Explaining Sudan's economic and political turmoil

Pour Micbile AFRICAN ECONOMIES AND THE POLITICS OF

PERMANENT CRISIS, r979-r999

NICOLAS VAN DE WALLE Michigan State UniversitY

CAMBRIDGE UNIVERSITY PRESS

/

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge cBz 2RU, uK 4o'West zoth Street, New york, Ny roorr-4zrr. usA ro Stamford Road, Oakleigh, vtc ar66, Australia Ruiz de Alarc6n 13, z8or4 Madrid, Spain Dock House, The l7aterfront, Cape Town goor, South Africa http://www.cambridge. org

@ Cambridge University press zoor

This book is in copyright. Subject to statutory exceprion and to the provisions of relevant collective licensinj agreements, no reproduction of any part may take place without the written permission of Cambridge University press.

First published zoor

Printed in the United States of America

Typeface Sabon rolr3 pt. Sy st em QaarkXPress [BTS]

A catalog record for this book is auailable from the British Library,

Library of Congress Cataloging in pwblication data

Van de ti(alle, Nicolas, 1957- African economies and the politics of permanent crisis, 1979_1999 / Nicolas van

de'Walle. p. cm. - (Political economy of institutions and decision)

Includes bibliographical references and index. r s n N o - 5 z r - 8 o 3 6 4 - o - r s r N o _ 5 z r _ o o g 3 5 _ o ( p b . ) r. Africa - Economic conditions - 196o_ i. Atrir^ _ Economic policy.

3. Afuica - Politics and government. I. Title. II. Senes. HcSoo .v3 j7 zoor

338.96-dczt

ISBN o 5zr 8q64 o hardback ISBN o 5zr oo836 o paperback

z o o r o 2 S r T A

List of Tables Acknowledgm

N,- $ C( S

Patterns in Reform Implementation, rg79-r999

The Zambian government first committed itself to civil service retrench- ment in rgTg,inresponse to growing fiscal deficits and donor pressures.t

Fairly reliable government statistics from the Monthly Digest of Statis- tics indicate there were rz9,6oo public employees that year, in addition to 136rzzo parastatal employees. The government's stated commitment to eliminate up to half of the civil service was praised by the \ilorld Bank

and heavily criticized by the ZCTU, the Zambian Trade Union Federa-

tion. A reform package was not put in place until ry86 and it resulted

in minor staff removals, through a lowering of the retirement age. In May ry88, a cabinet office task force on restructuring the public service submitted a new action plan calling for a z5 percent reduction in the

civil service. The plan was not implemented, but in August r99o the gov-

ernment contracted a local consulting firm to develop another ambitious retrenchment exercise, under the donor-financed Public Service and Retrenchment Project. Reform efforts continued throughout the r99os, though to little effect. In November t993,the government launched the Public Service Reform Program, vowing to cut the civil service by z5 percenr within three years, again with support from the donors. The

United Nations Development Program (UNDP) financed a team of experts to oversee a plan for retrenchmentr reorganization, and upgrad-

r. This account of civil service reform in Zambia is taken from the 'World

Bank, Public Expenditure Reuiew of Zambia, Report No. u4zo-ZA (Ifashington,

DC: ITorld Bank, December r, r99z); and Nicolas van de Walle and Dennis Chiwele, "Democratization and Economic Reform in Zambia," MSU Working Papers on Political Reform in Africa, no. 9, Michigan State University, 1994. I

also thank Malcolm McPherson for an immensely enlightening personal commu- nication on these issues, but I absolve him of any responsibility for the account given here.

Patterns in Reform Implementation, t979-t999

ing of the civil service. critics of the government actually blamed civil

serrrice retrenchments for the rise in unemployment'2 'what

was the impact of these various reform programs? The truth is

that we do not really know. The government stopped reporting credible

employment statistics in the early rg8os, and the different numbers that

are cited in various donor and government reports are contradictory' The

zambian central Statistical office reported r52,4oo "central and local

government employees" in ry93.3 IFI reports suggest some r3o'ooo

"central government employees" in the mid-r 99osra while a report by

theZambianPrices and Incomes Commission indicated rx',,61o "public

sector employees', in r99ors and a more recent government report esti-

mates at ta4,47s the number of staff in "public administration."6 There

have been well-publicized layoffs in the r99os) but also evidence of

steady new hiring. various accounts suggest a problem of ghost workers,

so rhat ar least some of these totals are fictitious. sifting through the dif-

ferent numbers suggests a back-of-the-envelope guestimate of r4o'ooo

people on the government payroll in the mid-r99os, and argues against

it "

id." that the reform era has seen a sharp drop in public employment,

but the picture is muddled at best.

.Mani observers- of ' f i"a appear-to--take--as-lacrrlaf-rUuctuai*

adffiI-t r.r.rltqd in un **ot,It o' F,thlicernplpyment d'rring the:gi&os'

Ulg-:22glt *deed, as we shall see, in a small number of countries' there

T"*-f.ffrignificant civil service retrenchments. Yet, Zarnbia is far from

unique i' tlis situation of intense reform activity and costly donor

,oppor, probably yielding few retrenchments' in the midst of poor sta-

tisiics and ambiguity abour actual numbers. Is civil service reform

z. See Peter Henriot, ,,Effect of Structural Adjustment Programmes on African Fam---

ili;r,' (Lusaka, Zambia: Jesuit Center for Theological Reflections, April 1995)'

3. Personal communication, Malcolm McPher-son'

1. For instance, Ian Lieneriand Jitendra Modi, "A Decade of Civil Service Reform

-' l" i"f_s"rr" ran Africa,,, IMF rqorking

paper wplgTh Tg (washington, Dc: Inter-

national Monetary Fund, Fiscal Af{airs Department, December-ry97\' p. 43-

s. ,.Reoort on Development of Incomes of Unionized Formal Sector \rorkers,'' tisi-tuut" (Lusaka, Zambia: Prices and Income Commission' May'

r99z)'

Table 1. , F 6 . E c o n o m i c R e p o r t , t g g 6 ( L u s a k a , Z a m b i a : M i n i s t r y o f F i n a n c e a n d E c o n o m t c

D e v e l o p m e n t , J a n u a r Y 1 9 9 7 ) , P . i z '

z. i" ,i.ii.."i.# of adi,-tst-.", ln tttt lecigl: Thandika Mkandawire and Charles

' S.f"a" argue that

,'retrenchment in thJcivil service has played an important role

i, ,*.ffi"i the ranks of the poor" throughout Africa. See p.T3 of their otherwise

excellent book, Our Contiient, Our Fiture: African Perspectiues on Structural

Adjustment (Trenton, NJ: Africa lforld Press, 1999)'

6 \

L

64

African Economies and the Politics of Pertnanent Crisis, r979-tAfrxca.n Lconomtes and tne rouucs ol feftnanent Lrrsxs, 1979-1999

unusual in this respect? How much policy reforr4 has actually occurred in Africa during the last fwenty years?

'What kinds of policy biases

have governments demonstrated during this period of austerity and fiscal crisis? Are there patterns across the region in terms of which policies have been modified and which remain largely as they did twenty years ago? Are the statistics always as deficient as in the case of public sector employment? At the outset of our analysis it is important to examine the empirical record carefully, given the heated rhetoric that has surrounded debates on the topic of structural adjustment and has . often obscured more than it has revealed. Twenty years of continuous fiscal crisis was bound to bring about a degree of de facto policy change, so we need to focus on actual government choices and areas in which the state appears to have acted in the context of a discernible political strategy, In facf, although there is variation across the economies of the region, a clear pattern of partial reform does emerge, with much progress in some areas, but little in others. Most strikingly, I show that o.srall 4r19llgrent....c-anssqLp-!isq haa*aqfuallnjas{ealsd*d*4td_t-s ,"!qjlt"-e9eg!L-9y-e:_ar-Sggo-v,g-qtsp,$!as"s@ !o S.lggifisantly...pare-d-sJvn-its development,.e-ffe{t, Thanks in part to substantial donor support to state strucfures, twenty years of crisis have resulted in a bigger state that does less for its citizens, particularly its poor and rural ones. These patterns are fully compatible with this book's thesis that lbe-prigr:r-s*ol-scqeasigps.lie,aare to be fgg+d.lglgi" lhe state..ald-rog,.as.-us-ually,p-qtireil.in rpersralercscu{e-s- .os*deei$jon makers.

This chapter begins the process of presenting evidence on behalf of the hypotheses advanced in the previous chapter. The rest of this chapter is divided into three sections. First, it is necessary to establish clearly the nature of the empirical record regarding policy outcomes in Africa. The next section examines the record of implementation of policy reform during the last twenry years. There Lu.s*bee.n-uslg progress en qtrbilizatien tolieics than-on".the-naore.complex-insuruflonal reforms, of strucJrrral-adtu*neq1-bUfgush-Af--tLe,,p.1ogress aFpea'q vr,l- neLable - ro*r-plrgf,$-al-and,,manipulatiorr-.hy.-state-ctlees.- A third section then turns to a political interpretation of the evolution of public spend- ing. I provide evidence to cast doubt on the argument that state elites have responded to fiscal austerity in ways that show they are beholden to specific societal interests. Instead, governments in fhe region have managed the reform process to protect the state expenditures that serve elite interests.

P atterns in Reforrn IrnP lementation' r 9 7 9- t 9 9 9

R E F O R M I M P L E M E N T A T I O N : H O \ X / M U C H ?

H o \ r s u s T a l N A S r n ?

How much has the policy environment actually changed during the last

twenry years? To believe borh rhe IFIs and their critics, policy reform has

been comprehensive: the'world Bank's latest report on the region, pub-

lished in mid-zooo, suggests that "substantial reforms contributed to the

resurgence of growth in the second half of the l99os'"8 On the other

hand, critics such as Mkandawire and Soludo also assert that "African

economies have been subjected to dramatic feformsr" but they argue that

these reforms are to be blamed for the region's "exceptionally poor [eco-

nomicl performance" during the r99os.n

In fait, I will argue that the policy environment has changed less than

is commonly argued. Belying the claims of its own front office, the \rorld

Bank has issued evaluation feports that estimated at only 6o percent the

implementation rate of the reform measures included in its structural

ndi.rrt-..rt and sector adjustment loans during the r98os for all less

developed countries.lo The effective implementation rate is probably

lower, ,in." at least some of these conditions involved policies the gov-

ernment would have undertaken in the absence of a Bank loan, while

others did not involve substantial difficulties - for example, sector

studies. Moreovero the Bank project management staff, from whom these

reports draw their findings, have every incentive to exaggerate the degree

of iropl"*"rrtation. Independent studies reveal a lower level of imple-

mentation. The most thorough data about implementation come from

Tony Killick. In one study, Killick surveyed 3o5 IMF programs in less

developed countries between -'979 and ryg3 and found that 53 percent

had not been completed during the loan period.11 In another study,

on \World Bank adjustment loans, he found that 75 percent of the

8. The world Bank, can Africa claim the zr" century? (I0-ashington, DC: Ifrorld Bank, zooo), p. 28. In a now well-established pattefn, the repoft.actually con-

sists of "

a.iuit.d diagnostic of the region's ptobl.-t and ends with a renewed

call fol further PolicY reform.

9. Mkandawire and Soludo, Owr Continent, Oyr Fwtwre, p' Bt'

t;. a;; *orld Bank, Adjustment Lending: An Eualuation of Ten Years of Experi'

ence, Country Economics Department, Policy, Planning, and Research (I(ash-

i.rgton, DC, \iqorld Bank, r98i1, Tables 4-7 and 4-4.The implementation rate is

.uin lo*., for the low-irrco-me countries that predominate in sub-Saharan Africa'

rr. Tony Killick, IMF Programmes in Deuelopi.ng countries: Design.and Impact

(London: ODI, x996). i<illick defined a program as uncompleted if the country

irad failed to impi"m.trt zo percent or more of rhe program's conditions.

66 t 3 /

African Economies and the Politics of Permanent Crisis, rg79-rg9g

disbursements of the second tranche of adjustment loans were delayed more than twelve months, suggesting a widespread problem of noncompliance.12

If implementation of policy reform has been less than complete, the obvious question is this: what patterns and bialses have characterized the reform process? There is clearly much variation across the forty-eight African economies, but extremely suggestive trends in implementation can be identified from a careful analysis of the different components of most reform programs. I structure the following analysis by making use of the standard distinction between stabilization-Wtlicies, which seek tq

-r_-e_g!ore m49I99EarcmiEbala!_c€"tl$.grttgr.l!9_!q-e"diumie$, and adiust- ment policies, which seek to alter the basic gg,onomie,igsdlulions of the country to foster hifi[er glowth in1h. -.ii"m to long re-9q1.

' ' . . . j - ^ * - - - _ : : : t = j - Y E . # - '

St ab ilization P o Ii ci e s

The majority of African economies have made substantial progress on implementing basic stabilization policies. These are designed to restore macroeconomic stability in the short term, a goal typically at the core of IMF programs.l3 Significant progress had been made on cutting fiscal deficits, which declined on average from well over ro percent (exclud- ing grants) at times in the r97os and r98os, to 9 percent of GDP in 1992, and to 4.5 percent in t997.14 Current account deficits have under- gone a similar improvement and were onlS on avetage,4 percent of GDP in ry97. Exchange rate policies have perhaps undergone the most dra- matic improvements. In the countries outside of the Franc Zone, repeated devaluations during the r98os and movement toward a more flexible exchange rate system has resulted in a definite trend toward more rea- sonable exchange rates. In the Franc Zone, in which currencies are pegged to the French franc, the devaluation of January 1994 served to bring down what had been typically fairly overvalued exchange rates. Thus, if between ry7 5 and 1984, eighteen of the region's economies had

rz. Tony Killick, Aid and the Political Economy of Policy Change (London: Rout- l e d g e , 1 9 9 8 ) , p . 3 o .

13. Tony Killick, IMF Programmes in Deueloping Countries, 14. See Stanley Fischer et al., Africa: ls This the Turning Pointi IMF Papers on Policy

Analysis and Assessment ('S(ashington, DC: International Monetary Fund, May 1998), p. r. For a broader assessment of fiscal adjustment, see Jayati Datta-Mitra, Fiscal Management in Adjustment Lending, a l0orld Bank Operations Evalua- tion Study (lfashington, DC: \forld Bank, ry97).

Patterns in Reform lmplementation, t979-t999

parallel exchange rates at least 5o percent higher than the official

"*.hung. rate, by r995-96, such a black market premium existed only

in three countries: Guinea-Bissau, Nigeria, and 56o Tom6 and Principe.l5

These continental averages disguise large intercountry variation, of

course, but they do imply a significant amount of progress on macro-

economic policy throughout the region. [t is nonetheless important to

put this progress in perspective. First, much of the progress has been

quit" ,...rrt, occurring in the r99os rather than the r98os, during which

there was remarkably little progress on much of the reform front. In a

sense, the delay in stabilization is striking given Africa's situation. Faced

with extfemely high fiscal and balance of payments crises from the late

rgTas on, and generally unable to raise significant capital on private

markets, African governments had little choice but to borrow from the

donors, despite their policy conditions. \fithout the option of printing

money,16 various pressures would probably have induced equilibration

of deficits in time, regardless of government policy and with or without

donor pressures to do so. As I will argue in Chapter 5, donor finance

and the ability to accumulate debt rapidly has served to delay this

inevitable adjustment to fiscal realities.

The slow rate of initial progress is true even of the countries usually

perceived today as success stories: the current star pupil of the interna-

iional financial institutions, Uganda, provides a good example. Lax mon-

etary and fiscal policies through the early rggos resulted in an inflation

rate ofover 50 percent in that country as late as r99z' In the same vein,

Kenya,s fiscal de{icit - in rggS a sterling 2 percent of GDR according to

the'World Bank17 - had been rr.8 percent of GDP as recently as 1993.

