Trade and Exchange-rate Reforms in North Africa
179
effective exchange-rate depreciation (appreciation). Volatility is computed, for a given
year, as the standard deviation of monthly changes in the real effective exchange rate.
Two measures of misalignment are considered: the black market premium and a
5
measure based on the Edwards (1988) models. The computation of these variables is
6
the same as in Sekkat and Varoudakis (2000).
The indicator of trade liberalisation is taken from Richaud and Varoudakis (1999).
They used the average tariff on imports (duty and other specific taxes) and tariff
dispersion and exemptions to classify tariff schemes as restrictive, relatively restrictive,
moderate, relatively open and open. Similarly, non-tariff barriers on imports and
exports, including quotas, bans, licensing schemes, foreign-exchange allocation,
government control on export and import marketing (state monopolies, etc.), were used
to classify trade regimes as restrictive, moderate and open. Finally an overall index of
trade liberalisation achievements ranging from 1 (the most closed) to 10 (the most open)
was computed (see the Appendix).
As shown in Table 1, Morocco and Tunisia display a slight but steady trend of real
effective exchange-rate depreciation. In the 1990s the REER seems to have been
stabilised in the case of these two countries. In contrast, Egypt experienced a sharp real
depreciation following devaluation in 1979, and then continuous real appreciation up to
the 1990s.
The average level of volatility is substantially higher in Egypt than in the other
countries in every sub-period. However, all three exhibit increasing REER volatility
over the period. This can be the joint outcome of the move towards more flexible
exchange-rate mechanisms throughout the 1980s and the internal inconsistencies of
macroeconomic policies which failed to deliver stable economic conditions.
During the 1980s, misalignment as measured by the black market premium was
substantial in Egypt, but low in Morocco and Tunisia. Looking at the model-based
measure of misalignment, North African countries show a pattern of declining REER
misalignment up to the mid-1980s. In the later part of the decade they exhibit an
increase in REER misalignment (which was substantial in Egypt). However, in the
1
990s misalignment decreased in Tunisia and Morocco while remaining substantial in
Egypt.
The trade policy index suggests that the three countries had restrictive trade
regimes (Trade=1) during the 1970s and the 1980s. Morocco seems, however, to have
liberalised earlier (Trade=3 in 1988) than the others. Based on achievements in trade
reform by the end of the period, Egypt still appears as a slow reformer (index lower than
4
4
) while Morocco and Tunisia can be considered as moderate reformers (index between
and 6).
5
6
. The black market premium is used as a crude measure of real exchange-rate misalignment. However, it
may also capture the influence of other distortions in the foreign-exchange market (see Pinto, 1990). The
measure of misalignment first suggested by Edwards (1988) is considered in the literature as more reliable.
The approach distinguishes among changes of REER due to changes in external or domestic
‘
fundamentals’ which bring about changes in the equilibrium REER (international terms of trade,
international transfers, technological progress) and misconceived domestic policies which may create
policy-induced misalignment’ (excess domestic credit creation, excessive foreign borrowing, excessively
‘
inward-oriented trade policies).
. To save space it will not be presented here. The series is available upon request.