Caribbean 6.7 9.0
Middle East and North Africa 4.9 6.3
South Asia 7.7 8.0
Sub-Saharan Africa 7.5 9.2
World average 2.7 3.8

In the absence of tariff commitments under regional trade agreements (RTAs) and
unilateral preferences, WTO members would effectively revert to MFN tariffs, which
in turn would result in a 40 percent increase in tariffs global average, from 2.7 percent
to 3.8 percent. Average import tariffs in advanced economies such as the EU and the
US are among the lowest, ranging from 1.5 to 1.6 percent. In contrast, in the emerg-
ing and developing economies (EMDEs) of South Asia and sub-Saharan Africa, the
tariffs actually applied are five times higher than those in advanced economies with
average rates ranging from 7.7 to 7.5 percent respectively. In addition, a comparative
analysis of the different tariff rates by country reveals that the SSA countries are the
most protectionist countries in the world while the countries of the European Union,
Asia and America occupy the position of the least protectionist countries as shown in
Figure 2. In addition, over the past two decades, developing economies have become
important users of anti-dumping measures in their trade protection policies (Kang and
Ramizo 2019); the aim being to eliminate the injury caused to the domestic industry
of the importing country. However, many developing countries in the 1980s and early
1990s significantly reduced tariff and non-tariff barriers, and exposed their economies
to external competition (Mengyun et al. 2019).
Given that all regions of the world experienced a surge in protectionism after the 2008
crisis, are the reasons for trade restrictions in other parts of the world valid for SSA?
Obviously, the factors explaining trade restrictions in the rest of the world cannot be
transposed to Africa because this continent lags behind others on issues related to devel-
opment in general and trade in particular. To this end, the structure of the economies of
the countries of sub-Saharan Africa could be at the origin of these trade restrictions. The
graphic representation of two indicators characteristic of the structure of the economies
of sub-Saharan Africa (Figure 3) reveals that between 2008 and 2018, 16 countries
(Benin, Burkina Faso, Burundi, Ethiopia, Ghana, Kenya, Madagascar, Mali, Mauritus,
Mozambique, Senegal, Sierra Leone, South Africa, Tanzania, Togo, Uganda) successively
and simultaneously realized the budget deficit and the current account deficit. This sit-
uation presumes the existence of the twin deficit thesis in most of the most protectionist
countries in the world (SSA countries). The existence of this twin deficit over more than
a decade can be explained by a combination of sets of factors such as the volatility of
the price of raw materials, particularly oil, the persistence of internal conflicts, climate
change which affects the level of domestic supply, the absence of a real industrialization
policy and governance problems.
478 G. Z. NKEMGHA ET AL.

Figure 2. Tariffs by country. Source: World Bank WITS database (2015). Note: Trade weighted averages for 2015.
Includes countries with population of at least 3 million. (A,B) Applied tariffs are a combination of preferential tariffs
(where these exist) and MFN tariffs. (C,D) Most Favored Nation (MFN) tariffs are normal, non-preferential tariffs
charged by WTO members to trading partners that are members of the WTO. (E,F) Bound tariffs are maximum MFN
tariff level.

Figure 3 shows that there is a positive association between the budget deficit and the
current account deficit. This means that an increase in the first deficit leads to a wors-
ening of the second. In parallel to Figure 3, the observation of Figure 1 above shows a
resurgence of trade protectionism since the 2008 crisis. Thus, the twin deficit and tariff
barriers have moved in the same direction since the 2008 crisis. In the end, we therefore,
see that an increase in the budget deficit (trigger) leads to a worsening of the current
deficit and the conjunction of the two deficits leads to the resurgence of trade protec-
tionism through the rise in tariff barriers: this phenomenon can be described as trade
protectionism syndrome. Because a syndrome is a set of signals (the coexistence of the
budget deficit and the current deficit) revealing an uncomfortable situation (considered
bad) and characterizing a pathological state (protectionism).
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 479

Figure 3. Evolution of the current account balance and balance budget in some selected sub-Saharan African
countries. Source: Authors through Stata 12 software.

From the above, the objective of this work is to verify whether the coexistence
between the budget deficit and the current account deficit explains the trade restric-
tions in sub-Saharan Africa. To sum up, the results reveal the existence of the trade
protectionism syndrome in SSA. Thus, reducing the budget deficit and speeding up the
industrialization process with the aim of being competitive in international markets will
reduce the trade deficit and promote free trade in SSA.
The rest of the paper is structured as follows: section 2 is devoted to the literature
review. The methodological approach will be the subject of section 3. The analysis of
results is reserved for section 4 and section 5 will conclude.

2. Literature review
This review is structured in two subsections: the theoretical review and the empirical
evidence.

2.1. Theoretical review of the explanatory factors of trade protectionism

Protectionism is the sum of government trade policies aimed at helping domestic pro-
ducers against foreign producers in a given industry, by raising the price of foreign
products, lowering costs for domestic producers, and limiting foreign producers access
to the market. The methods for obtaining such protection are diverse and can be grouped
into two categories: tariff barriers and non-tariff barriers. Tariff barriers refer to import
taxes and anti-dumping or variable import levies against dumping practices. As for
non-tariff barriers, also called invisible barriers, they are diverse and varied. It’s about:

• Quotas on the quantity of foreign products sold on the domestic market, which limit
the supply and increase the price of imported products;
• Regulatory obstacles that hamper imported products such as product classifications
and seemingly endless lists of standards and specifications (health, human safety and
environment);
480 G. Z. NKEMGHA ET AL.

