understanding the rural--urban voting patterns in the 1992 ghanaian presidential election. a closer...

Understanding the Rural--Urban Voting Patterns in the 1992 Ghanaian Presidential Election.A Closer Look at the Distributional Impact of Ghana's Structural Adjustment ProgrammeAuthor(s): Mahamudu BawumiaSource: The Journal of Modern African Studies, Vol. 36, No. 1 (Mar., 1998), pp. 47-70Published by: Cambridge University PressStable URL: http://www.jstor.org/stable/161637 .

Accessed: 09/05/2014 12:59

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to TheJournal of Modern African Studies.

http://www.jstor.org

This content downloaded from 195.78.109.147 on Fri, 9 May 2014 12:59:54 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=cup

http://www.jstor.org/stable/161637?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


The Journal of Modern African Studies, 36, I (I998), pp. 47-70 Printed in the United Kingdom ? i998 Cambridge University Press

Understanding the rural-urban voting patterns in the 1992

Ghanaian presidential election. A closer look at the distributional

impact of Ghana's Structural Adjustment Programme

Mahamudu Bawumia*

This article attempts to explain the rural-urban voting patterns in the I 992 Ghanaian presidential election. In this election, rural voters voted overwhelmingly for the incumbent and urban voters did the opposite. It is argued that Ghana's Structural Adjustment Programme ( I983-92) was distributionally favourable to rural households and unfavourable to urban households. A link is therefore drawn between the distributional impacts of the Structural Adjustment Programme and the voting patterns of rural and urban households.

The relationship between the state of the economy and the fortunes of political parties at the polls is one which has generated a lot of debate.' This debate has largely taken place within the confines of Western democracies, not least because of the absence of Western-style democracy in many developing countries. We are, however, seeing a movement towards 'democracy' in many developing countries, with pressures for economic liberalization going hand in glove with those for political liberalization. The increasing democratization by many African countries undertaking Structural Adjustment Programmes provides us with an opportunity to investigate the relationships between the welfare implications of these programmes and the voting behaviour of the electorate. Is voting behaviour in Africa any different from that in Western democracies?

* Assistant Professor of Economics, Hankamer School of Business, Baylor University, Waco, Texas, 76798-8003, USA. I would like to thank Professor Mahmood Hasan Khan, Professor Peter Kennedy, Professor Wayne Nafziger, seminar participants at Simon Fraser University, and two anonymous referees for very helpful comments and suggestions. All errors are mine.

1 See Peter Nannestad and Martin Paldam, 'The VP function: a survey of the literature on vote and popularity functions after 25 years', Public Choice 79 (I 994), 2 I 3-45.



48 MAHAMUDU BAWUMIA

TABLE I

Percentage of votes cast for Rawlings in the I992 election by region, and major urban constituency

Rawlings' share Difference Rawlings' share in major urban (I) - (2)

Region in region (I) constituency (2) +

Western 60-7 33-5 27-2 Central 66-5 45 9 206 Eastern 56.7 4I'5 I 5-2

Volta 96-2 94'I 2- I

Ashanti 32 8 I 6 8 I6-o Brong-Ahafo 6i 5 40?3 2 I*2

Northern 62-5 4I52 25I3 Upper East 5o 8 39.8 I Iso

Upper West 5IP? 55 9 -4'9 Greater Accra 5i 8 4I57 IOI

Source: Ghana Electoral Commission, The Presidential Election, 1993. Regional and constituency results

(Accra, I993).

After many years of economic decline, Ghana embarked on a programme of economic recovery in I983, under the auspices of the International Monetary Fund and the World Bank. Nine years into the implementation of this programme, the country was returned to civilian government, when the ruling Provisional National Defence Council (PNDC) under Flight Lieutenant Jerry Rawlings formed a political party, the National Democratic Congress (NDC), to contest presidential and parliamentary elections.2 The results of the presidential election suggested that Jerry Rawlings derived his support largely from rural constituencies.3

The I 992 results, as presented in Table i, indicate that even though Rawlings won the majority of the votes in each region (except Ashanti), he lost in many urban constituencies: Bolgatanga and Navorongo (Upper East) Koforidua and Nkawkaw (Eastern), Sunyani East (Brong-Ahafo), Yendi (Northern) and Sekondi and Takoradi (West- ern). In the Central region, Rawlings won in Cape-Coast (the capital) but with only 45-85 per cent of the vote.