In French west Africa, the devaluation of the CFA franc inJanuaty 1994,

followed by several good rainy seasons, were the key events that account

for the present apparent successful stabilization' As a result, in most

countries, it is too early to tell whether the progress will be sustained.

r5. calculated from vodd Bank, African Deuelopment Indicators, zooo (washing'

' to.r, DC, Iforld Bank, zooo), p. 5r. Unless otherwise indicated, the statistics

attributed to the World Bank thloughout this chapter arc from this volume.

16. It is interesting that with the notable exception olzaire, African govelnments on

the whole did not resort to the expedient solution of ptinting money. In the Franc

zone, strict rules apparently served as an effective deterrent. Governments appear

to have been dissuaded from doing so by the examples of disastrous hyperinfla-

tion elsewhere, notably in Latin America.Perhaps as important has been the influ-

ence of extensive "ni

long-rrunding technicaf assistance to central banks and

finance ministries from the IMF, thJFrench treasury, and other-Western donors'

rz. African Deuelopment Ind'icators, p. r9o'

6 8 69

African Econcsmies and the politics of permanent Crisis, r979_r9g9

second, progfess on macroeconomic md.nagemenf remains vulnerable to reversals. By their nat,.ue, most stabilization measures are the policy reforms that are the easiest to achieve in the short run, but also the hardest to sustain because they are the easiest to undo. This is particu, larly true for measures that are, in effect, the subject of annual decisions. For instance, balancing the budget this year does not necessarily make it any easier to balance it next year, or less necessary to do so to main- tain macrostability. In countries like Ghana, Kenya, or Gabon, past progress on stabilization was abruptly jettisoned in the months before an election by a government eager to hold on to power. Thus, Ghana's budget deficit went from 4.9 percent. of GDp in r99r to rz.7 percent in 1992, an election year in which Rawlings faced the voters for the first time. coupled with an unexpeced decline in oil revenues, omar Bongo's campaign to refain the presidency in Gabon in x999 led to a mush_ rooming of the deficit from z percent of GDp ro an estima ted 3opercent as he stoked public expenditures to ensure political support.ld similarlS the positive impact of the devaluation of the cFA frutt. t' Francophone west African economies will inevitably abate; even with inflation rates comfortably in the single digits, these counrries typically have higher rates than those prevailing in the \7est, so that real, effective exchange rates are almost sure to appreciate, slowly but surely, until the countries are faced with another competitiveness crisis.le

Third, much of the recent improvement in the basic macroeconomic picture is probably not sustainable without large amounts of external suppoft. Balance of payments and fiscal deficits often remain too large to be sustained without external assistance or rcgular debt forgiveness and rescheduling. Thus, only nine countries in the region hacl fiscal deficits of under 3 percent of GDp in both ry97 and, ti9g, if exrernal grants are not taken into account - a level that can be viewed as a bench- mark of prudent fiscal management in countries with limited access ro

18. calculated from the Economist Intelligence rJnrt's country Report: Gabon, Second Quarter, x999 (London: Economist Intelligence Unit).

19. A recent assessmenr of rhe cFA Zone is proviled in Bruno cabrillac, ,,La situation macro6conomique des pays AfricJins de ra zone Franc i la fin de I'ann6e ry98," Afrique Contempiraine rg9 Gggg), z3_9. More generally, on the operation of the zone, see Jean clement-ei al.,"iftermath'ii iii ceg Franc Deualuotion, r'ternationar Monetary Fund occasional paper # r3g (rgg6l; and the earlier essay by Nicolas van je'walle, ,.The Decline o'f ,h. F.".r. Zone: Monetary Politics in Francophone Africa,,' African Affairs 9o (r99r1, 3 8 3 - 4 o s .

Patterns in Reform Implementation, r979-r999

international investment and low levels of savings.2o In a number of

countries, progress on controlling government expenditures has been

achieved thanks to high levels of donor support, creative accounting and

the deferring of core expenditures into the future. There is something

profoundly disingenuous about donors and governments claiming

progress on fiscal balance when it is taking place in the context of aid

flows totaling a tenth of GDR usually with a pretty high loan compo-

nent. In several countries, significant debt relief largely explains ostensi-

ble improvements in macro-indicators in recent years. But this can only

be temporary if the underlying balance of payments and fiscal deficits

have not been overturned - unless one is willing to make the unreason-

able assumption that the current high aid flow is permanent.

In order to remain in good standing with the donors, moreover' gov-

ernments commonly resort to various budgetary manipulations' A

Iavoite ploy is to accumulate arreafs on civil service wages or other

obligations. ln tggg, reports from Francophone Africa indicated that

Gabon had accumulated seven to eight months of salary arrears since

1995, while these arrears had reached eleven months in Niger, six

to eight months in Togo, and up to eighteen months for some civil

servants in Congo-Brazzaville,zl Elsewhere, Guinea-Bissau's salary

arrears were six to eight months in early zooa,22 while Nigeria has

routinely run several months of such arrears for its state employees in

the last couple of years.23 Such salary arrears are particularly pronounced

in the Franc Zone countries, given strict rules limiting government

financing of deficit spending. Because civil service wages are invariably

the biggest single item in the budget, including these financial obligations

zo. African Deuelopment Indicators,Table 7-2, The countries are Cameroon, C6te d;Ivoire, Equatorial Guinea, Kenya, Mauritania, Senegal, Sudan, Swaziland, and Tanzania. Note that only Mauritania and Swaziland had also enjoyed deficits under 3 percent as recently as 1995.

zr. On Gaton, see March 6s Tropicaux, April r 6, 1999, P' 7 97 i affeats are reported for Niger in "Lassitude Pre-Electorale," in I'Autre Afrique 99 (October tz, 1999):

4o-r; in Togo, "Larm6e sans le sous," in Lettre du Continent, October 14, 1999, p, 3; and itt Cotgo Brazzaville, "Brazzaville Mortuary

'Workers on Strike,"

Panafrican News Agency' December 3r, 1998, zz. "lllection de Kumba Yala traduit une volont6 de changement," Agence France

Press, January zz, zooo. 23. See the following stories from Nigerian newspapers, all found or the Internet:

"Govetnor Seeks Advice," P.M. News,June 7, 1999; "Nigeria; Minimum \ffage: Enugu, Oyo Resolve V'age Crisis," This Day,August 3, zoooi and "Nigeria; IThy States Are Boiling," Vanguard Daily, J:une 24, zooo.

7o 7 r

African Economies and the Politics of permanent Crisis, rg79-r999

in the deficit would significantly worsen the fiscal situarion. yer the IFIs typically look the other way ro keep adjustment programs officially on track.

some oil exporrers like cameroon and congo have similarly sold oil forward several years into the future, despite strict IFI rules against the practice. According to the \florld Bank's statistics, congo's fiscar deficit in the r99os averaged ro.8 percent of GDp between r99o and 1996, excluding donor support, but the precise meaning of these numbers is unclear when rumors suggest that the government had sold forward five years' worth of national oil production - thought to represenr well above go percent of total govefnment revenues - in secret agreements wifh private oil companies,2a countries like Tanzania, Uganda, and zambia have adopted "cash budgets" in the r99os) whereby expenditures are monitored and strictly controlled throughout the year relative to avail- able revenues, in order to achieve abalancedbudget.2j cash budgets have been widely praised for the fiscal turnaround in these countries. But in all three countries, the practice has led to the accumulation of arrears, as line ministries pay for goods and services with promissory notes. In Zambia in 1997, the Ministry of Finance itself admitted that arrears accounted for the equivalent of r.5 percent of GDp.26 Reports suggested that the transport and communications secror alone totaled debts tf over roo billion kwacha to private contractors in late t999, whlle the official capital expenditures budget that year for the sector amounted to only 66.5 billion kwacha!27 Typically, rhese arrears are nor included in fiscal statistics,

24. These sales have been widely reported. see, for instance, .,Les, Milliards en I'Air du Congo," inl'Autre Afrique, May 19, r99g,99, which mentions rhe conserv- ative estimate o{}.s-3.2s billion French francs' worth of sales to oil companies on production through zoo4. see also various issues of the Economist Intelli, gence.Unit Country Reports on Congo in the mid-r99os.

25. David Stasavage and Dambisa Moyo, "Are cash Budgets a cure for Excess Fiscal Deficits (and at \(hat cost)?" center for the study if Afri.".r Economies, Uni- versity of oxford, Paper r/PS/99-rr, May 1999; and, Bruce Bolnick, "Establish- ing Fiscal Discipline: The cash Budget in Zambia," in Meriree s. irindle (ed.), Getting Good. Gouernrnent: capacity Building in the public sector of Deuelop- ing countries (cambridge, MA: Harvard Institute for Internationai Develoo- m e n t , 1 9 9 7 ) , p p , 2 9 7 - 3 3 2 .

26. Stasauage and_Moyo, "Are cash Budgets a cure," p. 13.My own interviewing of officials and businessmen in Lusaka in May oi .999 r"gg.sts this figure may be a significant underestimate.

27. see."zambia: .Gearing Up for Growth," in African Reuiew of Business and Tecb-

nology, March r7, zooo. One U.S. dollar bought roughly z,ooo kwacha ar fhe time.

Patterns in Reform Implementation, 1979-1999

Fiscal adjustment has been held back by the stagnation in revenue

levels that has characterized the adjustment pattefn of most African

economies. The impact of policy reform programs on government rev-

enues is the subiect of some disagreement. lawaniah and Branson find

that state revenues declined ftom t8.7 to r7.7 percent of GDP during

the course of Bank adjustment programs in their sample of Africa coun-

ffies,28 This accords with the findings of several studies that tax revenues

and level of foreign aid have had a significant negative correlation.2e On

the other hand, the IMF has usually argued that its own stabilization

operations have not resulted in systematic changes in the government's

tax effort.3o No one, on the other hand, has argued that the last two

decades have witnessed significant increases in government revenues,

despite the IMF's stated intention in virtually all of its adjustment opef-

ations to increase them.'World Bank data show that African government

revenues have stayed around an average of 19 to 2-o percent of GDP. As

shown in Table 2.r, this average disguises significant variation.

The table identifies seventeen counfries in which revenues were less

than 15 percent of GDP in ry95-98 and fifteen countries in which rev-

enues averaged more than z5 percent of GDP in that period. The low-

extraction countries are typically low-income states without mineral or

oil wealth, the presence of which clearly helps explain the performance

of the high-extraction countries. Nonetheless, the striking differences in

revenue levels for similar countlies does suggest that economic structure

explains only some of the variation and that institutional differences

must be significant as well. Compare' for example, Kenya and Uganda,

similar economies with small manufacturing and large commercial agri-

cultural sectors.

The table also compares these countries over time' to see whether

fevenue collection has improved over the last decade. Interestingly, only

two of the low-extraction countries improved their performance during

this period. In a much higher number of countries, fevenues declined sub-

stantially. Clearly, in many countries, the weakness of revenue collection

constitutes a real handicap for sustained macroeconomic stabilization,

28. Jayarajah and Branson, Structural and Sectoral Adiustrnent, p. 44. t9. iot example, Howard White, "Foreign Aid, Taxes and Public Investment: A

Further Comment," in Journal of Deuelopment Economics 45 G994): r55-64'

3o. For insrance, Karim Nashashibi et al,, The Fiscal Dimensions of Adiustment in Low-Income Countries, IMF Occasional Paper, number 95 fWashington, DC: International Monetary Fund, April rggz); Nashashibi provides data suggesting that government revenues went up in nine African cases and down in nine, rel- ative to a "base year."

7 2 7 3

Sierra Leone Chad Sudan Central African

Republic Niger Rwanda Madagascar Uganda Guinea Mozambique Guinea-Bissau Burkina Faso Tanzania Togo Cameroon Comoros Burundi

7.1. Lesotho 7 .4" o Seychelles 7.5 Botswana 8.0 Eritrea

8.1* Swazlland 9.1 Angola 9.4" Namibia

10.3 o o Gabon '1.0.7"

South Africa 17.3 Djibouti 11.5 Zimbabwe 12.3 Kenya 1"2.9 Mauritania 1.4.6" Cape Verde 1.4.6* 14.8 1.4.9"

38.4 50.20 0

46.8 44.4" 46.8 40.7"

35.5

3 4 . 5 33.7 34.3 2 8 . 1 3 3 . 9 o * 3 2 . 8 2 8 . 5 0 2 4 . 4 2 8 . 5 0 0

28.2 26.8 28.1 2 2 . 6 2 7 . 6 " " 32.7 27.00 7 7 . 7 2 6 . 3 0 " 25.9 26.1.

African Economies and the Politics of Permanent Cri.sis, r979-r999

Table z.r. Extractiue Performance of African Statri, r985-1988, 1995-1998 (gouernment reuenue a.s a percentage of GDP)

Low-Revenue Countries High-Revenue Countries

1985-88 1.995-98 1 9 8 5 - 8 8 1.995-98

Patterns in Reform Implementation, t97g-r999

chance of reversibility once they have been pushed through. The record

here is even more ambiguous and uneven across the economies of the

region. Certainly, widespread progress has been sustained in certain

areas. Almost all countries have undergone extensive price liberalization,

with significant progress on the elimination o{ domestic price controls

and the liberalization of bank credit and interest rates.31 Liberalization

measures were almost invariably at the core of the \ilorld Bank's condi-

tionality in its first generation adiustment programs. The r98os wit-

nessed considerable deregulation in the banking sector' which had long

been a privileged site for government intervention. Banking services had

constituted a public monopoly in well over a third of the countries in

the region in the late r97os; todaS no country retains such a monopoly.

In the agricultural sector, governments have done away with many of the

agricultural marketing boards that enjoyed statutory monopsonies over

the commercialization of export crops and often fixed producer prices

excessively low.32 Elimination of these marketing boards or the encour-

agement of private competition in the commercialization process was

designed to improve incentives for farmers.

The record on domestic llberalization is nonetheless ambiguous. Even

more than for stabllization, positive trends tend to be recent, with rela-

tively little real progress until the late r98os and early t99os. Ghana,

for instance, has often been heralded as an early reformer; on inspection'

3 r. A good, if biased summary of the progress achieved up to the early r99os is pro-

vided by the \(orld Bank, Adjustment in Africa: Reforms, Resuhs and the Road Ahead ('Washington, DC: Wodd Bank, ry94l.The country case studies on which much of that assessment is based have been published as Ishrat Husain and Rashid Faruqee (eds.), Adjustment in Africa: Lessons from Country Case Studies (\(ashington, DC:

.\forld Bank, 1994). See also David E. Sahn, Paul A. Dorosh,

and Stephen D. Younger, Structural Adiustment Reconsidered: Economic Policy and Pouerty in Africa (New York: Cambridge University Press, 1997); and Carl

layarajah and Sfilliam Branson, Structural and Sectoral Adiustment: The World Bank Experience, tg\o-t992, a \forld Bank Operations Evaluation Study (lilfashington, DC: Wodd Bank, 1995).