• Subsidies to domestic producers ranging from tax breaks to direct cash payments;
and
• Currency controls to limit access to foreign currencies or manipulate exchange rates
to inflate the price of foreign products and lower the price of domestic products.

These regulations and instructions are in fact seen as invisible trade barriers to inter-
national trade. Thus, domestic producers are protected against foreign competition.
Although there was a respectable drop in tariffs and import quotas after World War
II, the number and extent of invisible barriers increased dramatically. Indeed, today, the
main obstacle to world trade lies in invisible barriers rather than tariffs.
Theoretically, arguments for protectionism include: personal income, balance of
payments, employment, infant industries and rent seeking.

2.1.1. Income of individuals

Support for protectionism can be explained by the impact of trade on individual income.
Comparative advantage models explain how international trade affects the income of
individuals through changes in relative prices. The Heckscher-Ohlin (H-O) model
assumes full factor mobility at no cost between sectors and is often presented as a
long-term view. In contrast, the Ricardo-Viner (R-V) model assumes the existence
of sector-specific factors and is therefore often presented as a medium-term model.
According to the H-O model, unskilled workers in unskilled countries should support
free trade, while skilled workers should oppose it. The reverse would happen in skill-rich
countries. On the other hand, the R-V model predicts that factors specific to the import-
ing sector will lose out from trade liberalization, but factors specific to the exporting
sector will gain. Since the workforce is not very mobile across sectors in the short term,
individual trade policy preferences will depend on whether the person is employed in
an import substitution industry or export (Gawande and Krishna 2003).

2.1.2. Balance of payments

Advocates of protectionism are alarmed by the current account deficit of the balance
of payments. When the trade deficit persists and widens, politicians begin to exercise
protectionism to tackle the perceived injustice in the country’s trade relations. This argu-
ment, while popular, ignores key issues. First, studies of the economics of trade continue
to show that the trade deficit in itself is not detrimental to the economy. Second, focus-
ing only on the trade balance is only a partial and limited view of a country’s balance
of payments. A positive balance in another account, such as the capital account, which
measures capital inflows and outflows, accompanies a trade balance deficit. This is the
case with the American balance, and it is a fact seldom mentioned by arguments in favor
of protectionism. Third, protections that reduce imports invariably reduce exports and
do not change the deficit situation. Kaempfer and Willett (1987) investigated the issue
of using the import surcharge to reduce the trade deficit and concluded that the policy
would distort resource allocations and fail to reduce the deficit. They suggest that the
other deficit be corrected, the government budget deficit, as this promises to reduce the
trade deficit. Recent studies by the Institute for International Economics come to the
same conclusion: restricting imports is not recommended as this would harm GDP (US
Current 2008).
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 481

2.1.3. Employment
An industry that has not prepared for global or national competition loses market share
and jobs are lost. Workers, their representatives and employers lobby the government
very strongly for protections, and they often do. The protections reduce imports and pre-
serve some jobs, but the subsequent reduction in exports reduces employment in export
industries. Luttrell (1978) demonstrated that employment gains resulting from reduced
imports and losses resulting from reduced exports balance out with a net employment
effect close to zero! Third, if employment in the buffered industry is spared, it worsens in
industries that depend on imports, such as industrial users of imported goods, retailers,
trade-related service industries. In addition, the rise in the price of protected products
increases the cost of doing business in these industries and makes them less competitive.
Some end up moving or closing.

2.1.4. Infant industries

This is another popular argument in both developing and developed countries. A newly
established industry may not take advantage of the cost and production savings enjoyed
by competitors who have been in business long enough to develop production effi-
ciencies and innovative technologies. As a result, the newly established industry exerts
pressure on its government to protect it from international competition by imposing
trade restrictions on imports for a number of years until the domestic industry is likely
to establish its comparative advantage. Unfortunately, the protected industry contin-
ues to rely on its political power and allies to prolong the duration of its ‘childhood’
and resist the lifting of protections (Coughlin, Chrystal, and Wood 1988). These infant
industries enjoy the luxury of protection and often grow in size and begin to resemble an
oligopoly with significant political power to preserve and even increase levels and types
of protection (Pincus 1977).