The results show a rural-urban dichotomy in voting behaviour, with the rural areas generally voting for the incumbent (Rawlings), and the

2 For an overview of these elections, see Richard Jeffries and Clare Thomas, 'The Ghanaian elections of I992', African Afairs 92, 368 (2993).

3 The parliamentary elections were boycotted by the opposition parties after they alleged that the presidential election was rigged. The presidential election was, however, declared by an international group of monitors to have been largely 'free and fair'.



VOTING PATTERNS IN GHANA 49

TABLE 2

Average annual %change in selected basic indicators (I 970-1983)

Annual rate of growth

Indicator or decline (%)

Population 2-6 GDP -o-6 GDP per capita -3-2

Agriculture -0'2

Industry -42

Services I-5 Exports -6-2

Imports -9-0

Terms of trade - 37 Cocoa output -6-i Cereal output - 34 Starchy staples -4*7 Inflation 4-6 Minimum wage - I' Average earnings - s6- Cocoa-producer price - II-3

Source: Tabatabai, 'Economic stabilization', i 986.

urban areas doing the opposite. Table indicates that in every region apart from the Upper West region, Rawlings's total share of the votes was higher than his share in the major urban constituency in that region. This article argues that this difference in rural-urban voting patterns can be attributed to the distributional impacts of Ghana's Structural Adjustment Programme (SAP) on rural and urban households. The rest of the article will first examine the distributional impact of Ghana's economic decline in the decade before its adoption of the SAP in 1983. It will then examine Ghana's SAP and its distributional impact between 1983 and 1992. Finally, the voting patterns observed in the I992 election will be linked to the distributional impact of the SAP.

THE DECADE BEFORE STRUCTURAL ADJUSTMENT, I972-1I983

The period between I972 and 1983 is important because it was the decade immediately preceding the 1983 Structural Adjustment Programme. This period was one of continuing deterioration in the economy. For the most part, the policies of this period emphasised import substitution, underpinned by a restrictive foreign exchange regime, quantitative restrictions upon imports and price controls, with



50 MAHAMUDU BAWUMIA

the state playing a major role as producer. As shown in Table 2, the dramatic economic decline between I970 and i983 encompassed a decline in GDP per capita by more than 2 per cent a year, in industrial output by 42 per cent a year, and in agricultural output by 02 per cent a year. Real GNP per capita declined from 520 cedis in I974 to 320

cedis by i983 (Table 3).

The main foundation of the economy, cocoa, was in decline. Although in i983 food production was affected by the worst drought in Ghana's history, the decline was probably due to the massive migration suffered by the rural sector.4 This exodus was partly a result of the deteriorating economic conditions, and also of the I973/4 oil boom in Nigeria which induced more than 2 million Ghanaians to leave in search of greener pastures in Nigeria. Particularly hard hit was the government's tax base, as those activities that provided it with the bulk of its revenues shrank disproportionately. Central government revenues, which amounted to 2I per cent of GDP in I970, fell to only 5 per cent of a smaller GDP in i983.' The revenue collapse increased the reliance on the banking system to finance expenditures. Between I974

and i983 the monetary base expanded from 697 million to I,440

million cedis. The loss of monetary control accelerated inflation, which increased from i8-5 per cent in I974 to I I6-5 per cent by i98i in the midst of a regime of controlled prices (Table 3). The period of decline was also characterised by negative real interest rates, and domestic savings and investment decreased from I2 per cent and I4 per cent of GDP respectively to less than 4 per cent.6

In the meantime, successive governments continued the policy of overvaluing the cedi. Between I974 and i983 the Ghanaian cedi had been devalued only once (in I978, from i-i5 to 2-75 cedis to a US dollar), despite a hundred-fold increase in domestic prices. The current account deficit of US$ 27 million in I975 increased to US$ 294 million by I983 (Table 3). The current account deficits not only depleted gross official foreign reserves but also involved an accumulation of external debts. Arrears amounted to the equivalent of go per cent of annual export earnings in I982.7 Successive governments responded with import controls, which fell disproportionately on consumer goods. Consumer goods as a proportion of imports fell from 20Z2 per cent in I975 to ITI per cent in i980. The scarcity value of such goods was