32. Adjustment in the agricultural sector is reviewed inJacob Meerman, Reforming Agriculture: The World Bank Goes to Market, a World Bank Operations Evalu- ation Study ('$fashington, DC: \ilorld Bank, t997); Andrew Shepherd and Stefano Farolfi, Export Crop Liberalization in Africa: A Reuiew, FAO Agricultural Ser- vices Bulletin, Number 135 (Rome: FAO, ry99); and Pekka Seppdld, "Food Mar- keting Reconsidered: An Assessment of the Liberalizatton of Food Marketing in

Sub-saharan Africa," U.N. Universitn WIDER Research for Action Papet 34 (Helsinki: \trIDER, rg97\.lthank Chris Delgado for bringing these papers to my

attention. See also Peter Gibbon, Kjell J. Havnevik, and Kenneth Hermele, A Blighted Haruest: The World Bank and African Agriculture in the t98os (Trenton, NJ: Africa'!0orld Press, r993),

6 . 6 ) . 2 +

9 . 0 9.4

1 0 . 8 1.0.6 13.2

1.4.1" 1.0.6 L J . , t

70.9 14.4 26.2 1"9.3 1"2.6 17.4

Congo-B

Noles: High-revenue countries are those in which total government revenues averaged above z5Y" of GDP (excluding grants) in r991-98t low-revenue countries are those in which total government revenues averaged below r5% of GDP (excluding grants) in r 9 9 5 - 9 8 . o Revenues decreased by at least zV' of GDP between 1985-88 and ry95-98. oo Revenues increased by at least zY' ol GDP between r985-88 and ry95-98. Sources Calculated hom AfTican Deuelopnrent Indicatols,Table 7,6.

let alone more rapid growth and structural transformation. \fith such low extraction levels, the burden of adjustment must be disproportion- ately on the expenditure side.

Strwctural Adiustment P olicies

Structural adjustment policies are policy reforms that may not have a direct incidence on macroeconomic stability in the short run, but are designed to affect the long-term prospects for economic growth. In general, they can be thought of as "stickier" reforms than stabilization policies, with more complex implementation issues but with a lower

/ n / )

African Economies and the Politics of Permanent Crisis, rg7g-r999

however, reform implementation has clearly beerr belated as well as partial.33 President Rawlings came to office in r98r and economic reform is said to have started in 1983. Cocoa producer prices were increased almost immediately, quadrupling in real value between t983 and 1988, but exchange rates were not fully liberalizeduntil r99z-r993,the import licensing system was abolished only in t989, and the privatization of Ashanti Gold Fields was not completed untll ry94.

Furthermore, key elements of the old apparatus of state control have survived in some countries. Often, the policies have changed on paper, but in practice, something resembling the status quo ante continues to prevail.3a In some cases, the old policies were reinstated under a new name or with some new policy objective. In Kenya, the government restored fixed prices for maize and a tariff on maize imports in ry97. The government justified this return to the prereform status quo by allud- ing to food security and welfare concerns for small farmers, but one study estimates that the policy overwhelmingly favors big farmers, who dominate the maize market.3s

In other cases, governments ignore the spirit of their own liberaliza- tion efforts by continuing to interfere in officially deregulated markets. The hurdles facing private sector investors in many countries offer an excellent example of this situation. Officially, most governments welcome private investment and have removed all the red tape that pre- viously discouraged it. In practice, as a recent review of these issues put it, it is still the case that "in most African countries, the procedures for

73. Ghana's reform path has been much examined. See Jeffrey Herbst, The Politics of Reform in Ghana, t98z-t99t (Berkeley: Universiry of California Press, r99z); Eboe Hutchful, "Ghana," in Poul Engberg-Pedersen, Peter Gibbon, Phil Raikes, and Lars Udsholt (eds.), Limits of Adiustment in Africa: The Effects of Economic Liberalisationr9S6-1994(London:ZedBooks, r99S),pp.r4r-zr4;AliceSindz- ingre, "Politiques economiques, instabilit6s et secteur priv6," in Comi Toulabor (ed.), Ressozrc es politiques et l/gitimitd au Ghana (Paris: Karthala, in press); and Donald Rothchild (ed,.), Ghana: The Political Economy of Recouery (Boulder, CO: Lynne Rienneq r99r).

34. lnher excellent book on the informal secror in Dar es Salaam, Aili Tripp docu- ments cases in which old laws were so consistently evaded by the citizemy that public officials stopped trying to enforce them. In such cases, the actual policy regime is quite ambiguous. See Aili Mari Tripp, Changing the Rules: The Poli- tics of Liberalization and tbe Urban Informal Economy in Tanzania (Berkeley: University of California Press, t997).

3 5. T. S. Jayne et al., Do Farmers Really Benefit from High Food Pricesi Policy Brief, Tegemeo Institute for Agricultural Policy and Developmenr, no.r (September zooo). The authors estimate (p. 5) that the new policy offers an average subsidy of U.S.$r9,zoo to an average large farm of zoo acres.

Patterns in Reform lmplementation, t979-t999

setting up a company and entering into legitimate business are a night-

mate,"36 Despite Mozambique's long-standing reform efforts, fot

example, registering a company still takes six months, and the various

payments required in licenses and mandatoty notary fees typically

approach ro percent of capital.3T Elliot Berg has concluded that in

Senegal, much deregulation and price liberalization has little impact on

private sector investment, which continues to be undermined by "skep-

ticism about the government of Senegal's commitment to liberalization;

the lack of transparency in decision making and implementation coupled

with a generulized suspicion that the playing field is uneven; and the

persistence of administrative delays' weaknesses and harassment."38 A

recent report onTanzani4 often cited as an example of successful invest-

ment liberalization, sounds a similar tone:3e the authors recognize that

the country has made "substantial progress on economic reform" (p. r),

but argue that "the government continues to exhibit the laxity and

indifference to performance of the socialist eta" (p. zol, and note that

"virtually evefy infefaction with the government (by private business-

men) seems to require some sort of side payment to assure that neces-

sary licenses, approvals or clearances are processed favorably" (p, zz)'

The agricultural sector is also emblematic of this tendency of what

one is tempted to call virtual reform' Although policy reform appears to

have advanced more in southern and eastern Africa and less in west and

central Africaral nonetheless, much has changed throughout the region.

76. James J. Emery and Melvin T. spence, "Administrative Barriers to Investment in- Africar The Red Tape Analysis," unpublished paper

'Washington, DC: \forld

Bank, June 1999, p. 9. 37. Ibid, p. 18. 3b. fttioin.tg et a1., Sustaining Priuate Sector Deuelop?nent in Senegal: Slrategic-

Consid.eritiozs (Bethesda, MD: DAI, lune 1997), p. 17. For a slightly different emphasis focusing specifically on the import sector, see Ibrahima Thioub, Momar- coumba Diop, and catherine Boone, "EconomtcLiberaltzation in senegal: Shift- ing Politics in Indigenous Business Interests"' in African Studies,Reu,iew 4t GggS): Q-89. They agree with Berg that despite official trade liberalization, infighting befiveen business and "old style politician businessmen" continues to e"ist ovei the lucrative imported manufacturing market, but they emphasize the state's declining ability to control rent-seeking networks, in a context of declin- ing administrative capacity and political legitimacy.

39. Michael F. Lofchie and Thomas callaghS Diuersity in the Tanzanian Bustness Community and Its Implications for Growth, Report to the USAID Mission, Dar es Salaam, Tanzania, Contract #6zvox76'C-oo-5o35-oo' December 5, r995'

ao. So concludes an excellent fecent review of agricultural adjustment: Hans Bin- swanger, Robert Townsend, and Tshikala Tshhibaka, "spurring Agriculture and RuraiDevelopment," paper presented at the second research workshop, African Development Bank, Abidian, July 6-rr, 1999, pp. ri-rs.

7 6 77

African Economies and the Politics of Permanent Crisis, r979-r999

The expensive fertilizer, seed, and pesticide'subsidies of the past have often been withdrawn,al while the many parastatals that had once pro- moted a crop or region with extension services, input delivery, and other activities have often been closed or had their activities sharply curtailed. Typically, the agriculture sector reform programs that oversaw these reforms also mandated price liberalization, but here reform implemen- tation has been much more ambiguous. It is important to distinguish export from food crops and east and southern Africa from west and central Africa. Many key export crops have not seen their prices com- pletely liberalized. In some countries, poorly informed farmers remain at the mercy of unscrupulous local officials keen to retain their central role in the marketing chain, or of newly empowered private purchasing agents, often in some degree of collusion with state officials. Thus, in Cameroon's cocoa belt, the adjustment program of the early t99os resulted in the old price-fixing system being eliminated in favor of an annual "minimal indicative price," which theoretically established a floor below which purely private transactions could not go. In practice, the lafter prices were more or less enforced as the fixed price.a2 A World Bank review of the agricultural sector suggests this practice is not limited to Cameroon, but is widespread throughout Africa.a3 Reviewing the experience in six countries, Gibbon et al. find that the prices actually received by farmers probably declined during the r98os in half of the cases.aa

The policy environment for export agriculture has improved signifi- cantly more in countries like Kenya and Zimbabwe; here farmer orga- nizations with long histories stretching back to before World War II have effectively pressured governments for favorable treatment. In Zimbabwe, however, the policy regime has hardly moved in the direction of liberal- ization as large commercial farmers have fought for and retained signif- icant subsidies.as In west Africa, on the other hand, the absence of a

4r. Meerman, Reforming Agricultwre, pp.74-8j, covers the record for all input subsidy programs.

42. Georges Courade and V6ronique Alarg "Les planteurs Camerounais ont-ils 6t6 r66valu6es?" in Politique Africaine j4 G99414-87, p. 79 and passim. See also in the same issue of Politique Africaine, Bruno Losch, "Les agro-exportateurs face i la d6valuation," pp. 88-ro3.

43. Meerman, Reforming Agriculture, pp. 7o-r.See also Shepherd and Farolfi, Export Crop Liberalization in Africa, pp. r3-r9

44. Gibbon et al., A Blighted Haruest, pp. ro7-9.

4 j. Tor Skalnes makes this argument convincingly for Zimbabwe. See his book, The Politics of Economic Reform in Zimbabtue (New York: St. Martin's Press, 1995). Zrmbabwe is also marked by growing rural inequality as all farmers are far from

P atterns in Reform lrnplementation' r 9 7 9- t 9 9 9

European settler history has resulted in a much slower and haphazard

reform process for export agriculture' For cotton, usually the leading

export crop in the Sahel, farm gate prices continue to be fixed by a state

marketing board throughout French west Africa.a6 The same is true in

senegal for politically sensitive groundnut prices.aT Governments usually

acceJed to donor demands to improve price incentives to farmers, long

viewed as significantly too low, but have proved extremely reticent to

allow markets to set prices themselves, preferring to maintain their

power to fix Prices annuallY. The pi.ture is similarly uneven for the liberalization of food markets.

In west and particularly central Africa, few states were ever able to reg-

ulate food marketing completelS as private actors found myriad ways

to escape state controls.as The rgSos witnessed the abolishment of many

of the public agencies that states had once created to monopolize food

mark.ting and keep consumer prices down 'ae Typically, however' the

governmJnt has retained a central regulatory role if not a monopoly on

ihe marketing of imported foods, such as rice and wheat' In east and

southern Africa, on the other hand, reform has been chatacteized by

significant backsliding. Typically, controls on food transport and urban

.o.,ro*., prices have been liberalized, but governments have maintained

their attempts to set farm gate prices and have resisted donor calls to get

rid of food marketing institutions.so

b e n e f i t i n g f r o m t h e s a m e l e v e l o f p o l i t i c a l i n f l u e n c e . S e e a l s o C h a p t e r 5 o f t h e fine but ,io* ,o*ewhat dated studyby Jeffrey Herbst, State Politics in

Zimbabue

( B e r k e l e y : U n i v e r s i t y o f C a l i f o r n i a P r e s s , r g g o ) ; a n d M i c h a e l B r a t t o n , " T h e C o m - rades and the countryside: The Politics of Agricultural Policy in Zimbabwe,"

World Politics 39 $987\ 4-zoz.

+g. Cti.i" Freud, ipoliiiques des prix et performances des fillibres cotonnidres en

Afrique," Reuue Tiers Monde 15 ft999): 929-4r'

47. il;;;;'h" Rouis, "senegal: Stabilization, rartial Adjustment and Stagnation"' in

Ishrat husain and Rasf,id Faruqee (eds.), Adius*nent in Africa, Lessons from

Country Case Studies ('!0ashington, DC: \ilodd Bank, 1994), pp' 286-35r'

+s. V"li j;"1 and John Weeks' Africa Misunderstood

(London: MacMill.an' ry91)'

zq. Seppela. ,.FoodMarketing Reconsidered." see also the dated but still excellent-' '

f.iif, Uu, ,, ih, polirlrol E"ono*y of 'West

African Agricuyure .(Ca.mbridge, UK:

cambridge university press, rgliz). see also Daniel Maxwell, "The Political

Economy'of Urban Food S.iority in Sub-Saharan Africa," World Deuelopment

z 7 f t 9 9 9 ) : r 9 3 9 - 5 i '

So. S., ttr. 6". r.ui.* Lf th.r. issues in T. S. Jayne et al, " Successes and Challenges'

of Food Market Reform: Experiences from Kenya' Mozambique' Zambia and

ii^b^b*r," MSU International Development 'Working

Papers' no' 7z (Michi-

gan State University :'999). On Kenya, for evidence of government manipula-

t"ion of cereals -"rk.t ibri^Iir tio"th.oogh rg93, see the interestingcase study

by Gerrishon K. Ikiara, Mohamud Jama, and Justus O' Amadi' "The Cereals

7 8 79

African Economies and the Politics of Permanent Cri,sis, 1979-1999

Trade policy reform has progressed even less far.51 R.eform has long

sought to lower and simplify tariffs, eliminate nontariff barriers, make

customs procedures more transparent, and lower export taxes. Trade

reform has been a staple of all reform programs since the early r98os,

even though tariffs constitute a significant source of revenue for most

states in the region. Yet, in a majority of cases, reforms have been left

unimplemented, or, when undertaken, are often subsequently reversed.

Table z,z provides the most recent available data on trade restrictions.

Officially, trade protection remains significantly higher than in other

regions of the'developing world, with tariff rates averaging about z5

percent, roughly four times the non-OECD average.sz Afuica is the only

region in the world where the degree of openness has not significantly

increased during the last two decades.s3 The actual situation on the

ground is much harder to assess, however, given the extremely uneven

manner with which trade policy is implemented. In some countries, gov-

ernments have undone the impact of policy by tolerating high levels of

corruption within the cusfoms services or extending exemptions from

prevailing rates to favored firms. Lofchie and Callaghy argue that in Tan-

zania, "corruption in the customs bureaucracy is so extensive that Tan-

Chain in Kenya: Actors, Reforms and Politics," in Peter Gibbon (ed.), Markets, Ciuil Society and Democracy in Kenya (Uppsala, Sweden: Nordiska Afrikainsti- axeq r995), pp. 3r-68.

5r. The progress of trade policy reform is discussed in John Nash, "Trade Policy Reform Implementation in Sub-saharan Africa: How Much Heat and How Much Light?" \7orld Bank I7orking Paper (\Tashington, DC: lforld Bank, 1995); F. Ng and A. Yeats, "Open Economies Work Better! Did Africa's Protectionist Policies Cause Its Marginalization in l(orld Trade?"

.World Bank Policy Research

lWorking Paper, no. 1636. (\(ashington, DC: .World

Bank, 1996); and Bdatrice Hibou. I'Afrique est-elle pfotectioniste? (Paris: Karthala, ry96). Charles Soludu focuses on a sample of countries with a somewhat better trade reform record in his essay, "Trade Policy Reforms and Supply Responses in Africa" (Geneva: UNCTAD, September 1998). On the political economy dimension of trade policy reform, see Dani Rodrik, "\(hy Is Trade Reform so Difficult in Africa," Journal of African Economies 7, Supplement r (r998): 4-69; and Henry Bienen, "The Politics of Trade Liberalization in Africa," Economic Deuelopment and Cultural Change 3 8, no. 4 $99o1: 71312.

jz. Ng and Yeats, "Open Economies'Work Better!" 53. This is the conclusion of Jeffrey Sachs and Andrew \farner, "Economic Reform

and the Process of Global Integration," Brookings Papers on Economic Actiuity r (r995): r-r19; of the'World Bank, Global Economic Prospects and tbe Deuel- oping Countries, 1995 (I7ashington, DC: \7orld Bank, ry96); and of Robert Sharer et al., "Trade Liberalization in IMF-Supported Programs,"

'World' Eco-

nomic and Financial Suruey (Yashington, DC: International Monetary Fund, Febrtary ry98).

