Cotton, War and Growth in Chad (1960-2000)

Cotton, War and Growth in Chad (1960-2000)
Chapter 3 of volume 2

Jean-Paul Azam[1] and Nadjiounoum Djimtoingar[2]





Table of Contents


1. Introduction 1
2. Cotton-Led Sahelian Growth in Chad 3
3. The Three Main Phases of the Chadian Growth History 9
4. The History of the Cotton Sector 13
5. The Political History 18
6. Conclusion 27
Appendix 1 : The Johnson-Brecher-Diaz-Alejandro Immiserizing Growth Model
34
Tables 37
List of Tables

Table 1: GDP Growth Rate of Chad at Local 1987 Constant Prices
Table 2: The Impact of Cotton on GDP Growth
Table 3: Growth and Inflation in the Wake of the 1994 Devaluation
Table 4: Exports and Cotton Exports in CFA F Billion
Table 5: Revenues from the Export Tax on Cotton Fiber during 1985/1986
Table 6: Distribution of Health Infrastructure between the North and the
South in 1966
List of Figures

Figure 1: Actual and Fitted GDP Growth Rate (Constant Local 1987 Currency)
Figure 2: Tracking of GDP Growth by Equation (8)
Figure 3: Growth Rate of Per Capita GDP at Constant International Prices
Figure 4: Per Capita GDP at International Prices (Log Scale)
Figure 5: Comparison of Per Capita GDP at Local and International Prices
(Log Scale)
Figure 6: Cultivated Area and Cotton-Seed Output
Figure 7: Total Exports and Cotton Exports (in CFA F Billion)
1. Introduction

Chad is a small Sahelian economy of 7.7 million inhabitants in 1999, which
is landlocked, partly a desert, and whose political life is dominated by
the perennial rivalry between the northern semi-nomadic Muslims and the
southern Animist and Christian farmers. Like many other African countries,
Chad has a highly polarized society, where identity and differences are
defined along ethnic, cultural, religious, and economic lines. The
northerners speak an Arabic dialect and, in their majority, rejected
western education during the colonial days and afterwards. The southerners,
in contrast, seized what they saw as a unique opportunity to develop, and
speak Sara and French. The northerners live on cattle raising and other
livestock, as well as on trade and smuggling with neighboring Algeria,
Libya, Sudan, and Egypt (via Sudan) in the north, as well as with Cameroon,
the Central African Republic, Niger and Nigeria in the south. The
southerners produce the main export crop, cotton, and own most of the
productive human capital; and their land contains huge oil reserves, which
have been known since the mid-1970s, and began to be exploited starting in
2003. At that point, Chad switched from being a poor landlocked country to
becoming a resource-rich country. However, this change in economic
opportunities was partly endogenous, as shown below. It is only after peace
was secured, and institutions were put in place to make it credible, that
foreign oil companies ventured into exploiting the oil.
Until then, the growth experience of this country was dominated by the
cotton sector, often disturbed by civil war, and complicated by the Libyan
intervention. This is brought out strikingly by section 2 below, which
shows econometrically that, after controlling for shocks due to political
violence, growth in Chad simply reflects the evolution of growth in other
cotton-driven Sahelian economies. These economies are quite similar to one
another, all belonging to the CFA Zone and dominated by the cotton sector.
The cotton sector was dominated for most of the period by a French company,
the CFDT (Compagnie Française des Textiles). Hence, a large component of
the growth experience of Chad can be regarded as determined exogenously,
reflecting only the ups and downs of the Sahelian cotton economy. In the
case of Chad, this can also be checked econometrically by a direct analysis
of the impact of the output of the cotton sector on the whole economy,
which is also performed in that section. Cotton also provides the
transmission channel through which the 1994 devaluation of the CFA Franc,
an event that owes nothing to the Chadian history itself, impacted the
country's growth path. The devaluation triggered a cotton-led export boom
that pulled the economy upwards, for a while.
Political shocks provide the idiosyncratic features of the Chadian
growth history. The most important one occurred in 1979-84, comprising of
three years of outright civil war followed by two years of near-genocidal
massacres perpetrated in the cotton-growing area. This is by all accounts a
watershed in Chadian history. Section 3 shows that the pattern of growth is
different before and after this period. Hence, a natural periodization of
Chad's economic history emerges, as war draws a clear separation between
two contrasted growth episodes. There was a gradual build-up leading to the
outbreak of the civil war, when GDP per capita was on average decreasing.
This superimposed a downward trend on the cotton-led growth path between
independence in 1960 and the outbreak of the civil war, reflecting the
pessimistic expectations of potential investors. This may be called the
expected disaster episode, when the shadow of conflict loomed large over
the Chadian economy. Potential investors were anticipating the probable
breakdown of the post-independence political equilibrium, with potentially
dramatic redistribution of property rights and political power. The post-
conflict period shows a different profile, with mere stagnation or a slight
recovery, at a significantly lower income level than before. The new
political equilibrium only emerged with a lag, after the civil war ended.
While the 1980s were still highly unstable and violent, including an
attempted invasion by Libya, Idriss Déby imposed a lasting settlement in
1990. It was based on a degree of power-sharing and a relatively credible
framework for sharing the benefits of economic growth between the two ethno-
regional groups.
As cotton and war will have been shown in sections 2 and 3 to be the
crucial determinants of Chad's growth history, the subsequent two sections
will describe in greater detail the history of the cotton sector (section
4) and the political history that led, through the civil war, to a change
in power from the (relatively) rich and educated southerners, who ran the
country just after independence, to the poorer northerners, after the
turmoil period of the war (section 5).