2.1.5. Rent seeking

The authoritative studies of Tullock (1967) and Krueger (1974) suggest that the govern-
ment intervenes in economic activities, such as in commerce, in response to lobbying
efforts by interest groups seeking to protect their interests by government. The process
is called ‘rent seeking,’ which is common when government protection is sought. As
the government rewards ‘rent-seekers,’ the public begins to distrust the fairness of the
market system and begins to realize that it is fruitful to pressure the government to step
in and back out protected. For example, the public sees that the efforts of the American
farmer’s lobby to secure the protection subsidies benefit farm households by raising their
income above the national average. Data from the Department of Agriculture shows that
in 2006, the average farm household income was $ 77,654.17 percent higher than the
average US household income, and that in 2008, the average income of farm households
is expected to reach $ 90,000 (Etter and Hitt 2008). The farm bill was assessed by the
Bureau of Management and Budget, which concluded that ‘the agricultural economy
has never been stronger and there is no need to increase support and subsidies to Rich-
est 2% Americans’ (Gordon and Cho 2008). As a result of ‘rent seeking’ more and more
people are beginning to follow the path of seeking government intervention and pro-
tection. Lamb (2006) provides an illustrative example of rent-seeking by the California
agriculture lawyer who used this and other political tactics to block avocado imports
from Mexico for more than 80 years and continue to maintain Mexican exports to only
7 percent of annual consumption of avocados in the United States.
482 G. Z. NKEMGHA ET AL.

What about empirical work on the explanatory factors for protectionism?

2.2. Empirical evidence on the determinants of trade protectionism

The empirical evidence for the determinants of trade protectionism can be categorized
into two groups: economic determinants and non-economic determinants.

2.2.1. The economic determinants

The literature devoted to the influence of the macroeconomic context on the protection-
ist attitude of States is scarce. Baldwin (1985) finds that for the United States, sectors with
low wages and a high labor/output ratio tend to be highly protected. This implies that
the protection is used to help low income groups. He also finds that protection levels are
positively related to industry employment levels, supporting the adding machine model,
which implies that an industry’s trade protection is related to its voting strength. Finally,
he notes that protection is inversely proportional to the number of companies in the sec-
tor. This suggests that oligopolistic industries find it easier than competitive industries
to overcome the free rider problem to pass protectionist legislation (confirming a result
found by Pincus (1977) for the Tariff Act of 1824).
Denslow and Fullerton (1996) stress the influence of risk aversion on people’s atti-
tudes. Using a survey of consumers in Florida, they find that confidence in future
national performance and education are the most important determinants of consumer
attitudes towards trade policy. Similarly, Scheve and Slaughter (2004) show that eco-
nomic integration and foreign direct investment, in particular, have increased worker
insecurity in the UK by increasing wage and job volatility employment. This result
could partly explain the attitude of people towards free trade in the UK. Regarding the
macroeconomic context, Beaulieu, Yatawara, and Wang (2005) find that countries that
are more trade supportive tend to be countries that have grown faster over the past
decade. Caplanova, Orviska, and Hudson (2004) analyze attitudes towards EU mem-
bership in Central and Eastern European countries. They find that good economic
performance increases the likelihood of sustaining membership. Does the good eco-
nomic performance of a membership take into account the situation of the balance of
payments?
Another relationship that has received very little attention in the literature is how the
balance of payments situation or trade policies influence the opinion of citizens. To this
end, the study by Melgar, Milgram-Baleix, and Rossi (2013) suggests that people reject
the possibility that trade liberalization could alleviate the balance of payments problem
in dependent countries by lowering prices. Similarly, people value exports positively,
especially in smaller countries. Therefore, these results clearly prove that the public does
not adhere to free trade and Smith’s laissez-faire attitude, at least through a more neo-
Mercantilist view that considers a favorable trade balance to be associated with a healthy
economy; a situation that should be achieved according to this point of view through
protectionist measures.
The main contribution of Melgar, Milgram-Baleix, and Rossi (2013) is to test the
influence of certain macroeconomic factors on individual preferences for free trade.
They find that a sharp increase in the unemployment rate or the inflation rate increases
protectionist attitudes, indicating that individuals are not convinced that free trade will
lead to lower prices or the creation of jobs. In an unstable macroeconomic context, the
fear of adjustment costs outweighs the positive effect that free trade could have through
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 483

a reallocation of resources. Thus, they conclude that the overall appreciation of the con-
sequences of trade liberalization for the economy as a whole is inconsistent with the
optimism of trade theory for free trade.
Nicita (2013) analyzes the effect of exchange rate misalignments on government
decisions regarding trade policies. The methodology consists of estimating fixed-effects
models on a detailed panel data set comprising about 100 countries and 10 years
(2000–2009). The results indicate that exchange rate misalignments affect international
trade flows substantially. Because the authors find that the undervaluation of the cur-
rency favors exports and restricts imports.
In addition to economic determinants, non-economic factors also justify trade pro-
tectionism.