Patterns in Reform Implementation, r979-t999

Table z.z. Trade Restrictions in Africa

Country

R a t e s ( % ) Restrictions (%)

1984-87 L991.-93 1984-87 1991-93

C6te d'Ivoire Ethiopia Ghana Kenyao Malawi'' Mauritius*' Nigeria Tanzaniao Ztmbabwe'' All Africa Sub-Saharan Africa Latin America South Asia East Asia All Countries

Z J . J

2 9 . 0 2 9 . 6 39.2 1 6 . 7 34.9 2 3 . 8 32.'t 8 . 7

2 6 . 3 L O . - )

2 6 . 6 6 1 . 7 1 7 . 9 27.0

4 3 . 7 1 5 . 2 z / . o

3 2 . 8 2 9 . 8 1 0 . 1 25.7 2 5 . 3 1"2.3 47.5 1 6 . /

22.1.

6 . 6

48.4 67.3 9 6 . r

1 7 . 0 62.2 2 . 5

4 3 . 7 44.0 30.2 4 7 . 6 21..2 37.7

3 7 . 8 9L.3 3 s . 2 8 . 8

79.7 93.5 42.9 4 7 . L 8 . 6

20.4 3 . 5

2 3 . 7

Dut^ ut" from ry84-7 and rg88-go. Regional averages are taken from 15 sub-Saharan

African states, rr from Latin America, 5 from South Asia, and 7 Irom East Asia'

Source: Data are from United Nations Conference on Inade and Development. Cited in

World Economic Forum, The Africa Competitiueness Report (Geneva: World Economic

F o r u m , r 9 9 8 ) , p . 4 r .

zania is) for all practical purposes, a duty-free zone."s4 In other coun-

tries, the trade sector continues to face redtape and excessive interference

from government bureaucracy' so that reform may not have changed the

effective rate of plotection faced by business given various transaction

costs that appeaf to have increased. The uncertainty and fandomness of

the situation in many countries, even when the effective levels of pro-

tection are quite low, are almost certainly a strong deterrent to potential

trade growth.

Public enterprise refofm offers an interesting contrast; after proceed-

ing at a snail's pace during the r98os, it has picked up speed in the

r99os.55 Over time, privatization has become more prominent in the

5 4 . 5 5 .

Lofchie and Callaghg Diuersity in the Tanzanian Business Community, p. z,

A useful survey is provided in Paul Bennell, "Privatization in sub-saharan Africa:

Progress and Prospects during the r99os," 'World

Deuelopment L5 Gggl\'

ry{5-8q; see also \forld Bank, Bureaucrats in Business: The Economics and

8 o 8 r

African Economies and the Politics of Permanent Crisis, rg79-r999

reform agenda, as various schemes to improve the management of public

enterprises invariably failed to stem their thirst for costly subsidies.

Nonetheless, during the rg8os, pivatization appeared stalled by a

combination of political opposition and the technical difficulties of

arranging for sales. Governments worried about the implications of large

retrenchments. In the mole recenf past, howevet' pfivatization transac-

tions have been completed in a number of countries. According to

Bennell, the period of rgSo-rg87 witnessed some zz7 pivatization

transactions, while 657 transactions occurred between t988 and r995,

and over 3oo in r994-g5 a1one.56 Initial resistance to privatizaticin

appears to have been overcome primarily by the growing unwillingness

to bear the burden of subsidizing loss-making parastatals combined with

the attraction of the revenues to be generated by selling off public assets.

Bennell (p. tZgo) thus points out that ptivatization transactions raised

almost z billion in the region between r99o and t995'

In many cases, governments remain unwilling to part with the biggest

of their public enterprises. The case of Zambia is particularly revealing.

That country has undertaken an ambitious ptivatization program since

1992, but hesitated throughout the t99os to privatize the ZCCM, the

giant copper mining concern.sT Successive governments opposed selling

the company for ideological and nationalistic reasons, and ra;tionalized

that the company was the engine of the national economy. By the late

rggos, however, zccMwas losing more than $r million a da5 the result

of decades of systematic underinvestment and poor management. The

government announced its intention to sell in t996, but now priva-

tization was delayed by investor hesitation in light of the company's

financial difficulties, as well as by the accusations of rent-seeking and

corruption on the part of senior state managers. President Chiluba had

entrusted the sale to a special team father than the national pdvatiza-

Politics of Gouernment Ownership ('lfashington, DC: World Bank, r995); and Elliot Berg, ,.Privatization in Sub-saharan Africa: Results, Prospects and New Appro""his," in Jo Ann Paulson led.), African Economies in Transition: Volume tt Tb, Rrlor* Experience (New York: St. Martin's Press' 1999), pp' zz9-89'

56. Bennell, "Privatization in Sub-Saharan Africa"' p.1789.

!7. This "..oo.rt oI the ZCCM privatization has been compiled from_Colin Barra-

clough, "Trouble in Lusaka,-" Institutional Inuestor (international edition) xz, no. i3'ft99g): 47.lhave also relied on the the 6ne series of articles on the trans- u.tio" Uy ftr.o'Bull in Profit Magazine throughout 1997 and r9_98. The general

zambiancontext is explored in Lise Rakner, Nicolas van de walle, and Dominic Mulaisho, ..Aid and Riform inZambia," unpublished report for the

riforld Bank project on aid and reform,'SVashington, 1999.

Patterns in Reform Implementation, t979-t999

don agency. The head of the team, Francis Kaunda, received a salary of

$r6,ooo a month and was accused of slowing down negotiations pur-

posely to maintain his position. Negotiations were also slowed down

when foreign investors learned that company assets were being privately

sold to company officials and politicians' and they demanded a precise

inventory of company holdings. In particular, some zoo houses owned

by ZCCM had apparently been secretly sold off. A deal was almost

struck with potential buyer Anglo-American in late 1998 but was

allegedly undone by Anglo's refusal to pay the large side payments

demanded by the president's office to seal the deal. Agreement was

nonetheless reached with Anglo in April zooo; by that point, the

company's disastrous finances were weighing heavily on government

finances and the donors were conditioning any increases in aid on an

agreement,5s In other cases, the fear that privatization will benefit certain ethnic

groups or foreign business interests has stalled divestiture, or the process

has suffered delays following accusations that the transactions appeared

to benefit members of the political class, usually by significantly under-

valuing assets and selling them to members of this class at fire sale

prices.se As Lewis comments about the privatization process in Nigeria,

much of which was conducted through offerings on the national stock

market: "The dispersal of government assets' especially in a tight

economy, created numerous opportunities for windfall gains. The weak

regulation of securities transactions facilitated the use of insider infor-

mation for substantial profit. . . private placements and closed bidding

provided easy avenues for directing favors to friends and clienfs'"60 But

few governments still retain the pronounced ideological opposition to

privatization they openly expressed in the rgTos.tt No matter how con-

tentious privatizationtransactions have been, once achieved, they are not

challenged. Thus, even with the regime changes brought about by democ-

latization in the early t99os) there have been no examples of renation-

alization of privatized firms.

58. ZCCMT belated privattzation is chronicled in "Zambia Finalizes Sale of State Copper Mine Group," The Financial Times, April 4, zooo, p. 38.

59. This argument is well made by Thandika Mkandawire, "The Political Economy olPrivatlzarlon in Africa," in Giovanni Andrea Cornia and Gerald K. Helleiner (eds.), From Adjustment to Deuelopment in Africa (New York: St. Martin's Press, r994, pp. t9z-2r6.

6o. Peter M, Lewis, "Economic Statism, Private Capital and the Dilemmas of Accu- mulation in Nigeria,"

.World Deuelopment zz (tgg+|, 437-5r, 446.

6r. See Bennell, "Privatization in Sub-Saharan Africa," for a similar argument.

8 38 z

African Econornies and the Politics of Permanent Crisis, 1979-1999

Finally, there has been little progress on civil service refotm, which

has long been discussed but rarely implemented. The objective of reform

is both to cut down on unnecess ary staff and to improve the conditions

of service for the remaining staff. Virtually every country in the region

has undergone some kind of civil service refofm program, supported by

international finance.62 The donors have long recognized the need for

reform but have hesitated to promote it, given its likely cost' complex-

ity, and perceived political sensitivity. As a result, it has rarely been at

the forefront of adiustment programs, relegated instead to a back burner.

Governments have tr."n -iuing to commission studies on thb civil

service. to institute costly and largely ineffectual donor-funded programs

of voluntary retirement or campaigns to get rid of ghost workers, but

very few have been willing to lay off sizable proportions of the civil

service or to increase salaries to keep up with inflation. For their part,

the donors have rarely focused their conditionality on civil service reform

progress. Three notable exceptions to these pattefns are Uganda, Guinea,

and Ghana, where progfams resulted in the civil service being cut by

Lo)ooo) 3o,ooo, and 6o,ooo positions, respectively in the late r98os'53

Elsewhere, there has been remarkably little progress'

It would have been astonishing if over two decades of fiscal crisis had

not served to cut the size of the civil service relative to the population,

as it appears to have done; on avetagercivil servants amount to r pefcent

of the population, down from r.3 percent in t99l., belying the image of

6 2 . See the informative collection of essays in David L. Lindauer and Barbara

Nunberg (eds.), Rehabilitating Gouernment: Pay and Employment Reforms in

Afri.ca (\lashington, DC: \forld Bank, ry94).ln her own contribution, Nunberg

cites fifiy-five rfforld Bank loans wirh a civil service reform component extended

to Africa between rg8r and rggr. see "Experiences with civil service Pay and

Employment Reform: An Overview," pp'rr9-59. See also Bamidele Olowu'

"Redesigning African civil Service Reforms," Journal of Modern African s-tudies

lz Gssg\: i-4; and Alice Sindzingre, "Dimensions economiques des r6formes

'd" l'"iit'rn Afrique sub-Saharienne,;' in Comi Toulabor and Dominique Darbon

(eds.), R,forme de I'etat: Reconstruction institutionelle et modes de r'gulation

( P a r i s : K a r t h a l a , i n p r e s s ) . Louis de Merode and Charles Thomas, "Implementing Civil Service Pay and

Employment Reform in Africa: The Experiences of Ghana, Gambia and Guinea,"

ln Davld L. Lindauer and Barbara Nunberg (eds.), Rehabilitating Gouernment:

Pay and Employment Reforms in Africa (Iilashington, DC: \(orld Bank, t994),

pp'. 6o-z:'o.'The Ugandan numbers are ambiguous, given the-high number of^g|ort -ork.rs and ihe distinction between general public employment and the

iivil service. The overall public employment rolls were cut by an amazing

r6o,0oo employees between t99o and r996' See Table 9 of Lienert and Modi,

"A Decade of Civil Service Reform."

Q .

8 4

Patterns in Reform Implementation, t979_t999

an overdeveloped state.6a Nonetheless, the best available estimates from the IMF suggest that between ry86 and 1996, the number of civil ser- vants actually increased in eleven of the eighteen countries for which there are complete data. The more ambitious elements of the reform agenda, focusing on increasing civil service profession alizationand effec- tiveness, have progressed even less. The same IMF data indicate thaf average real salaries increased in only seven countries during this same period, suggesting that in mosf countries working conditions are not improving.65

Neglected lsswes

This chapter has so far focused on the agenda of reform defined by the international financial institutions since the late t97os. It has not said anything about an array of other issues African governments need to address, in all likelihood, for their economies to sustain rapid growrh. Issues like judicial se.tor reform have been mostly neglected by the donors until recently despite obvious deficiencies and the general recog- nition that improved performance is a prerequisite of growth.56 In c6te d'Ivoire, for instance, one review of judicial reform efforts by the donors from r99o to 1996 notes, "Donor involvement was half hearted, even perfunctory' French aid was limited in volume and in scope. . . . The Bank's commitment seems to have been lukewarm and partiarrwith inter- mittent involvement. Judicial component funds were cut for . . .lack of activity. No conditionality was attached until 1996.,,67

The health sector presenrs similar situations. In spire of any progress in reforming the health sector across the region, access to health services remains far inferior to every other region of the world. An African child in 1995 was 6o percenr more likely to have been vaccinated against measles than she would have been in r98o, but she remained more than

ri

64. Lienert and Modi, "A Decade of civil service Reformi' p. ++. The size of the African state and its economic implications were recently inaiyzed, by Arthur A. Goldsmith in his fine essay, "Africa's overgrown state Reconsid"r.d, B,rr"urr-

o ) .

6 6 .

cracy and Economic Growth," World politics 5r (1999): 5zo-46. I b i d , p . 4 3 . A general survey is presenred in Richard E. Messick, "Judiciar Reform and Eco- nomic Development: A survey of the Issues," wortd Bink Research obseruer ta ( r 9 9 g ) , r t 7 - 3 6 .

67 . Elltot Berg, Patrick Guillaumont, Jacques pegarien an, and Jacky Amprou, ..Aid and

Reform in the c6te d'Ivoire," case Study for the world Bank project-..Aid and Reform in Africa," unpublished paper, V/ashington, DC, December r99 9, p. 6c'.

8 5

African Economies and the Politics of Pertnanent Crisis, 1979-African Economies and the Politics of Pertnanent Crt'sr's, r979-r999

a third less likely to have been vaccinated than the average Latin

American child!6s Furthermore, two decades of crisis leave most coun-

tries tragically unprepared to address the current AIDS/HIV crisis' which

threatens to reduce life expectancy in the region from fifty-nine years

today to forty-five in zoro, or the various other health epidemics that

may be lurking around the corner, from the ebola virus to drug-resistant

strains of tuberculosis and malatia.6e perhaps even more seriously, all across Africa, investment budgets

have been sacrificed to pfotect the more politically sensitive recurient

budget from the cuts in expenditure made necess ary by the economic

crisis and mandated by reform programs, particularly in cases in which

fevenue generation does prove stagnant. The donors have recognized

that reshaping the patterns of public expenditures to favor development

is just as important in the long fun as cutting the fat out of budgets, but

the Bank it .lf ,..og.rizes that its efforts to do so have largely failed.7o

Thus, in the rggos, gross public invesrment has declined to an averaBe

of less than 6 percent of GDP. The proportion of expenditures devoted

to public infrastructure has declined sharply.

Skimping on maintenance and investment is an inevitable feature of

macroeconomic stabilization programs and not a cause for alarm in the

short run. Over time, however, the prolonged nature of the African crisis

is turning what could be justified as a necessary short-term expedient

into a critical obstacle to renewed economic growth. The African Devel-

opment Bank estima ted in ry99 that a third of the roads built in the last

two decades had eroded due to inadequate maintenance.Tl These accu-

mulated maintenance deficiencies represent significant obligations on

future budgets: the'World Bank estimatedin ry95 that additional outlays

of over $r.5 billion ayeat for a decade were needed simply to restore

the existing road network in the region to an appropriate level.7',Poor

roads result in high tfansport costs, which undermine Africa's competi-

68. Calculated from the l7orld Bank, World Deuelopment Indicators, p' 94' 69. See the blunt assessment by the National Intelligence Council, Tbe Global Infec-

tious Disease Threat and Its Implications for the lJnited States (Washington, DC:

National Intelligence Council, January zooo), especially pp' t-4-7'

7o. The world Banik, The l?npact of Pwblic Expenditure Reuiews: An Eualuation,

Report Number 18573, dperations Evaluation Department (\fashington, DC:

I7orld Bank, November 13, ry98), 7r, The African Development Bank, "Infrastructure Development

in Africa," in

African Deuelopmeit Report, t999 (Oxfor4 UK: Oxford University Press,

x999), p. t3r. zr. fhr-Sif"rtd"Bank, A Continent in Transition: Sub-Saharan Africa in the Mid-

rggos (.$Tashington, DC: rl7orld Bank, November, r995), p' 58'

P atterns in Reforrn Imp lementation, t 9 7 9- r 9 9 9

tiveness. In 1998, the Kenyan newspaper' The Nation, reported that the

country's main commercial highway between Nairobi and Mombasa was

so poorly maintained that the 5oo kilometer trip took three days for a

truck laden with merchandise.T3 Similar levels of maintenance expendi-

tures are likely to be necessary for other government assets, from build-

ings to public utilities and enterprises. 'lflithout

making these essential expenditures, it is difficult to see how

many of these economies can sustain growth, There is already much evi-

dence that inadequate public investmenf has an impact on growth and

poverty alleviation. During the r98os, Africa was the only region of the

developing world where agricultural labor productivity actually declined.