4 Tabatabai, 'Economic stabilization and structural adjustment in Ghana', Labour and Society 2, 3 (i986). 5 Ibid.

6 International Monetary Fund, Ghana: recent economic developments (Washington, I987). 7 Ibid.



VOTING PATTERNS IN GHANA 5I

co o tH o - c o C-4

m co ? " & . coi-.)& b C 0 CO;1~ Ci4 - 0 LO LO CO mC - o L CO O c

eqCO ~~ ~~~~~~~~0 eq

Nq LO d CO CO - C) 0b C 00O

0i eo dq 01 1

o r- co C C

t Lo Lo 1

- roe~LO~~ b ? - L ew Q0 U t LO 0 d " ?

Co:CO 0r0 01 N L eq ~o )- c1 cs

0 co 00 03 0 0 o c0 C( - O 0 L

03 0) LO LO I-1 r- ",L co c ot cn "It - Co O - co C 01

0 co cnll 0 r-"0 d nH o LO LO lt 1 00 es 00 co ?c

CO ) LO

C) co~~~Ct

* co ( 1 -0 0 1(0 0 0 "C0(0 CLOOr

Oco 0 t M- CO -.t - CO 0 r o CO- o - c (o 00 t0 - m 0 1

4-) co - m tO N 0

Ut~~~~~~~~~~L CO LO+ LO o C o 0 t- - C 4 m b r* Lo b Lo

O CON Nt- Co oo to 0o o cC) 0O .- , o co d C o tco 4 ; O - q

U- Q m cod 00 co d c C4??