2.2.2. Non-economic determinants

Previous literature has shown that trade policy preferences also depend on non-
economic factors such as the median voter, lobbies, political ideology, and demographic
values and characteristics.
Dutt and Mitra (2001) empirically explore the Stolper-Samuelson argument. They
find that all countries protect, but an increasing gap by which the median family capital-
labor ratio is lower than the country average raises protectionist barriers in capital-rich
countries and lowers them in abundant countries in the labor force, providing interim
support for the median voter theory. Baldwin and Magee (2000) provide further support
for the theory, with their conclusion that a high proportion of less educated workers in
an electorate, i.e. a lower human capital endowment per worker, makes a representa-
tive American more likely to vote for protectionist legislation. Similarly, Kaempfer and
Marks (1993) find that votes in favor of accelerated authority (for the Bush adminis-
tration to negotiate the Uruguay Round and NAFTA) were positively related to average
wages (human capital) and negatively work-related. If the middle voter influences trade
policy, what about industries represented by lobbies?
Unsurprisingly, Goldberg and Maggi (1999) and Gawande and Bandyopadhyay
(2000) both found that everything else being the case, tariffs are on average higher
in industries represented by organized lobbies, while Gawande, Krishna, and Robbins
(2001) find that industries with organized foreign lobbies have lower trade protection
on average than in industries without such lobbies. Consistent with this research, Magee,
Brock, and Young (1989) found that Republican administrations generate more protec-
tion than Democratic administrations. Their mechanism is that Republicans tend to be
more ardent fighters for inflation, resulting in a recession and pressure for protection-
ism. Besides political ideology, do values and demographics influence the protectionist
attitude?
Regarding the influence of religious beliefs, Guiso, Sapienza, and Zingales (2003)
argue that attitudes towards dealing with ‘others’ and accepting ‘others’ differ among
religious denominations. In their study for the United States, they find that Catholics,
Baptists, and Methodists are more likely to support trade restrictions than those with no
religious affiliation. Without a doubt, attachment to the country and national pride count
in any debate about foreign policies. O’Rourke and Sinnott (2006) argue that these values
play an important role in trade policy preferences as these could translate into a sense of
national superiority and rejection of foreign products. Finally, many demographic vari-
ables may be relevant in explaining political preferences. For example, with respect to age
and gender, previous empirical studies have shown that older people are more likely to
484 G. Z. NKEMGHA ET AL.

support restrictive import policies than younger ones. The same can be said for women
as compared to men. Moreover, some of the empirical studies, (Mayda and Rodrik 2005)
or (O’Rourke and Sinnott 2006), find that married people are also more likely to support
trade restrictions.
There are several arguments in favor of protectionism. However, the record of trade
in modern history continues to show that protectionism is ultimately harmful to society.
It emerges from this literature that no study has yet analyzed the link between the
twin deficit and trade protectionism. However, the economic structure of several coun-
tries on the African continent shows a coexistence of the budget deficit and the current
account deficit over more than a decade. Thus, the aim of this study is to fill this gap in
the literature by analyzing the factors that contribute to the formation of trade protec-
tionism in the countries of sub-Saharan Africa by highlighting their common structural
characteristic like the twin deficit.

3. Methodology
This methodology is structured around two main axes: the presentation of the data and
the estimation technique.

3.1. Data
We investigate a panel of 16 African countries with data for the period 2008–2018 from
World Development Indicators (WDI). The periodicity and countries under investiga-
tion are chosen according to data availability constraints. The full description of the data
is presented in Tables 2 and 3.
The dependent variable is trade protectionism measured by the tariff rate. This
variable is increasingly being used in the recent literature (Nicita 2013). The main inde-
pendent variables are Current Account Deficit (CAD) and Budget Deficit (BD). To
reduce the bias that may arise from possible variable omissions, four control variables are

Table 2. Summary statistics. Source: Authors.

Variables Observations Mean SD Minimum Maximum
Tariffrate 176 8.0226 3.0383 0.68 17.22
Current Account Deficit 176 −8.9395 7.9902 −65.0289 −0.4957
Budget Deficit 176 −3.9251 2.0691 −10.718 −0.4
Industrial employment 176 12.3410 8.0707 1.505 33.43
Manufacturing, added value 176 8.7814 3.8775 1.533 19.037
Gross Domestic Product 176 5.0432 3.5647 −20.6 20.72
Terms of Trade 176 121.2451 31.219 48.72 190.261

Table 3. Pairwise correlation. Source: Authors.

Tariffrate CAD BD Indempl Manuf GDP Tot
Tariffrate 1.000000
CAD 0.1037 1.000000
BD 0.0211 0.2193 1.000000
Indempl −0.3105 0.3161 0.0818 1.000000
Manuf −0.3166 0.1987 0.0338 0.3325 1.000000
GDP 0.1978 −0.0782 0.0361 −0.1207 −0.2245 1.000000
Tot −0.0344 0.3281 −0.0327 0.1767 −0.0476 0.10171 1.000000
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 485

included in this study. They comprise (i) industrial employment, (ii) added value of man-
ufacturing, (iii) Gross Domestic Product and (iv) terms of trade. A detailed description
of variables as well as definition of variables are presented in the appendix. Table 1 gives
the summary statistics of the variables and the Pairwise correlation analysis is presented
in Table 2.

3.2. Estimation method

The aim of this paper is to investigate the impact of twin deficit on trade protectionism in
sub-Saharan African countries. According to the recent literature on trade protectionism
(Nicita 2013), we formulate the following quadratic form model:

Tariffrateit = β0 + β1 Tariffrateit−1 + β2 CADit + β3 BDit + β4 BDit ∗ CADit + β5 Xit