According to Khan, between r98o and r99o, it declined by .4 percent a

year, while increasing by ,.1percent in South Asia and 3.9 percent in the

Middle East and North Africa.Ta The decline appears to have quickened

in the r99os. There are complex causes for this decline, but the dismal

condition of rural infrastructure is surely a primary one. For example,

only 4.6 pefcenr of sub-Saharan Africa's (SSA) agricultural land is irri-

gated, compared to 38.4 percent in Asia'75 Africa's road network is by far

rhe most poorly developed and the least satisfactorily maintained.

Dunsfan Spencer has estimated that the humid and subhumid regions of

tropical Africa would have to increase the density of their rural road

network by a factor of five to reach the level enjoyed by India in 19 5o.76 In turn, declining productivity is one cause of the production stagna-

tion that character\zes the region's agricultural sector. Ali and Thorbecke

link all these factors to striking patterns of growing rural poverty in

Africa, with 59 percent of the region's rural population living below

the poverty line.77 The region, they conclude, "in comparison with other

77. "Despair Road Thwarting Trade," The Nation (May ro, t998):3''zq.

cit"d in Ali Abdel Gadir Ali and Erik Thorbecke, "The state of Rural Poverty,

Income Distribution and Rural Development in Sub saharan Africa," paper pre-

oared for a conference of the African Economic Research Consortium on Com-

iarative Development Experiences in Asia and Africa, Johannesburg, November

6, r9g7 (revised version, April r998), p. ro.

7 5 . I b i d , p . 1 3 .

72. Swdy "itia

in Christopher Delgado, "Agricultural Transformation: The Key to' Broad-Based Growth and Poverty Alleviation in Africa," in Benno Ndulu and

Nicolas van de walle (eds.), Agenda for Africa's Economic Renetual (!(ashing-

ton, DC: Overseas Development Council, t996)' p. t67,

77. Ali, arld Thorbecke, ..The state of Rural Poverry" p. g. ol course' these totals

disguise wide variations across the region, from Cote d'Ivoire with a poverty head

corint ratio of 38 percent to the Central African Republic with an incredible 78 percent.

8 78 6

African Ecorutmies and the Politics of permanent Crisis, 1979-1999

regions, suffers from greater, more severe and more persistent poverty, more unequal distribution of income, declining food producdon per caprta and agricultural labor productivity; and. a continuing population explosion" (p. rt).

These various lacunae almost certainry arso explain why the reason- ably positive economic climate of the mid-r99o, hu, not produced more than a small increase in investment, which remains too low in most coun- tries.78 rn t996, the African region received ress than r percent of global foreign direc invesrment (FDI), and two-thirds of thai-enr to Nigeria and Angola for oil exploration.Te The enormous capital flight that hls characterized Africa these last twenty years was ,rot ,.u.rr.J despite the apparent improvement of the mid-rggos. one estimate is that 40 p.r..rrt of the pfivare wealth of Africans is held ourside the region.80 Nigerians alone held an estimated $5o billion outside their country.rt In surveys, businessmen cite poor governance, uncertain property rights, the dismal condition of the infrastructure, and the cost implications of poor health conditions on labor productivity to explain their reticence to invest in the region.82 Economic studies suggest that mala,,a cosf the region the equivalent of r percenr of GDp a year in the mid_r99os, and UiVAIDS lowers GDP in some countries by .25 percent a year, a testament to the dismal state of the public health system in the region.83

There has been a slight increase in investment in a minority of coun- tries like uganda and Ghana, which have been rewarded for their rcra- tively stable macroeconomic policies in recent years.sa Both have received

78. Fischer' Is This the Turning point? see also FranEois Bost, ..IjAfrique subsahari_ enne, oubli6e par les investjsseurs,', in Afrique Contemporaini il9 ft999): 4 r - 6 r ,

79. see Jeff sachs, "Foreign Direct Investmenr in Africa,,' it The Africa competi- tiueness Report, r998 (Geneva, Switzerland: Iforld Economic Forum, ,igAL p.

8o. i'aul collier, Anke Hoeffler, and Catherine patillo, ,.Flight capita l as apottlolio Choice," International Monetary Fund'Working f^prr\WVtgitrr, fW"rt irrgrorr,DC: Internadonal Monerary Fund, Decemb

", i9991, p. 7.8r. cited by Lewis Machipisa, "Africa Loses Miilionr ihio"gtt capital Flight,,, on

Interpress Service, May rz, t999, found. on a Lexis_Nexi. Int.rn"t searfh. lt. S^9. Jeff Sachs, "Foreign Direct Investment in Africa,,, p, 37.83. Cited in Vijaya Ramachadran,Inuesting in Africa, noii.y nr.uy no. z5 (\7ash-

ington, DCr Overseas Development Council, in press). a+' 91 Uganda, see Robert L. sharer, Hema R. De zoysa, and carvin A. Macnon-

ald, U_ganda: Adiustment uitb Grotuth, r9g7_r994; Occasional pape, Nu_b., rzr (Vashingron, DC: International Monetary Furri, Mrr.h ,9951'; ii Gh^nr, Robert P. Armsr-rong, Ghana Country .Assistance Strategy Reuiew: A Study inDeuelopment Effectiueness flJ7ashington, DC: Vorld Baii, ry9e1,.

- ---

Patterns in Reform Implementation, 1979_:-999

several hundred million dollars in FDI in recent years.8s unfortunately, the perception that the progress on stabilization is not sustainable will tend to be a self-fulfilling prophecy; if private sector agents do not view the reforms as fully credible, they will take a wait-and-see attitude before changing their economic.behavior. Even fhese relatively successful coun- tries have in fact underinvested in basic public goods, such as education

and infrastructure. Surveys suggest they suffer from many of the ills,

albeit in attenuated form, of their neighbors. A recent survey of barriers to private sector investment gave poor marks to both Ghana and Uganda on such issues as rcdtape,bureaucratic harassment, and government cor- ruption, for instance.s6

The conclusion is inescapable that the belated progress on macroeco- nomic stabilization witnessed in the r99os will be difficult to sustain, given its shaky foundations and Africa's pressing developmental needs. A small number of countries appear to have begun to put their economies on a solid footing, but in most, improved statistics in the late r99os will likely proved short-lived.

The uneven implementation of reform policies is summarized in Table 24. \X/ith the caveat that generulizing across all cases is difficult, the course of reform can be broadly summarized in the following manner. \(hile there has been some undeniable progress on changing economic policies, this is uneven across the region's economies, given to partial implementation and vulnerable to reversals. Basic stabilization reforms have a better implementation record, but they have proven less likely to be sustained. Fiscal adjustment, in particular, has suffered dramatic setbacks after bouts of clear progress. Exchange rate reform has usually been sustained with the greatest consistency. At least outside the Franc Zone, there is a clear pattetn of countries moving fo more realistic and flexible levels in a way that appears broadly sustainable. These countries have not resorted to the printing presses to finance deficits in the r99os. But here too, there are no institutional checks to quick reversals.

Structural adjustment reforms have proceeded more slowly, but they Appear less vulnerable to reversal. In some areas, reform appears to have taken hold. In particular, privatization, though slow to start, has not been questioned once it has been implemented. There are no examples of rena- tionalization in the region, suggesting that the days of nationalization

85. UNCTAD, Foreign Direct Investment in Africa: Performance and Potential (Geneva: United Nations Commission on Trade and Development, July r999).

86. Emery and Spence, "Administrative Barriers."

8 8 8 9

o\ H I o\ N o\ H

R

q)

\)

\\)

\) . L\

F

\\)

h a

o

..1 0,)

r-

I

d

u d

U

(n ! o o

(s

O

o

d

>.

€ o

f, E tg,g $ E

' u a e e r o € E g i g F i * qtEgx- E X o E e c r J * 1 ? I

f E es'l€5f $ E H g E € E 9 U g - i :

e E s a , : E E : F # H F , d H ' L T E T : € N ; i = <

' E R F ; H . . l . 9 c s . E ' H= F l ! 6 ' ) g E s i f " € H H ;

;[s;l i+e;?nEEsc $bld $ EI5 s iE g; H.; 5E€ E } F5 g; g E HE I g figFEg FF3'!53EfrE5

f , o l F5 E 6 E x x E E

: : E t s E E H H s + $ g * s H E E x E EEHHH EE . Y + . Y i ! + ' ; ' : i ' l g

E T E E.EEEE ii - t s x !

> . ! . H o . r € . e s H t p E = d i i i ! !

H & h t s E n s € P " ,

. Y i 1 3 . 9 8 q g E H E g - S E E F o t h . g o E ! : S = - b 9 ( ! S . y g . 9

x a S o F * r . H . { E l E E F i . 9 ( ! n ' S j i 6 x o ? o i o

o

sr € H; EEESS; ** ; E ' E f r = E E A ' , , r F d 5 ; ; < i . ; 6 d ; r j d T i

; F i

P atterns i.n Reform Implementation, r 9 7 9- t 9 9 9

are over. A broad array of domestic markets have been liberalized, from

credit to agricultural inputs. In other areas, there has been virtually

no sustained progress. Civil service reform is one such glaring example.

Why has there not been more progress on policy reform? How can we

explain why certain types of reform have been implemented across the

region and not others? \fhy has the speed of reform varied across policy

areas?

T H E E V O L U T I O N O F P U B L I C E X P E N D I T U R E S

A more careful look at the evolution of public expenditures holds impor-

tant clues to help answer the questions posed at the end of the last

section. I first look at patterns of public employmenf and expenditure

across major world regions in Tables 2.4 and 2.5. caution must be used

in interpreting these numbers. Cross-national expenditute data are noto-

riously unreliable as definitions of expenditure types vary across coun-

tries, and some governments choose to move expenditures from one

category to another for political reasons. For instance, there is much evi-

dence that the category of. capital expenditure often includes various

recurrent expenditures and subsidies (notably to the parastatal sector),

as many developing countries wish to appear to spend less on recurrent

expenditures than they actually do.87 The African subtotals are even

more suspect, given missing values for half to two-thirds of the countries

in the region, depending on the indicator. It is difficult to be sure how

representative of the entire region are the cases for which we have data.

In Table 2.5,1recalculated regional averages for only the low-income

countries in Africa; at least in the expenditure category of subsidies and

current transfers, there was a substantial difference in the estimated

regional average between the sets of countries.

Nonetheless, these data can give us a sense of the degree to which

state expenditures respond to societal demands. In this respect' the data

in Table 2.4 about public sector employment suggest the extent to which,

at least in a comparative sense, the African state is not unusually large.

To be sure, African state structures grew extremely quickly in the decades

following independence and they are unusually large relative to the size

g7. To citeiust one example, the wodd Bank estimated in the eatly rg:osthat 55' percent of the health sictor investment budget in Senegal should really have been iesignated as recurrenr expenditure, See the \florld Bank, Senegal: Public,Expen- ditie Reuiett,, Africa Region, Sahel Department (rwashington, DC:'World Bank, December 3r, r99r), P, zo,

9 r

African Economies and the Politics of Petmanent Crisis, 1979-1999

Table 2.4. Gouernment Employment, Eaily t99os (as percentage of populatron)

Central Local Teaching &

Health

Africa (n = 20) A s i a ( 1 1 ) Eastern Europe & Former USSR (17) Latin America & Caribbean (9) Middle East & North Africa (8) OECD (21)

Overall

worr' Couii"ge is of total civilian public sector; regional averages are unweighted.

Source: Salvatore Schiavo-Campo, Giulio de Tommaso, and Amitabha Mukheriee, Az

International Statistical Suruey of Gouernment Employment and Wages, Public Seclor

Management and Information Technology Team, Technical Department for Europe,

Central Asia, Middle East, and North Africa (\Tashington, DC: World'Bank, undated). From \i/orld Bank website.

of what is usually a small formal sector. Nonetheless, the growth came

from a very low base, so that, as a proportion of their overall popula-

tions, they remain significantly smaller than their counterparts in other

regions of the developing world.88 Table 2.4 calls forth the following styl-

ized facts. African states are relatively small; they are primarily based in

the capital and involved in central administration. Roughly 40 percent

of this employment concerns the educational and health sectors' a pro-

portion that is roughly similar to those of countries in other regions'

although again a smaller share of the national population. In addition,

the available data indicate that few African sfates suffered a significant

decline in the absolute numbers of public employees in the r98os and

r99os, though they usually declined relative to the overall population,

as the growth in public employment did not keep up with that of the

overall population, which grew at 2,5 to 3 percent a yeat.

There are probably many reasons for these patterns. Nonetheless, this

cross-regional comparison suggests a relatively small number of public

employees in Africa working for a state that does not provide many

88. Goldsmith, "Africat Overgrown State Reconsidered: Bureaucracy and Economic Growth."

General Government

l l l r ^ i l q ? | o \ i l t a I o \ l l - e l H i l P . r I

i l ( n ; . I

ll".rlR ll-'* l l r - l l u - l R l l s , ! l * I I E E I [ o s I l l ^ ! l A

l l 5 r - l o o

ll- 13 l l i l l R l l o l o \ l l u E l * l l 9 e I i l e t r | i l E X l

l l r a l g l l 1 3 t l l l , - i l | o \

l l d B l s i l @ ! |

Is* I l l > ' l g l l 1 3 t l ll tn l l u r l e l l o u I

l t € E l l l S s l s ll rl t l tl ll ll t l tl tl t l il t l tl

$\

\ \) * $

\) !\ q)

6\ H

s- lI]

$

\ \) rt

rq

\)

L\) n o

t \

-t

.i ()

r 6r-

2 . 0 2 . 6 6 . 9

3 . 0 3 , 9 7 . 7 z+. /

0 . 8 1 . 0 5 . 1

1 . 1 1 . . 6 J . +

2.4

0 . 9 0 . 3 0 . 9 0 . 7 1 . 0 0 . 8

1..2 0.7 r . 4 0 . 9 1 . 8 2 . 5

1 . 11..2

9 z

t{- O\ cO b

N t'- f-- c o d

\ a \ . ) + . C O \ O d d . N ' N h

O c.l oo N e . l N

h V ) b O N N N N

o o ^ OO \O aO -1 O\ F T N H N t )

c O l A F l t i l { . O \ i H i i l t a \ l

. N t r .