2. Cotton-Led Sahelian Growth in Chad

A large part of the growth experience of Chad is simply shared with other
Sahelian economies. This comes out clearly in Figure 1. As mentioned above,
we take a regression-based approach to identifying the relevant growth
episodes in Chad, using only variables that can reasonably be regarded
either as exogenous to Chad's growth process, such as the growth experience
of other Sahelian countries, or at least as unquestionably predetermined,
like the outbreak of political violence. The latter might arguably be
regarded as influenced to some extent by the Chad's cumulated growth
experience over a distant past.


***insert Figure 1 near here****

***insert Table 1 near here****

Figure 1 shows that a remarkable tracking of the growth rate of this
economy over the 1970-1996 period can be achieved by using a very simple
econometric equation, presented in column 1 of Table 1. Figure 1, and
column 1 of Table 1 from which it is derived, show clearly that the growth
rate of GDP in Chad, measured at constant 1987 local prices, can be tracked
rather closely by taking into account only the growth experience of Burkina
Faso and the various episodes of political violence. The series comes from
World Bank (1998). Although episodes of fighting or massacre have occurred
in Chad from 1965up until almost the present, we have identified 1979-1984
as the most acute period of civil war in constructing the dummy variable
'war' used in Table 1. In fact, the years 1983-84 witnessed more a series
of massacres perpetrated in the south of the country than a civil war
proper. Similarly, Libya occupied a portion of Chad, the Aouzou stripe just
across the border, beginning in 1973; but it was only in 1986 that the
Libyan army launched a massive attack in the center of the country. This
triggered the victorious counter-attack of 1987, which removed all Libyan
presence from Chad. Hence, the 'warlib' dummy indicates the years 1986-87.
Some killings of southerners have taken place ever since the outbreak of
the civil war in 1979. But they took another, much larger, dimension in
1982-84, after the victory of the northerners in the civil war, under
Hissein Habré, and again in 1993, under Idriss Déby. During that year of
political trouble, the 'National Conference' took place for venting popular
resentment (see Buijtenhuijs 1998). The two episodes of massacre are
captured by the 'repsouth' dummy variable.
The estimated equation has a simple implicit error-correction
specification à la Banerjee et al (1993), with the GDP of Burkina Faso
being included both as the current growth rate and as the lagged level, in
logarithm, while the lagged value of Chad's GDP is also included. Hence,
these equations are based on the so-called 'one-step Hendry' method for
estimating error-correction models and testing co-integration. As Banerjee
et al (1993) show using Monte-Calro simulation, this method is clearly
preferable to the more common 'two-step' Engle-Granger approach (Engle and
Granger 1987) when using a sample size below 200 observations. In this
case, the two-step method is liable to produce highly biased estimates.
This equation suggests that GDP in Chad and Burkina Faso are co-
integrated. This is confirmed by a direct application of the Johansen
cointegration test, including War, WarLib and RepSouth as exogenous forcing
variables. The likelihood ratio found for this test is 1.54, to be compared
with a 5 % critical value of 3.76. The co-integration equation reads:
LyChad = 1.74 + 0.59 LyBur. Notice the small slope of this equation, whic