+ μi + vt + εit (1)

where Tariffrateit , represents the level of trade protectionism for country i in the period
t, CADit is the current account deficit for a country i in the period t, BDit is the bud-
get deficit for a country i int t period, BD ∗ CAD is the interaction variable between the
budget deficit and the current account deficit. Xit is a vector which includes all control
variables, μi is an unobserved country-specific effect, vt is time specific effect, and εit is
the error term. The introduction of the lagged dependent variable in equation (1) aims
to check whether the level of future trade protection depends on the level of previous
trade protection. By analyzing the effect of the import surcharge on the trade deficit,
Kaempfer and Willett (1987) concluded that this policy would distort resource alloca-
tions and fail to resolve the deficit. To this end, they suggested that correcting the budget
deficit promises to reduce the trade deficit. Referring to this work, we understand that
the current account deficit may be the channel through which the budget deficit exerts
its effect on trade protectionism. It is for this reason that we decided to introduce the
interaction variable BD ∗ CAD. Due to the fact that equation (1) is a dynamic model,
there is higher probability that the lagged value of trade protectionism (Tariffrateit−1 )
is correlated with the error term (Nickell 1981) raising the problem of endogeneity. To
address this endogeneity issue, we apply the System Generalised Method of moment
(GMM) proposed by Arellano and Bond (1991), Arellano and Bover (1995) and Blun-
dell and Bond (1998). GMM is useful for several advantages. First, GMM estimator has
been widely used to address the endogeneity problem that appears in panel data estima-
tion (Arellano and Bover 1995; Blundell and Bond 1998). Second, GMM estimator also
take into account the biases that appear due to country-specific effects. Third, GMM also
avoids simultaneity or reverse causality problems. The consistency of the GMM estima-
tor depends on two things: the validity of the assumption that the error term does not
exhibit serial correlation (AR (2)) and the validity of the intruments (Hansen test).

4. Empirical results
The results are presented in the various tables below. Table 4 was used to verify the exis-
tence of the twin deficit hypothesis before proceeding to estimate the trade protectionism
model. The Pesaran (2004) test allowed us to conclude that the variables budget deficit
and current account deficit admit a transverse dependence with regard to their respec-
tive probabilities 0.03 and 0.00 (these probabilities are less than 5%). The analysis of the
486 G. Z. NKEMGHA ET AL.

Table 4. Results of the effect budget deficit on current account deficit. Source: Authors.
Dependent variable: CAD OLS GLS Driscoll-Kraay
BD 0.8468∗∗∗ 0.8468∗∗∗ 0.8468∗∗∗
(0.003) (0.00) (0.00)
Cons −5.6154∗∗∗ −5.6154∗∗∗ −5.6154∗∗∗
(0.00) (0.00) (0.00)
R2 0.0481 0.048
F-stat 8.79 52.52
Wald (Chi2 ) 8.89
Prob (F-stat) 0.00 0.00
Prob (Chi2 ) 0.00
Nb. obs 176 176 176
Nb. countries 16 16 16

impact of the budget deficit on the current account deficit by the Ordinary Least Squares,
Generalized Least Squares (it is robust to autocorrelation and heteroskedasticity) and
Driscoll-Kraay methods (it is flexible for dependencies across countries, heteroskedas-
ticity, and autocorrelation) shows a positive and significant association between the two
variables, which attests to the existence of the twin deficit in the countries of sub-Saharan
Africa. This result is compatible with the work of Ahmed and Ansari (1994).
With the twin deficit hypothesis verified, we will analyze the impact of the coexistence
of the two deficits on trade protectionism. The results from this estimation are confined
in Table 5. The GMM estimation regressions mutually satisfy the Hansen instrument
validity test and the AR series correlation test (2). In addition, the coefficient of the
lagged dependent variable (which is almost one) is positive and significant over the eight
columns. This means that the future level of trade protection of an economy depends on
the previous level of trade protection of that economy.
Analysis of Table 5 shows that the current account deficit and the budget deficit each
have a positive and significant effect on tariff barriers. Thus, an increase in the current
account deficit and the budget deficit of 1% leads to an increase in customs tariffs of
0.29% and 1.05% respectively (model 7). This result can be explained by the fact that in
the event of a twin deficit, governments can resort to borrowing or increase their tariffs
(because the latter contribute significantly to the budgets of African countries) with the
aim in particular of filling their budget deficits. In doing so, they also reduce their current
account deficits. Similarly, we can also observe through the interaction variable (BD ∗
CAD) that the current account deficit is the channel through which the budget deficit
exerts a positive impact (0.0675) on tariffs rate. This result is consistent with the work
of Kaempfer and Willett (1987) and (US Current 2008). In the end, we see through this
result that an increase in the budget deficit (trigger) leads to a worsening of the current
account deficit and the conjunction of the two deficits leads to the resurgence of trade
protectionism through the rise in tariff barriers: this phenomenon can be qualified as
trade protectionism syndrome. Because a syndrome is a set of signals (the coexistence of
the budget deficit and the current account deficit) revealing an uncomfortable situation
(considered bad) and characterizing a pathological state (protectionism).
Industrial employment has a positive and significant effect on tariffs. Thus, an
increase in industrial jobs of 1% leads to an increase in trade restrictions of 0.03% (model
7). This situation is explained by the fact that the protections reduce imports and pre-
serve certain jobs. It is for this reason that workers, their representatives and employers
Table 5. Results of the estimation of the trade restriction model (GMM).
1 2 3 4 5 6 7 8
CAD 0.0160∗∗∗ 0.0108∗∗∗ 0.1239∗∗ 0.1530∗∗∗ 0.1189∗∗ 0.3050∗∗ 0.2929∗ 0.1985∗∗∗
(0.00) (0.00) (0.050) (0.00) (0.03) (0.02) (0.055) (0.00)
0.0482∗ 0.3858∗∗ 0.4243∗∗ 0.3516∗∗ 1.0993∗∗ 1.0572∗∗ 0.4701∗

THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT

BD
(0.09) (0.03) (0.01) (0.04) (0.02) (0.050) (0.07)
BD∗CAD 0.02401∗ 0.0321∗∗∗ 0.0260∗∗ 0.0723∗∗ 0.0675∗ 0.0473∗∗∗
(0.08) (0.00) (0.03) (0.02) (0.07) (0.00)
Indempl 0.0241∗ 0.0154 0.0318∗∗ 0.0384∗ −0.0351
(0.07) (0.243) (0.03) (0.07) (0.143)
Manuf 0.0476∗∗∗ 0.0706∗∗∗ 0.0501∗∗∗ 0.0560∗∗
(0.00) (0.00) (0.00) (0.04)
GDP −0.042∗∗∗ −0.0306∗ −0.0288∗∗
(0.00) (0.08) (0.034)
Tot −0.0087 0.0081
(0.41) (0.238)
CAD∗BD∗Manuf −0.0011∗∗
(0.033)
Lag Tarrifrate 1.0012∗∗∗ 0.9826∗∗∗ 1.0044∗∗∗ 1.0588∗∗∗ 1.0831∗∗∗ 1.0865∗∗∗ 1.0766∗∗∗ 0.9892∗∗∗
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
CONS 0.0233 0.3392 1.5841∗ 0.9583 0.0486 2.7105 3.7943∗∗ 1.2297
(0.97) (0.25) (0.09) (1.01) (0.960) (0.14) (0.02) (0.33)
Prob(Chi2 ) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Prob(AR2) 0.362 0.359 0.312 0.295 0.304 0.291 0.297 0.294
Prob(Sargan) 0.467 0.771 0.906 0.914 0.889 0.891 0.892 0.742
Nb. observations 160 160 160 160 160 160 160 160
Nb. countries 16 16 16 16 16 16 16 16

487
488 G. Z. NKEMGHA ET AL.

Table 6. Robustness check (with other

control variables).
Variables Coefficients
CAD 0.5983∗∗
(0.03)
BD 1.4829∗
(0.055)
BD∗CAD 0.1637∗
(0.050)
Indempl −0.2788∗
(0.050)
Manuf 1.4121∗
(0.06)
GDP −0.1016
(0.10)
Tot −0.1264∗∗
(0.03)
CAD∗BD∗Manuf −0.0097∗∗
(0.08)
Mobile 0.0233∗
(0.054)
INF 0.0863∗∗
(0.03)
Goveff −4.9024∗
(0.07)
Lag Tarrifrate 0.9520∗∗∗
(0.00)
Cons 10.2625∗∗
(0.03)
AR(1) 0.032
AR(2) 0.29

Prob (Sargan) 0.63

Prob(Chi2 ) 0.00
Obsersations 160

are putting so much pressure on their government for protections. This result is not com-
patible with the work of Luttrell (1978) who demonstrated that the employment gains
resulting from the reduction in imports and the losses resulting from the reduction in
exports are balanced with a net effect on close employment from zero!
The value added of the manufacturing sector has a positive and significant effect on
trade restrictions. Thus, an increase in value added in the manufacturing sector of 1%
leads to an increase in customs tariffs of 0.05% (model 7). This result is justified by the
fact that a newly established industry may not take advantage of the cost and produc-
tion savings enjoyed by competitors who have been in business long enough to develop
production efficiencies and innovative technologies. It is for this reason that it exerts
pressure on its government to protect it from international competition by imposing
trade restrictions on imports for a number of years until the domestic industry pre-
sumably establishes its position comparative advantage. Unfortunately, the protected
industry continues to rely on its political power and allies to prolong the duration of
its ‘childhood’ and resist the lifting of protections (Coughlin, Chrystal, and Wood 1988).
As for the gross domestic product, it has a negative and significant effect on trade
restrictions. Thus an increase in economic growth of 1% leads to a reduction in trade
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 489

restrictions of 0.03% (model 7). This result is consistent with classical free trade the-
ory which states that there is a positive association between economic growth and trade
liberalization and. The results of Jouini’s work (2015) confirm this position.
The results of the last column of Table 5 (model 8) show that the value added of the
manufacturing sector makes it possible to attenuate the syndrome of trade protectionism
(that an increase in the budget deficit leads to a worsening of the current account deficit
and the combination of two deficits leads to renewed trade protectionism through the
raising of customs barriers). This situation is explained by the fact that an increase in
manufactured products will make it possible, on the one hand, to increase State revenues
and, on the other hand, to reduce the trade deficit in order to promote greater openness
through lower trade restrictions. The fact that the coefficient of the interaction variable
BD∗CAD∗MANUF is low justifies the embryonic nature of the manufacturing sector in
sub-Saharan Africa.
In order to test the sensitivity of our results, we opted to introduce other control
variables (inflation, number of mobile phones and government effectivness) in model
8 of Table 5 and to keep the technique of GMM system. The results of this estimate
are reported in Table 6. It emerges from Table 6 that each deficit constitutes an explana-
tory factor for trade protectionism. In addition, the current account deficit is the channel
through which the budget deficit hinders trade openness. Finally, industrialization is pre-
sented as a solution to mitigate the effect of the twin deficit on trade openness. Overall,
these results corroborate those in Table 5.