O \ o | t " c o N N ' c o " i

i d N \ f N . d c o c o 6 l N c o . H

h \ o h o \ h \ f ( f ) h a O N

6 + . O . # @ h h h . a N

s ' i

i E g - o . - F * A c C ' - a , ! l

d . ' i ! €

t q d , 9 \ J o S ' i A ' 4 V L

! i = 9 Q . \ < 6 d . t d v c o r , . 9 e P F E " o b g c c F = o c t r c i i ' 6 X

i 9 ' 6 ] " : < e 3 + < l d s 7 - o F l t ' E P E T o ; s f i i > 3 8

@ o\ o\ H

G

F (,

bD

d

; o\ o\ H

\J \

s 6

F

+ x € 6 F i b D - a t s ! : N 9 o ! t s

E 'rr . 9 9

a a .o .q 6 o i i.o :J

\) w)

d 9

3 . !

d F o - O

9 U X C

< V o " , o

d . 9

t s s - i o

o * bo q! P H o 6 6

1 . 9 a t q H

l F c l o I l b o [ j | 6 . i

t P = l r - r O l "

o

t , ; u t o H t b D : t d x

l o ' ? 1 9 o

l g . E

t d t r

t . : 6

l : 6 t - o l o q l l - { t 9 ^ d t @ +

1 6 o l > 6

1 ! ? u

i l 6 : i l F q -

l l - - I ,

i l ! ? t r l l 6 6 i l 4 d

l l 9 E i l - 6 l l 9 z i l . Y o l l b o = i l o i

i l d 5 l l o d

l l € > i l <

l t s ; l l o 6 | z o

African Economies and the Politics of Permanent Crisis, 1979-1999

services and is largely absent outside the capital. This impression is sup' ported by studies on the qwality of public expenditures' or the efficiency

of expenditures at improving the welfare of the population. For instance, research by Gupta, Honjo, and Verhoeven has concluded that health and

education expenditures in Africa consistently underperformed similar

expenditures in Latin America and Asia in terms of improving health and

education attainments, largely because of the disproportionate amount

spent on salaries for state personnel relative to other outlays.8e Some anecdotal evidence also suggests that staff and expenditures in social' sectors may not be oriented toward providing social services' In the early r99os, more than ro percent of Senegal's public teaching staff occupied

nonteaching jobs in the central ministry offices in Dakar,e0 for example,

while in Tanzania, the primary enrollment rate went fuom 93 to 66 percent between r98o and 1994-96, even as the number of teachers was

increasing from 8r,ooo to just under rog,ooo during this same period.el

In Mali, a rWorld Bank studyez noted that the government did not allo-

cate any budgetary resources Ior primary health carc and health educa-

tion programs in t997, which were entirely funded by external donors (p. ar).That year, over 8o percent of government health expenditures went to, in order of importance, central administration, logistics, sec-

ondary care, and uaining of staff. Mali is hardly unique: another study

notes that donors financed 'oneady all capital investments in health in

the public sector" throughout Africa during the t99os.e3 Table 2.5 offers additional evidence of this phenomenon by provid-

ing a comparative look at what governments spend money on. Based on

89. See Sanjeev Gupta, Keito Honjo and Marijn Verhoeven, "The Efficiency of Gov- ernment Expenditure: Experiences from Africa," IMF Working Paper, Fiscal A{fairs Department (\(ashington, DC: International Monetary Fund, November t997). Paul Collier and Jan Gunning cite other studies showing that social expen- ditures in the region consistently underperform public expenditures in other regions, and conclude that this is because "the public sector has been used to

create employment rather than to deliver services, and this reduces productivity" (p. Zo). See their essay, "Explaining A{rican Economic Performance," Journal of Economic Literature 37 (March ry99): 64-xtx'

9o. The'World Bank, Senegal: Public Expenditure Reuietu, p, 21,

9r. The'SVorld Bank, African Deuelopment Indicators, Tables r3-r4 and x3-t7.

92. The'World Bank, The World Bank and the Health Sector in Mali: An OED Country Sector Reuiew, Report no. 18rrz (I(ashington, DC: \(orld Bank, June 3 o , r 9 9 8 ) .

y. DavidH. Peters, Kami Kandola, A' Edward Elemendort and Gnanaraj Chellaraj, Heahh Expenditures, Seruices and Outcomes in Africa, Vorld Bank Health Development Network, Health, Nutrition and Population Series (Washington,

DC: \forld Bank, t999), p. 14.

Patterns in Reform Implementation' 1979-1999

a different sample of countries, with different regional aggregates' the

data divide overall expenditure into six broad categories. At this broad

t.u"l "f

generalitS Airican government expenditures appear to follow

allocational patterns roughly similar to those in other developing regions.

The one area in which that is not true is for the category of "cur-

rent transfers and subsidies" in which African governments spend less

than other governments, particularly once Africa's four middle-income

economies are putled out of the sample. This category combines trans-

f., p"yrr,"rr,, to individuals (i'e., for pensions and welfare) with subsi-

dies to firms (for example, subsidies for public enterprises). combined

with other data, however, it does offer tentative evidence that African

governments spend a smaller percentage of their overall expenditures on

i*prouir,g the welfare of the citizenry than do governments in other

,.g'io.r*. 6ood., for instance, indicates that in the early r98os' African

gi.rrr*..ta. devoted 3.4 percent of total expenditure to 'osocial-security

lnd welfare" compared iith t.e percent for Asia, v.7 percent for Latin

America, 8.r percent for middle Eastern states' aod 36'r percent for

industrial couirtries.,o For his part, Pradhan provides data suggesting

expenditures for this same category in Africa to be, on average' half the

level in Latin America and a third of the level in South Asia.es

This evidence accofds pretty well with my argument that most African

states are fairly autonomoos f,om social pressures' It suggests a state that

is not particularly responsive to societal demands and for which expan-

sion is an end in itse* rather thas a means to promote economic devel-

opment and PovertY alleviation. The evolution of public expenditures during the economic crisis

of the

r98os and r99os tends to confirm these patterns' Comparative data

across time also suggest that African state structures have weathered the

economic crisis better than their Latin American or Asian counterparts.

As a proportion of the economy, state expenditures have actually crept

upward by one or two percentage points of GDP over the course of the

l"st t*o de."der, according to IMF data for low- and middle-income

countries throughout the developing world.e6 In this sense' the evolution

94.RichardGoode,GouernmentFinanceinDeuelopingCountries(.Washington,DC: Brookings Institution, t984), p' 48'

g S. S""ty pl"dhan, .,Evaluating lr"Uiic Spending,"

.World Bank Discussion Paper no.

3rj lW"thington, DC:'World Bank, ry96)', 9e . iei,io, ir,.taL.,'.w.*id B"nk, woitd-Deuelop.ment

Indicators, 1998 (rufashing- '

ioi,DC: World'S^ii, ,ggtg),'Tables 4--rz and 4-t3.In ar least some cases, the

iir" i, e*pt"itt"d by a decline in GDP during this period'

9 4 9 5

African Economies and the Politics of Permanent Crisis, rg79-rg9g

of public expenditures in Africa does not look unusual. However, when you factor in the role of foreign aid, a different picture emerges. Between the late t97os and the late rggas, aid as a share of GDP has grown from an average of under 5 percent of GDP to well above a tenth of GDP across the countries of the region, compared to an average of under r percent of GDP for low- and middle-income countries in other regions of the developing world. Since the major proportion of aid takes the form of goods and services provided to states, in effect the size of government expenditures relative to the domestic economy is bigger than national statistics suggest, and has grown, not shrunk, despite two decades of dei- astating economic crisis.

The evolution of expenditure shares over the last two decades further reinforces the sense that most African states have responded to the crisis by looking out for themselves. As discussed above, government con- sumption has been protected more assiduously from austerity than was public investment, which collapsed in many countries. This leads to a stylized fact about adjustment in Africa: a cofttr?non std.te response to eco- nomic collapse appe*rs to haue been to protect its otun posi.ti.on and to lessen instead its deuelopmental arnbitions. The patterns of reform in the agricultural sector described in the previous section illustrate this dynamic. The state has withdrawn from the costly and often not very effective developmental apparatus it built up with donor support in the r97os. extension activities have been curtailed, as has feeder road con- struction. Most significant, the very costly fertllizer subsidies of the past have been eliminated in most countries. Typically, no pro-active policies for agriculture have replaced the old ones. The state's core services in the countryside have been reduced to virtually none, in many instances, unless there is a donor project in the area. Extension agents continue to receive salaries, but they no longer have a development budget and rely on the local administrative structure (the prefect, or police officer or perhaps a district officer) or NGO for such basic things as transporta- tion,eT But, on the other hand, as also discussed earlier, the state has often found it useful to retain a role in setting agricultural prices, refusing to accede to donor demands to completely abandon what it conrinues to view as an element of control over the sector. To call this "neoliberal reform" is to miss the point that the reform moment has been used by

9 7 . S u c h w a s t h e c a s e f o r a S e n e g a l e s e M i n i s t r y o f t h e l n t e r i o r s e r v i c e f o r r u r a l d e v e l - opment, which I visited some sixty miles sourh o{ Dakar in July 1998. The enrire development budget for a team of seven officials overseeing an arca of several hundred square miles amounted to jo,ooo CFA francs a year, roughly $roo.

P atterns in Reforrn Implementation, r 9 7 g- r 9 9 9

states to withdraw from their rural development responsibilities while retaining their ultimate ability to regulate the sector.

Much the same can be said about the social sectors. Many critics of

structural adjustment programs have argued that they impose high social costs as sfate retrenchment ends up hurting state programs in areas such as health and education. The evidence in Africa for this claim is nonethe- less mixed. Sahn and his colleagues have argued that government expen- ditures for health and education underwent only small cuts in the r9Sos and early r99os.e8 But other studies indicate a deterioration in social ser- vices in the region during this period, suggesting a less effective public

effort. Thus, school enrollment rates or vaccination services appear to be in decline in a large number of countries,ee

How can these anomalous findings be reconciled? Data quality may be one tactor, and there may simply be significant cross-national differ- ences. Nonetheless, the data are consistent with the interpretation that states have responded to fiscal austerity by cutting investment and devel- opmental activities much mofe than overall expenditures, increasingly focusing on recuffent consumption (such as salaries and perquisites). The retreat from developmental tasks is particularly striking in Africa's poorest countries, where governmental commitment to development was typically weaker even before the economic crisis and where fiscal aus- terity is particularly severe.

The evolving state role in Mali's education sector can serve as a telling illustration of these dynamics. In 1995, the World Bank estimated the gross enrollment rate at 3?- perce$t, up only slightly from z6 percent in r98o.100 Roughly a third of total enrollments were accounted for by private Koranic schools and other private schools, in rapid expan- sion, suggesfing fhe public sector provided educational services to barely a fifth of the school-age population. According to a r99z article by G6rard, privately run primary schools increased from z in 1985 to ro5 in early 1992, so that the public education sector appears to be

98. See Sahn et al., Structaral Adjustment Reconsidered, p. ro2. gg.

'World Bank, A Continent in Transition: Sub-Saharan Africa in tbe Mid r99os

($(ashington, DC: 'World

Bank, r995), estimates (p, z9) that gross enrollment ratios have declined during the r98os in twenty of the forty-four countries for which there are data. The decline can be directly linked ro civil strife in roughly a third of the cases. Similarln measles immunization coverage has declined or stagnated in thirteen countries.

roo. I(orld Bank, "Fact Sheet, Mali-Education Sector Investment Program (EDSIP)" at the

'World Bank web cite uut'u.t'uoildbank.org/picslpid/m140650

( r s s 8 ) .

9 6 9 7

African Economies and tbe Politics of Permanent Crisis, 1979-1999

receding.lol Enrollment rates vary significantly befween town and rural

areas. Thus, again in :r995, the enrollment rate varied between from 9o

percent in Bamako, the capital, to r r percent in the north region' This

low rate of enrollment resulted mostly from an insufficient number of

schools and teachers, even though the government has regularly allocated

some 20 pefcent of its national budget to education and has benefited

from substantial support from the donors. France has traditionally been

a significant source of money and educational policy ideas. In rg95'the

\(orld Bank agreed to undertake a $5o million education investment

program, the fifth of a series of loans to the sector since t962, It was

never disbursed, however, as the Malian government failed to meet the

loan's preconditions. Another similar education loan was being negoti-

ated as this book was being written in ry99. Although government funding declined slightly during the r99os, the

Malian government and the donors pushed through a series of ambitious

reforms, the main objective of which appears to decentralize primary

responsibility for education to the local level.102 Backed by the donors,

fhe government appeared to want to reduce its budgetary commitment

but retain a critical regulative role. One initiative consisted of providihg

partialfunding to parent associations to allow them to build schools for

pfivate instruction. To qualify for funding, the association had to provide

25 percent of the total cost, a figure G6rard viewed as prohibitively high

for most villages.1o3 How will the government maintain its regulatory

function as it withdraws from direct provision? Though a major archi-

tect of this policy the \florld Bank itself admitted that Mali's Ministry

of Primary Education is poorly equipped to play such a role, as it suffers

from ,,weak capacity for policy analysis and planning, overly centralized

decision-making, poor personnel and financial management. " 104 Clearly,

the numbers of nonteaching personnel will have to be increased consid-

erably if the state is to play this regulatory role effectively, and one is

allowed a degree of skepticism about the likely results of such an under-

taking, given the pasr record of donor-inspired institution building in

Mali's education sector. In the meantime, public educational expendi-

ror. Etienne G6rard, "Entre etat et population: I'ecole et l'education en devenir," Politique Africaine 47 ft992): 59-69'

roz. Ibid. ro3. Ibid, p. 66. Note, however, that a

'world Bank education specialist familiar with-

the pioject suggested this approach had worked reasonably well. Confidential interview, \0forld Bank, November 9, 1999'

ro4. Vorld Bank, "Fact Sheet," p. z.

Patterns in Reform Implementation, 1979-1999

Table 2.6. The Euolution of Primary School Gross Enrollment Ratios, , t 9 8 o - t 9 9 5

friAh Enr-ollment Countries Low Enrollment Countries

Botswana CongolZaire Tanzania Mauritius Cameroon Swaziland Lesotho Nigeria Kenya Togo Madagascar Congo, Republic of

+ 1 7 . 8 -20.2 -2s.7 + L 3 . 9 -9.8

+1.6.4 +5.9

- 1 0 . 8 -29.3 -1..3

-40.1. -:22.5

Burkina Faso Niger Mali Burundi Guinea Mauritania Djibouti Ethiopia

+2'J..3 +3.9

+14.6 +26.2 + 1 0 . 8 + 3 8 . 6 + 1 . 3 - J . U

Nofes: High enrollment deEned as a primary school gross enrollment ratio of 90 percent

or above in rg8o; low enrollment defined as a primary school gross enrollment ratio of

40 percent or below in r98o. The numbers in both columns are found by subtracting the

enrollment ratio in r98o from an average of available ratios for 1994-96.Some countries

are missing due to nonavailability of data.

Source: \Xloild Bank, African Deuelopment Indicators ('Washington, DC: World Bank'

zooo), p. 31o.

tures will be buttressed, even as the number of children educated by the

state continues to decline.

This evolution is not always quite as stark as in Mali. Two distinct

patterns can be discerned in national education statistics. In the poorest

African countries like Mali, enrollment rates have continued to creep

upward during the period of economic crisis, though less rapidly than

during the pre-crisis period, often with substantial assistance from the

donors. In the richer countries, on the other hand, in which virtually uni-

versal primary education appeared to have been attained before the

crisis, enrollment rates have declined over the last twenty years.tot This

evolution is demonstrated quite clearly in Table 2.6.