5. Conclusion and policy implications

A large part of the literature on the political economy of protectionism is concerned with
the consequences of trade restrictions on macroeconomic variables. However, studies
on the causes of trade protectionism are few in developed countries and almost non-
existent in developing countries, particularly in Africa; continent that did not escape the
resurgence of post-crisis protectionism in 2008. It is on this basis that we focused our
interest on the analysis of the causes of trade protectionism in sub-Saharan Africa by
highlighting the structural factors common to these countries. We used a panel of 16
countries which recordered simultaneously and successively the current account deficit
and the budget deficit during the period 2008-2018. The empirical evidence is based on
the system Generalized Method of Moment (GMM). To sum up, the results reveal the
existence of the trade protectionism syndrome (an increase in the budget deficit (trigger)
leads to a worsening of the current account deficit and the conjunction of the two deficits
leads to the resurgence of trade protectionism through the rise in tarrif barriers) in SSA.
Thus, reducing the budget deficit and speeding up the industrialization process with
the aim of being competitive in international markets will reduce the trade deficit and
promote free trade.

Disclosure statement
No potential conflict of interest was reported by the author(s).

References
Ahmed, S. M., and M. I. Ansari. 1994. “A Tale of Two Deficits: An Empirical Investigation for Canada.”
International Trade Journal 8 (4): 483–503.
490 G. Z. NKEMGHA ET AL.

Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel Data: Monte Carlo Evidence and
an Application to Employment Equation.” The Review of Economic Studies 58 (2): 277–297.
Arellano, M., and O. Bover. 1995. “Another Look at the Instrumental Variable Estimation of Error-
Components Models.” Journal of Econometrics 68 (1): 29–51.
Baldwin, R. E. 1985. Political Economy of US Import Policy. Cambridge: MIT Press.
Baldwin, R., and S. Evenett. 2009. The Collapse of Global Trade, Murky Protectionism and the Crisis:
Recommendations for the G20. London: Centre for Economic Policy Research (CEPR).
Baldwin, R. E., and C. S. Magee. 2000. “Is Trade Policy for Sale? Congressional Voting on Recent Trade
Bills.” Public Choice 105: 79–101. doi:10.1023/A:1005121716315.
Beaulieu, E., R. A. Yatawara, and W. G. Wang. 2005. “Who Supports Free Trade in Latin America?”
World Economy 28 (7): 941–958.
Blundell, R., and S. Bond. 1998. “Initial Conditions and Moment Restrictions in Dynamic Panel Data
Models.” Journal of Econometrics 87 (1): 115–143.
Caplanova, A., M. Orviska, and J. Hudson. 2004. “Eastern European Attitudes to Integration with
Western Europe.” Journal of Common Market Studies 42 (2): 271–288.
Constantinescu, I. C., A. Mattoo, and M. Ruta. 2015. The Global Trade Slowdown: Cyclical or Structural?
Policy Research Working Paper 7158. Washington: World Bank.
Coughlin, C., K. A. Chrystal, and G. Wood. 1988. “Protectionist Trade Policies: A Survey of Theory,
Evidence and Rationale.” Review – Federal Reserve Bank of St Louis, 12–26.
Denslow, D., and T. M. Fullerton. 1996. “Consumer Attitudes Toward Trade Liberalization.” Applied
Economics Letters 3: 179–182.
Dutt, P., and D. Mitra. 2001. “Endogenous Trade Policy Through Majority Voting: An Empirical
Investigation.” Economic Review 64: 291–303.
Etter, L., and G. Hitt. 2008. “Farm Lobby Beats Back Assault on Subsidies.” Wall Street Journal, March
27, A1, A12.
Evenett, S., and J. Fritz. 2017. Will Awe Trump Rules – The 21th Global Trade Alert Report. London:
Centre for Economic Policy Research.
Gawande, K., and U. Bandyopadhyay. 2000. “Is Protection for Sale? Evidence on the Grossman-Helpman
Theory of Endogenous Protection.” Review of Economics and Statistics 82 (1): 139–152.
Gawande, K., and P. Krishna. 2003. “Political Economy of U.S. Trade Policy: Empirical Approaches.”
In Handbook of International Economics, edited by J. Harrigan, and E. Kwan Choi, 213–250. Oxford:
Blackwell.
Gawande, Kishore, Pravin Krishna, and Michael Robbins. 2001. “Foreign Lobbies and US Trade Policy.”
Manuscript.
Goldberg, Penelopi, and Giovanni Maggi. 1999. “Protection for Sale: An Empirical Investigation.”
American Economic Review 89: 1135–1155.
Gordon, B., and S. Cho. 2008. “The Farm Bill and Free Trade.” Wall Street Journal, May 7, A19.
Guiso, L., P. Sapienza, and L. Zingales. 2003. “People’s Opium? Religion and Economic Attitudes.”
Journal of Monetary Economics 50 (1): 225–282.
IMF (International Monetary Fund). 2016. Subdued Demand: Symptoms and Remedies. World Economic
Outlook October 2016. Washington, DC: International Monetary Fund.
Jouini, J. 2015. “Linkage Between International Trade and Economic Growth in GCC Countries: Empir-
ical Evidence from PMG Estimation Approach.” The Journal of International Trade and Economic
Development 24 (3): 341–372.
Kaempfer, W. H., and S. V. Marks. 1993. “The Expected Effects of Trade Liberalization: Evidence from
U.S. Congressional Action on Fast-Track Authority.” World Economy 16 (November): 725–740.
Kaempfer, W., and T. Willett. 1987. “Why an Import Surcharge Wouldn’t Help America’s Trade Deficit.”
The World Economy 10 (1): 27–38.
Kang, Jong Woo, and Dorothea Ramizo. 2019. “: Impact of Antidumping Measures on International
Trade: Growing South–South Tensions?” The Journal of International Trade & Economic Develop-
ment. doi:10.1080/09638199.2019.1676295.
Krueger, A. O. 1974. “The Political Economy of the Rent Seeking Society.” American Economic Review
64 (3): 291–303.
Lamb, R. 2006. “Rent Seeking in US-Mexican Avocado Trade.” Cato Journal 26 (1): 159–177.
Luttrell, C. B. 1978. “Imports and Jobs – The Observed and the Unobserved.” Review – Federal Reserve
Bank of St Louis, 2–10.
 THE JOURNAL OF INTERNATIONAL TRADE & ECONOMIC DEVELOPMENT 491

Magee, S. P., W. A. Brock, and L. Young. 1989. Black Hole Tariffs and Endogenous Policy Theory: Political
Economy in General Equilibrium. Cambridge: Cambridge University Press.
Mayda, A. M., and D. Rodrik. 2005. “Why are Some People (and Countries) More Protectionist Than
Others?” European Economic Review 49 (6): 1393–1430.
Melgar, N., J. Milgram-Baleix, and M. Rossi. 2013. “Explaining Protectionism Support: The Role of
Economic Factors.” ISRN Economics, 1–14.
Mengyun, Wu, Jabbar Ul-Haq, Naeemuz Zafar, Huaping Sun, and Jingyu Jiang. 2019. “: Trade Liberaliza-
tion and Informality Nexus: Evidence from Pakistan.” The Journal of International Trade & Economic
Development. doi:10.1080/09638199.2019.1593489.
Nicita, A. 2013. “Exchange Rates International Trade and Trade Policies.” International Economics 135
(136): 47–61.
Nickell, S. 1981. “Biases in Dynamic Models with Fixed Effects.” Econometrica: Journal of the Economet-
ric Society 49 (6): 1417–1426.
O’Rourke, K. H., and R. Sinnott. 2006. “The Determinants of Individual Attitudes Towards Immigra-
tion.” European Journal of Political Economy 22 (4): 838–861.
Pincus, J. 1977. Pressure Groups and Politics in Antebellum Tariffs. New York: Columbia University Press.
Scheve, K., and M. J. Slaughter. 2004. “Economic Insecurity and the Globalization of Production.”
American Journal of Political Science 48 (4): 662–674.
Tullock, G. 1967. “The Welfare Cost of Tariffs, Monopoly and Theft.” Western Economic Journal 3:
224–232.
US Current. 2008. US Current Account Deficit. Peterson Institute for International Economics. Accessed
May 23, 2008. www.iie.com/research/topics/hottopic.cfm?HotTopicID 14 9.
World Bank. 2015. Having Fiscal Space and Using It. Global Economic Prospects January 2015. Washing-
ton, DC: World Bank.
WTO (World Trade Organization). 2016. Report on G20 Trade Measures. Geneva.

Appendices

Table A1. List of countries.

Benin Ghana Mauritus South africa
Burkina Faso Kenya Mozambique Tanzania
Burundi Madagascar Senegal Togo
Ethiopia Mali Sierra Leone Uganda
492 G. Z. NKEMGHA ET AL.

Table A2. Variable definitions.

Variables Signs Variables definition (measurement) Sources
Manufacturing Manuf Manufacturing, value added (% of World Bank (WDI)
GDP)
Gross Domestic Product GDP Gross Domestic Product (GDP) (annual World Bank (WDI)
%)
Tarrifrate Tarrif Tarrif rate, applied, weighted mean, all World Bank (WDI)
product (%)
Employment in industry Indempl Employment in industry (% of total World Bank (WDI)
employment)
Budget deficit BD primary deficit which is the difference IMF
between public revenue and public
expenditure, apart from servicing
the external debt.
Terms of Trade Tot It is the quotient of the export price World Bank (WDI)
index divided by the import price
index, with a reference year
Inflation INFGoveff Consumption price index World Bank (WDI)WGI
Government Effectiveness Government Effectiveness captures
perceptions of the quality of
public services, the quality of the
civil service and the degree of its
independence from political pressures,
the quality of policy formulation and
implementation, and the credibility of
the government’s commitment to such
policies. Estimate gives the country’s
score on the aggregate indicator, in
units of a standard