Of the countries with enrollment rates under 4o percent in r98o, only

Ethiopia's enrollment rate went down between r98o and t996.In sharp

ro5. These numbers should be treated with caution as most cross-national data on education are self-reported by governments and subject to wide differences in qualiry. See Joel Samoff, "The Facade of Precision in Education Data and Statistics: A Troubling Example ftom Tanzania," Journal of Modern African Studies z9 \r99tl 669-89.

l r J I'l'l'

l l l l ' r [ t l r :' l r r l d p l , : {}irj,' fr l F l j ',il

11'; rlrF! 4 [ r ,

9 8 9 9

i l ' I

African Economies and tbe Politics of Permanent Cri'sis, 1979-1999

confrast, eight of the twelve countries with rates over 90 percent in r98o

underwent a decline during this period. In addition, there is some evi-

dence that the quality of public sector educational services appears to be

declining, notably in rural areas.105 lnZambia, with an enrollment ratio

of s9 perc"nt in the mid-r99os, perhaps 75 percent of recent school

l"uu.rr-ur. functionally illiterate, as a result of the low qualrty of edu-

cation services. Given declining budgets, "the pupil/textbook ratio in

primary schools in r99z was 8: r in Mathematics, 5: r in English and

aboat zo: r in social studies" and conditions were so insalubrious that

many schools close during the rainy season to head off the threat of

cholera.1o7 In Mozambique, in the mid-rg9os, more than half of all

teachers had completed only primary school education'1o8

A recurring pattern throughout the region is the combination of a

growing do.ror role in developmental activity and the fetreat of the state

Ippu."Lr from what wefe once considered central developmental func-

tions. There is considerable variance across countries and across sectors'

but the best estimates suggest that somewhere between a third and two-

thirds of education and health services now completely bypass the state

and are the result of a combination of donor, NGO, and private sector

efforts.loe As Robinson and'White note, in many countries, there is a de

facto return to the situation during the colonial period when the state's

role in social sectol provision was peripheral to the effort of the Chris-

ro6, Ibid, see also Simon Appleton and John MacKinnon, "Enhancing Human- Capa.lti.s in Africa," in itetttro Ndulu and Nicolas van de.Walle (eds.), Agenda

foi afacar Econoruic Renewal (Washington, DC: overseas Development

'Council, ry96), pp. ro9-5o' especially pp' t-o-t'' For general

evidence

of growing-ro.ui'pou.ityi r.. eti and Thorbecke, "The State of Rural

PovertY." ro7.Seepp.r3r-4,'WorldBank,Zambia:Prospectsforsustainableand.Equitable'

Groith, R.pott no. fi57o-ZA (Washington, DC: \(orld Bank, August 23'

1 s g j l . ,os. p.i.t'ratlon and LuizPereirada Silva, "Recognizing Labor Market Constraints:

Government-Donor competition for Manpower in Mozambique," in David L'

Lindauer andBarbara ].{'unbetg (eds.), Rehabilitating Gouernment: Pay and

Employment Reforms in Africa (\Tashington, DC: I7orld Bank, ry94). p' 87'

,o9. S.. 1or.ph Semboja and Oie Therkildsen (eds.\, Seruice Proukion under Stress

' in i,astAfrica (London: James Currey, 1995); Meredeth Turshen' Priuatizing

Health Seruices in Africa (New Brunswick, NJ: Rutgers University Press, 1999),

especially pp. 86-9t; and Mark Robinson, ..Privatizing the Voluntary Sector:

N'COs as iobli. S..ui.. Contractors?" in David Hulme and Michael Edwards

(eds.), NGOs, States and Donors: Too Close for Comfort? (New York: St'

Martin's Press, 1997), PP. 59-78,

Patterns in Reform Implementation, r979-t999

tian missions.1lo The current fiscal crisis is leading many governments to

forsake the dominant postcolonial ambition to gerLeralize state pfovision.

In Uganda, for example, donors financed 77 pefcent of health spending

in flscal year t99zl93 while the government's share was 23 percent' In

the case of expenditures on hospitals, however, the government's share

was 64 percent, suggesting the often noted preference of governments in

the region for city-based curafive care that benefits the better off.11l In

Tanzania,58 percent of secondary school enrollments are accounted

for by NGO and church organizations' up from 28 percent in 1965' In

zimbabwe, church missions provide more rhan two-thirds of all hospi-

tal beds in rural areas.tlz

S ou er eignty ExP enditur e s

'Ihe retreat from external obligations is a fairly standard governmental

response to economic austerity, one that is far from unique to Africa.

More distinctive, howevel, has been the rest of the political fesponse.

Even as African states have withdrawn from development tasks and as

salaries for public servants have been allowed to decline precipitously in

real tefms, the top of the state appafatus often has retained conspicuous

consumption pattefns. Thus, if the previous paragraphs have suggested

that the entire state apparatus has been relatively protected from the

ravages of the crisis, it is important to underscore the fact that within

the state, there also appear to have been relativewinners and losers' In

particular, spending patterns suggest a clear shift in spending away from

ih. lo*"t rungs of the civil service to its highest levels and to the polit-

ical elite at the top of the state. I am speaking in particular of what the

French call expenditures "regaliennes," literally translated as regal

expenditures, but which can better be defined as expenditures linked to

the exercise of state sovereignty. These include defense spending, partic-

ularly when it is not related to a direct external threat, as well as

rro. Mark Robinson and Gordon r$fhite, "The Role of Civic Organizations in the Provision of social Services," United Nations University !flIDER Research for

Action Paper no. 37 ft997), see pp. 9-r J. rrr. Emmanu.i eblo and Ritva Reinikka, "Do Budgets Really Matter? Evidence

from Public Spending on Education and Health in uganda,"'wodd Bank Policy Research Papir (ITashington, DC:

'Wodd Bank, undated), P: t4' ,.

rrz. These last two statisticJare provided in Robinson and \(hite, "The Role of

Civic Organizations"' PP, rL-r3.

I O I

African Economies and the Politics of Permanent Crisis, rg79-rg99

expenditures for infernational diplomacy) government offices, and other activities that benefit an extremely small stratum of the population. They are all closely tied to the exercise of national sovereignty, be it in sym- bolic or ceremonial fashion, but they do not have a developmental func- tion and they are not necessarily linked to the exercise of state power, although they do play a role in mainraining political stability.

It is interesting to examine the evolution of these sovereignty expen- ditures over time, since donors have traditionally shied away from engag- ing in conditionality over them, despite their very evident impact on the government's ability to finance development. Although the available data are unfortunately very weak, they suggest that these kinds of

'expendi-

tures have been remarkably unaffected by economic crisis. Defense expenditures are perhaps the most obvious type linked to sovereignty. The best available data are published by the Stockholm Inrernational Peace Research Institute (SIPRI). They show that the twenty-one coun- tries in the region for data are available spent an average of z percent of GDP on defense, quite a modest amount and probably an underesti- mate.113 SIPRI also provides much better coverage of the military indi- cator perhaps least amenable to manipulation: the number of soldiers in the armed forces: Overall, for the forty countries on which it reports data, the number of soldiers grew by an average of 67 percent between 1985 and r9g6.tto Recognizing the great variety of defense needs across African countries between these two dates, I excluded six countries that were involved in significant military activities during this period (Angola, Burundi, Rwanda, Senegal, Sierra Leone, and Uganda) as well as three countries in which defense needs enormously declined during the period (Ethiopia, Mozambique, South Africa) and recalculated the rare of increase. For the remaining thirty-two counfries, the number of soldiers still increased by z8 percent during this period. For every country like Ghana (a 54 percent decline) or Tanzania (down 14 percent) that had decreased the number of soldiers to save money in a context of fiscal crisis, there were states like Gabon (96 percent increase), Zambia (up Zl percent), or Togo (up g+ percent) that had continued to increase the number of soldiers despite no apparenr external military threat.

r13. SIPRI numbers, cited in the ry99 edition of the United Nations Development Program's Human Deuelopment Report (New York: Oxford University press, 1999), pp. r88-9r. African governmenrs are widely believed to underreporr their defense expenditures, which moreover benefit from sizable foreign assis- tance, notably through subsidized arms sales.

rr4. Ibid. IINDP reports index numbers with a base year of 1985.

Patterns in Reform Implementation, t979-1999

Another striking if imperfect piece of evidence regarding the continu-

ation of sovereignty expenditures by African governments is offered by

the evolving size of government cabinets. I collected data on cabinets at

three points of time: April{une ryTg,JulylAugust 1988, and August/Sep-

tember ry96.Table 2.7 reports these data, with regional averages for the

three dates.115 Thus, it shows that in 1979, the average cabinet in the

region included r9.r ministers and that this average had climbed to zz.6

by ry96. These averages disguise significant variation, and I have also

included data on the biggest and smallest cabinet at each point. Each

data point constitutes a snapshot in time. In some cases, the cabinet size

recorded underwent big changes soon afterward - because of a military

coup, for example; because certain positions were demoted; or because

other prestigious governmental bodies (advisory economic and social

councils, politburos, a committee for national salvation, etc.) were estab-

lished or eliminated. The data are nonetheless instructive about general

trends. In most countries, a ministerial appointment is an important position

for a member of the political elite, bringing with it patronage opportu-

nities as well as significant perks and status-enhancing privileges. African

states have long been notorious for their large cabinets, with ministerial

appointments that often have little relevance to policy-making priorities

or the size of actual budgets. Large cabinets are rarely a response to eco-

nomic policy concerns, of course. TypicallS they seek to facilitate com-

promise and cohesion within the political class. Lijphart has shown a

strong positive correlation in Western democracies between the number

of parties in government and the size of the cabinet.116 Coalition gov-

ernments almost invariably have larger cabinets. In Africa, similarly,

ethnic diversity appears to have increased cabinet size. A simple corre-

lation between the size of the cabinet in each of the three dates in

rr5. These totals did not include the head of state even if he or she presided over the cabinet. It did include all positions accorded cabinet or council of ministers rank by the head of state, including deputy ministers and secretaries of state. To minimize problems of comparability, it did not include provincial, prefectoral, or regional ministers, even if they were accorded cabinet status. It also did not include other consultative executive bodies that existed in a number of governments, rypically soon after a military coup. Thus, Somalia's five-member politburo is not included in that country's total for r97g; nor is Nigeria's twenty-eight-member Armed Forces Ruling Council in 1988, or Madagascart tlventy-fwo-member Supreme Revolutionary Council at the same date.

r16. Arend Lijphart, Den'tocracies: Patterns of Majoritarian and Consensus Rule in Ttuenty-One Countries (New Haven, CT Yale University Press, 1984).

i * t , , .

I i t i , l

la i lr!:

Fi''' F r l i i ' 1 I

lr'''

. i i { l

. i 1 , .

l,'lr ;lt li+: ilF |jt,

i!

r02 r o 3

African Economies and the Politics of Permanent Crisis, 1979-1999

Table 2.7. African Gouernttent'Cabinet Size, r979-t996

G a m b i a ( 1 1 ) G a m b i a ( 1 3 ) Gabon (35) Gabon (54) C6te d'Ivoire (31) Cameroon (34)

o Micro states include all states with population under one million (Cape Verde, Comoros, Dfibouti, Equatorial Guinea, Sao Tome, and Seychelles). Category of Smallest Cabinet also excludes them.

Source: Data from Europa, Africa South of the Sahara (London: Europa Publications Limited, 1979, 1986, and ry96); data for 1979 most1y from April 1979 (N = +6); data for 1986 mostly from June 1985 (N = 46); data lor 1996 mostly from September 1996 (N =

47). Totals do not include heads of state if they sit in cabinet; does include deputy minis- ters, secretaries of state, and other subordinate cabinet posts, if they are clearly listed as having cabinet status. The totals do not include provincial governorships or prefects.

Table 2.7 and the country's level of ethnic fragmentation reveals a high

positive correlation.llT It is likely that leaders seek to preserve national

unity by including elites from as many salient ethnoregional groups as possible. Large cabinets may thus hold a political function. Certainly, it is difficult to imagine what policy advantages they could advance; large cabinets complicate policy making, posing significant coordination prob- lems within government. Certainly, African cabinets are rarcly convened for much more than ceremonial functions.

Cabinets typically have significant expenditure implications and have high opportunity costs in the typically small budgets of African states laboring under fiscal austerity. C6te d'Ivoire, for instance, the third biggest economy in the region after Nigeria and South Africa, had a total national budget in ry96 of U.S.$A.S billion and a cabinet including thirty-two ministers and four deputy ministers. A more typical state budget in the region is probably under a billion dollars in total expen- ditures ̂ yea\ including debt servicing obligations. For the smaller min- istries, the recurrent cost obligations of maintaining the minister's office constitute a significant component of the available budget. It might be noted that the fastest growing country in the region, Botswana, had cab-

rr7. Simple correlation coefficients were estimated between cabinet size in 1979, t986, and ry96 and, a measure of ethnic fragmentation. The correlations were, r e s p e c t i v e l y . 5 5 ( p = . o o o t ) , . 4 4 ( p = . o o o 3 ) , a n d . 4 6 ( p = . o o o z ) .

Patterns in Reform Implementation, t979-r999

inets of eleven to fourteen members during this period, while European

countries often have cabinets of fewer than two dozen members, though

they oversee national budgets considerably larger than the biggest in

Africa. Throughout the region, the cost of maintaininglarge cabinets has been

compounded by fairly rapid turnover of ministerial seats. Ghana, for

instance, had eight ministers of finance between r97o and r9go, twelve

ministers of education, and twelve ministers of agriculture.ll8 A large

number of cabinet reshuffles can be due to political instability and the

turnover of heads of state. More often this shifting appears to be

sympromaric of the management of the political class by the head of

state. Shuffles expand the number of individuals who can benefit from

ministerial largesse, allow the president to prevent the emergence of

rivals from within the cabinet, and inspire loyalty from ministers who

are reminded that their presence in the cabinet depends on presidential

goodwill. Whatever the reasons, it is not unusual for ministers to retain

certain perquisites and privileges fol a couple of months after they

have left office, again adding ro rhe fiscal burden of large, unstable

cabinets. The evidence from Table 2.7 suggesrs that cabinet size has held steady

if not actually increased since the beginning of the debt crisis. This evi-

dence is once again instructive regarding the persistence of government

consumption privileging the higher levels of the political class with little

benefrt to the state's developmental functions. Even as the investment

budget has withered and key developmental fasks have devolved to

donors or been abandoned, the luxury of the big cabinet has been sus-

tained. Indeed, recent events in Kenya demonstrated the degree of attach-

ment to this luxury. There, the cabinet has slowly but surely grown from

around twenty members to twenty-seven over the last fifteen yeafs. In

mid-l 999rMoi responded to pressures from a donor community increas-

ingly impatient with the corruption of the regime by cutting the number

oiministries down to fifteen, but he chose to retain all twenty-six of his

colleagues within the cabinet, presumably with their fu|[ salaries and

b e n e f i t s . l l e Unfortunately there are few other system atic data on this kind of gov-

ernment consumption that one can examine to back up the claim that

rr8. See Nicolas van de Walle and Timothy Johnston, Improuing Aid to Africa (Bal'

rimore: Johns Hopkins University Preis, for the Overseas Development Council,

r y 9 6 ) , p . 8 6 . --9. S.. joh,' Ny"g", "Ridicule and Praise Follow Casualty Free Reshuffle in-

Kenyar" Agence France Presse news release, September 7, 1999'

1979

Average Number of Members

- excluding microstates* Smallest Cabinets"'

Largest Cabinets

1.9.1. 2 0 . 3 G h a n a ( 1 1 )

20.9 22.0 Namibia (8)

22.6 2 3 . 9

Gambia (13) Lesotho (13) Cameroon (42) Sudan (38)

' ! r l

x { , - ! 1 ,

l : r l . . t

d l " '

i i l i Iqii

:l '. ' i'1'{ r l t i i

r l l . il.-

r o 4 r o 5

African Economies and tbe Politics of Permanent Crisis, 1979-1999

the crisis has not stopped excessive spending by senior state decision makers. The tendency of cabinets to increase in size is mirrored by an increase in the size of other national bodies. Legislatures, another impor- tant site of elite accommodation politics, grew from an avetage of 89.5 seats across the region following the first postindependence election in the early t96os, to r4o seats in t989, and, to r49 in t997.120 Legislators are probably not as expensive to maintain as ministers, but they too con- stitute a burden in a small budget. In mid-t999, the Nigerian legislature was embroiled in controversy with the revelation that each of that country's 593 new legislators would be granted a furniture allowance of up to 3.5 million Nigerian naira (roughly U.S.$: S,Soo), an impolitic sum of money in a country in which the average civil servant brought home about $200 a month.121

Presidential commissions and advisory councils have tended to pro- liferate, as have the number of administrative districts, municipalities, and provinces. In Kenya, President Moi created the Presidential Com- mission on Soil Conservation and the Presidential Commission on Music in the early r99os, both of which had permanent staffs and secretariats, despite having terms of reference that directly ovedapped with, respec- tively, the Ministry of the Environment and Natural Resources and the Ministry of Culture.122 But the champion of this kind of multiplication of offices is perhaps C6te d'Ivoire. Its legislature has regularly expanded, from 7o members at independence to rzo in r97o to r.47 in r98o and

ry5 in 1985. Over time, the Economic and Social Council has been expanded from z5 members to r2o, the number of administrative com- munes from 34 to 196 (another expansion is planned to 365), and the number of departments from 4 to 49,123

rzo. The first two numbers are cited in Bratton and van de lffalle, Democratic Exper- iments, p, 7r.The numbers for ry97 were calculated by the author from various sources. They concern lower houses of parliament and exclude ex officio members.

rzr. "Obasanjo's One Hundred Days," Africa Confidential (Augusr 27, 1999): t-3. The Economis, relates other shenanigans in the Nigerian legislature in "Self Service for Nigerian Senators," The Economisr (August rz, zooo): 4r.

rzz. See lVorld Bank, Kenya: Re-Inuesting in Stabilization and Growth through Public Sector Ad.justment (Washington, DC: V/orld Bank, r99z), p. 47.

rz3. On the Ivoirian administration and its political functions over time, see Richard Crook, "Patrimonialism, Administrative Effectiveness and Economic Develop- menx in C6te d'Ivoire," African Affairs 88 (r98fl: zo5-28; Richard Crook and James Manor, Democracy and Decentralization in Soutlt Asia and West Africa: Participation, Accountability and Performance (New York: Cambridge Univer- sity Press, 1999); and Tessy Bakary, "C6te d'Ivoire: I'etatisation de I'etat," in

Patterns in Reform lmplementation, r979-t999

One could multiply the examples of African governments maintain-

ing costly practices for top elites amid austerity budgeting: in one

country, the number of generals in the army is in constant progression

despite no obvious foreign threatlza in another, the government spends huge sums preparing the capital city for aflOrganization of African Uniry

summit;12s and in yet another, the senior civil service maintains an out- landish travel budget,126 One could also mention in this context, the con-

tinuation of the practice of presidential slush funds that are transparently designed to promote political stability. Thus, President Chiluba of

Zambia had the legislature approve a presidential fund of rz billion kwacha in ry98 (approximately 6-8 million), most of which was dis-

tributed at his discretion, notably to soccer clubs, church groups' and hospitals.127 Finally, and in much the same spirit, what should one make of the distribution of honorifrc medals in Cameroon during a national holiday ceremony in May 1989? As related by Achille Mbembe, slightly over 3rooo eminent citizens were rewarded with gold, silver, and

dark-red decorations, at a cost of 3o million CFA francs, or roughly

U.S.$rro,oo.128 Thatyear, r98g, was a disastrous one for the Cameroon economy, with a 6 percent decline in GDP.

But perhaps the most striking such practices concern the expenses that presidents have continued to lavish on themselves' even as their nation's

fiscal crises grew in significance. Presidents all over Africa are notorious for the presidential palaces they have built for themselves, albeit usually

Jean FranEois M6dard (ed.), Etats d'Afrique Noire: Pormation' mdcanismes et crises (Parts'. Karthala, 1994,pp. 53-9r.

tz4. In Cameroon, President Biya more than doubled the number of generals from seven to fifteen in 1993 in response to the severe internal instability his regime was then facing. See Celestin Monga, "R6gime militaire i visage civil," Jeune Afrique Economie (March ry97)': r.r8.

rz5. Togo is thus said to have spent $6-7 million to convene a t997 meeting of the OAU on resolving the Congo crisis. See "Remettre le Togo en Marche," Marchis Tropicawx et Mdditdrrandazs (April 4, t997\ p. 706.

rz6. InZambia, the government allegedly approved z7,ooo nights abroad by senior Zambian civil servants in ry97, Many of these nights out of the country were paid for by donors, but the same source claimed that the government spent seven billion kwacha that year on international travel, roughly $5 million in that yearb very volatile exchange rate (confidential interviews, Ministry of Financeo Lusaka, M:ay x999). This number may not seem outlandish when it is considered that the Zambian government sent seventeen officials to the May 1999 Consukative Group meetings in Paris.

rz7. Barraclough, "Trouble in Lusaka." rz8. Achille Mbembe, "Provisional Notes on the Postcolony," Africa 6z (x992)t

z - i 7 , P ' 2 7 '

i l , ,

i l i . t

l 1 | ! 1

|,l l;i

l l l l r i . i i ,

I l l l r - I t|il'],

l1 !!

, l i '

i l . ' ! r L

f f t , l - .

It f'ts |it' 4F

T o 6 r 0 7

African Economies and the Politics of Permanent Crisis, 1979-1999

officially with their own personal fortunes. Paul Biya had a personal airport built at his own presidential rctreat of Mvomeka'a in the early r99os, even as his government was disregarding \7orld Bank condition- ality and building a new international airport in Yaounde, the capital.tze \(hen he traveled abroad for official functions in the mid-r99os, Zimbabwean President Mugabe's entourage routinely included more than three dozen officials, assistants, bodyguards, and family mem- bers.130 In ry99 alone, Mugabe reportedly made state visits to sixteen countries in fourteen weeks outside the country. Given an entourage. of some Iorty officials on these trips at an average daily allowance of

'

ro)ooo Zimbabwean dollars (4o zim$ = r U.S.$), these 1999 trips cost the Zimbabwean taxpayer some U.S.$98o,ooo, not including airfares and incidentals.l3t Not to be outdone, Omar Bongo of Gabon made a visit to Europe and the United States in April-May ry99 with more rhan seventy people. One objective of the trip was ro gain symparhy for Gabon's worsening economy and lobby Western gove(nments for debt relief.l32

The most egregious excess by an African leader is surely the con- struction of the Basilica of Our Lady of Peace of Yamoussoukrou by President Houphouet-Boigny in his native village. A replica of Saint Peter's Basilica in Rome, albeit twice as big, containing the equivalent of thirty acres of marble and fully air conditioned, the church is said ro have cost several hundred million dollars. Officially, it was built entirely with the president's personal fortune, the ostensible reason for which the donors kept silent about the church's construction in the late r98os. During this same period, the donors agreed to reschedule Ivoirian foreign debt on an almost annual basis, allowing it to grow quickly to $rz billion by t994, the year of Houphouer-Boigny's death. It might be added

rz9, Since his predecessor, Ahmadu Ahidjo, had built an airport in his own home town of Garoua in the late r97os, Cameroon can now claim four airports able to land wide-bodied jets. The oldest and busiest airport remains the one in the country's economic capital, Douala.

r3o, As I was able to ascertain first when he came to Michigan State University to receive an honorary degree in 1992, and second at the Summit of the Global Coalition for Africa in Maastricht, rhe Netherlands, in November r995. These travel expenses, which have apparently come to be criticized by the press in Zimbabwq are described, with other examples, in a PANA news release, .'Zim-

babweans Question Mugabe's Travelling Show," October 19, 1999. r 3 r . I b i d . r3z. Economist Intelligence Unit, Country Report for Gabon, Thfud Quarter, 1999

(London: The Economisr Intelligence Unit), p. rr.

P atterns in Reforrn Implementation, r 9 7 9- t 9 9 9

that between rg8o and rgg5, personal income in c6te d'Ivoire fell by

5 o p e f c e n t , 1 3 3

It may be remarked that these scandalous examples are disingenuous

and that all governments of the wodd spend what seem to be large sums

on sovereignty expenditures, but that they are unimportant, relative to

the size of national budgets and given their political role in maintaining

sability. Their political importance will be assessed in later chapters, but

here it can be suggested again that in the small economies of Africa,

where budgets are also small, sovereignty expenditures do have a vetY

real opportunity cost in terms of development expenditure. Thus' a 1993

\forld Bank report noted that "every year since 1984,Zambia has spent

mofe on its diplomatic services than on 'land, water and national

resource development' (over 2.55vs. z,o%" [of the budget])' The cost of

maintaining Zambianforeign services abroad were more than how much

Zambia spent on the provision of primary school services (net of teacher

salaries) in the entire country."134

C O N C L U S I O N

It is now possible to summarize the pattems of implementation of policy

refofm over the last twenty years. In broad general terms, there has been

more progress on stabilization policies than on the more complex agenda

of institutional and structural change. On the one hand, this can be

imputed to the technical difficulties of undertaking real lasting change.

Much of the stabilization agenda is what are often called "stroke of the

pen,' reforms, which can involve only a handful of officials in a country's

ministry of finance of the central bank for successful implementation. A

reform like privatization, however, involves a host of public and private

actors, and the technical requirements of components like company val-

uation almost inevitably take time and arc subject to controversy. This

is true of most adjustment policies. Even when defining the parameters

of policy reform is relatively straightforward, as in much trade reform

and many price liberalization policies, implementation of the policy is

complex and difficult. Thus, we witness alarge and growing gap between

the reform that is in the books and the reality on the ground.

On the other hand, differences in implementation should not be

imputed solely to technical factors. First, fiscal considerations have also

r33. See 'olvory Coast Buries Founding Father," Neu York Times,Febtuaty 8, 1994, p . 4

t14. Zambian Public Expenditure Reuiew, p. 55.

I

1 '' 1 " ' l '

h{ l1t:

t l l i l

lll' it u'

i,i'*'

;;j, I r l i ,l;l:

l:: ilt: ,jp.. I irr't'f'

I L

1 i ; I ll'

l r

r o 8 t09

African Economies and the politics of permanent Crisis, r979-rg99

played a key role. As the crisis has unfolded, the fiscal implications of policies has become a key factor in the willingness of states to maintain or change them. Thus, the reticence to undertake pivatization reform has to a considerable extent been overcome by the lure of revenues from the sale of public enrerprises (pEs). similarly, the withdrawal of fertilizer subsidies in agriculture has been perhaps the mosf consistent element of reform in the agricultural secror largely because of their budgetary impli- cations. The elimination of food subsidies in east Africa follows the same logic, while much of the rest of the prereform agricultural policy frame- work has remained in place, because its cost to the treasury is less consequential.

\fhen reforms do not have a net positive fiscal impact, they also appear more vulnerable to delays and reversal. This is true of trade policy reform, for example. It will be noticed that fiscal and rent-seeking inter- ests coincide and interact. Trade policy reform is difficult because srates continue to rely on tariffs for revenue, which still amounted, on average, to over 6 percent of GDP in the region in the r99os.i35 Ar the same time, uncollected tariff revenues plague many governments and help perperu- ate the fiscal crisis. Much the same could be said of pE reform, where privatization sales have been charact erized by nonrransparent sweetheart deals that have lessened their benefits to states. In these cases, rent- seeking interests compete with revenue exigencies, often delaying reform or creating situations of ambiguity about the effective status of reform, a climate of uncertainty this chapter has noted on a number of occa- sions. state elites face a clear ttade-off here and, as argued in the next chapter, as the crisis has continued and state capacities have declined, fiscal discipline has clearly waned. In any evenr, the point is that much ofthe reform process often appears to be less the product ofcareful strat- egy on the part of the state and more of a fire sale mentality in which the search for easy budgetary gains outweigh other consiclerafions.

Relatedlg I have argued that political factors have been important constraints on implementation, although not in the manner in which they are usually viewed. Policies that might have been considered politically difficult because they directly affected powerful consrituencies have in fact been implemented..we saw that governments have allowed civil ser- vants to lose much of their purchasing power. Marketing boards have been closed with large job retrenchments. Even more striking, we saw

135. The International Monetary_Fund, Tax policy Hand.book (\fashingron, DC: International Monetary Fund, r99g), Table 3.

Patterns in Reform Implementation' t979-t999

rhat consumer food prices have been almost entirely liberalized, includ-

ing in east Africa, where the link between controlled maize prices and

political stability was long an adage of faith.

Political factors assert themselves in other ways. This chapter has

shown that the most plausible interpretation of implementation patterns

is the desire of the state to protect itself from fiscal austerity. Of course,

governments do not operate in a vacuum, and they worry about the reac-

iiotts of different groups of citizens to their actions. In some countries,

some groups have an evident impact on policy making in certain areas.

Farmer organizations in east and southern Africa were identified as

having had a significant impacr borne of a long legacy of institutional

linkages. Governments also fear certain groups more fhan others. I noted

that the accumulation of salary afreals has characterized much of.Fran^

cophone Africa. Interestingly, the evidence from these countries suggests

that arrearc invariably are higher for civil servants in the countryside

than for those in the cities, and that arlears are lowest for the police and

military staff. These divergences of treatment clearly constitute political

management of the arrears situation and suggest that governments are

aware that there are limits to how far they can push certain gfoups.

Nonetheless, far more striking is the extent to which governments

have worried less about protecting certain societal constituencies than

they have in protecting state elites from austerity, generally, and the

higher levels of the state elite more specifically. As fiscal pressure has

"ss.rted itself, budgets have sacrificed an increasing proportion of the

state's developmental activities. First, it skimped on the investment

budget, then, increasingly on core social services, as basic health and

education services were devolved to the private sector and donors. In

much of Africa today, pafticulally in the poorer countfies' the tendency

is for the state to do little besides pay for a civil service that is no longer

capable of actually promoting development. AdmittedlS this trend is

further along in the poorer states of west Africa than it is in the richer

and more institutionalized states of east and southern Africa, but even

in a gountry like Kenya, we can witness the state withdrawing from

service provision. Governments have sought to accommodate themselves to the fiscal

crisis by reducing their developmental obligations, but at the same time,

they have wanted to retain control over key markets. Thus, they have

resisted donor demands that they give up the prerogative to regulate key

prices in the economy, notably in the politically sensitive ̂tea of food.

Th.y ,.t ."t from the provision of education or health services, but

f 1 , ,

l,' l r : ' ll1 ll.

r i l I i

1 1 , ,

t i l ! i

t l l I

1 i 1 , .

11:i,

,,1* ln |1:, itr, tt' r l .

I F

r I o I I I

African Economies and the Politics of Permanent Crisis, r97g-r999

continue to want to regulate, license, and tax these sectors with the tech- nical assistance of the donors.

On the other hand, even as developmental ambitions have been cur- tailed, what I have called sovereignty expenditures have been little affected by the crisis, and the top of the state apparatus continues to benefit from a level of resources that seems impervious to the crisis. Support for the military seems stable, while high-level positions in the cabinet or legislature appear to be continuing their proliferation. osten- tatious expenditures at the very peak ofthe state also continue to be parr of the landscape, amazingly impervious ro Africa's economic decline.

Deci.sion Ma

The Nigerian government

limit and regulate foreign

mitment to what came to

Second National

implemented through the

ry72 and ry77.IheY economy would exclude would tolerate them, as increased to at least 4o perception that ten Years of

dominance of both

Eastern (mostly Lebanese) plained they were discnmtn

to finance their exPansion

companies were not rel

exert greater control over

industrialization. The impact of the two

would affect well over a

wholly transferred either t

public sector' Lebanese the use of front men, vari

helped many survive. Im

r. This account is informed State and Control of the Press, 1987); see also lfilli 1988), pp. 53-9, andTom

r ] . z

(Boulder, C0:'Westview