*4 0

0~~~~~~~~~~~~~

eq0 to oc tO eq 0 0 s

2v LO(0 Ct O I ? cr) o: oo uC O ( O

LO.0U

o

_o * - ' O - ' ;

~~~ ~~~~~-o ~ ~ ~ h

z ,-~ Q 0 (tQ Q t0 V

o 0~o



52 MAHAMUDU BAWUMIA

high, and by and large they assumed the title of' essential commodities' within the urban Ghanaian community.

Formal sector wage and salaried employees were severely affected by the decline in national income. By i983, the real minimum wage had fallen to I I per cent of its I974 value.8 For many Ghanaians this was not a living wage 'and for many employees the real remuneration for formal sector work could not support the food requirements of one single adult, let alone those of the family, plus their clothing, education, health and housing requirements '. Putting this in context, the estimated minimum socially acceptable household budget would have been of the order of 3I,500 cedis a month in mid-i984.10 At that time the minimum wage was under i,000 cedis a month, and upper middle level civil service salaries were 2,000 cedis a month. Rent-seeking (affectionately termed kalabule by the Ghanaian public) on the part of formal-sector workers provided the 'magic' needed to overcome their precarious situation. This rent-seeking took the form of obtaining goods (or chits/licences which allowed them access to scarce goods or foreign exchange) from official channels and selling them on the parallel markets. The kalabule economy benefited mainly urban residents (especially traders and public officials) and impacted detrimentally on the welfare of the urban poor and rural dwellers. However, rural residents living in the cocoa-producing border areas also benefited from smuggling.

During the kalabule era, the rural sector was relatively worse off because access of rural residents to the 'magic' needed to survive the harsh realities of the economy was very limited. However, the distribution of rents within the urban sector was skewed towards a small proportion of residents.1" In a survey conducted in i984, Huq reported that about 53 per cent to 6o per cent of respondents from Accra did not have access to goods supplied through official channels. The comparable figure for those with jobs at the University of Cape- Coast was 33 per cent. In Assin Awarabo (a village in Central Region), between 8o per cent and I oo per cent did not have access to goods from official channels.12 One has to remember that during most of the kalabule era, Ghana was under military rule. These military rulers had more to fear from urban unrest and were not too responsive to the silent cries from the rural areas. Rural dwellers saw their incomes dwindle

8 UNICEF, Ghana: adjustment policies and programmes to protect vulnerable groups (Accra, mimeo, Nov. 1986). 9 Ibid. 10 Ibid.

" M. M. Huq, The Economy of Ghana: the first 25 years since independence (Basingstoke, i990)

p. 42- 12 Ibid., pp. 42-3-




TABLE 4

Ghana: index of real government health and education expenditure per capita, I969/70-I98i/82

Education expenditure Health expenditure per capita (0975/76 = ioo) per capita (0975/76 = ioo)

i969/70 773 716 1975/76 I00 I00

I 978/79 94*4 84'9 1979/80 55'2 47'2 i980/8i 35 7 35'8 i98i/82 28'7 22-6

Source: World Bank, Ghana: towards structural adjustment (Washington, DC, i985), vol. II, cited in UNICEF (I986), table 4-5.

because of the fall in agricultural cash crop production and the high inflation rates. Real incomes of cocoa farmers in I 983 were I 2 per cent of their value in I 974. Food producers' incomes, due to the increase in food prices, seem to have been less severely affected than most other groups.

However, the trend in the real income of food-growing farmers is less certain and more diversified, given the declining per capita production. Farmers in the Northern and Upper regions, for example, fared particularly badly with respect to prices and production. Their main cash crops, yams and groundnuts, showed the lowest real price increase, while they were worst affected by drought." Bequele argues that while the larger farmers were protected by the rise in food prices, small farmers were made worse off by that rise since many of the small farmers were net purchasers of food.15 Also, part of the decline in cash crop and food production was more apparent than real.

Between I972 and I 982, the government budget fell from I 8-3 per cent to I o I per cent of GNP.16 This led to a fall in real expenditure (in absolute and per capita terms) in the education and health sectors (Table 4). As in many other sectors, equipment in health institutions fell into disrepair due to lack of spare parts. Basic drugs such as nivaquine and aspirin, and consumables such as bandages, needles and syringes, were in desperately short supply, and were often unavailable

13 Tabatabai, 'Economic stabilization'. 14 UNICEF, Ghana. 15 A. Bequele, 'Stagnation and inequality in Ghana', in D. Ghai and S.v Radwan (eds.),

Agrarian Policies and Rural Poverty in Africa (Geneva, I 983). 16 UNICEF, Ghana.



54 MAHAMUDU BAWUMIA

in rural clinics. The poor road system and shortage of vehicles in good running condition (again due to shortages of spare parts) made the distribution of the few supplies extremely difficult in rural areas. The country lost more than 50 per cent of its doctors between i98i and I983, and about 8-5 per cent of nurses in I982 alone,'7 and probably more between I974 and i98i.

Of the health personnel remaining in the country, the vast majority were located in the urban areas and largely concentrated in the Greater Accra, Eastern and Ashanti regions."8 Ubugo and Umo note that although the nine administrative capitals in Ghana accounted for only I5 per cent of the population, they controlled 55-4 per cent of the hospitals and 5I -3 per cent of the hospital beds. Hospital records show that annual attendance dropped considerably.'9 For example, in Korle Bu Hospital (the nation's foremost), the outpatient attendance in I983

was II7,000 compared with I98,000 in I 979. This trend also reflects the increasing importance of private medical services during this period.

Nutritional studies undertaken in i982 and i983 'indicate that the caloric and protein intake of children was below 69 per cent and 87 per cent of requirements respectively... and 5 per cent of children under 5 years were severely malnourished, and 50 per cent were under go per cent of the standard weight for age'. 20 There was a decrease in calorie availability. Average calorie availability as a percentage of requirements declined from 92 per cent in i960 to 68 per cent by i980 with an attendant increase in child malnutrition. This was the result of the decreased food production, the increase in food prices and the declining purchasing power of Ghanaians. As households became poorer, child labour became more crucial to making ends meet. This contributed to high-school drop out rates, especially among girls, as well as irregular attendance. Child labour in the informal urban sector was most apparent in hawking, and a modern form of 'child pawning' into domestic service to reduce the domestic costs of poor households.2' During the period economic decline reached a level much higher than in the previous twenty years.

After an initial coup dietat in I979, Flt. Lt. Jerry John Rawlings handed over power to the civilian government of the Peoples National Party (PNP) in the same year. Apparently unhappy with the economic

17 Ibid. 18 Ubugo and Umo, 'Impacts of the external sector on employment equity', World Employment

Research Papers (Geneva, i986). '" UNICEF, Ghana. 20 Ibid., p. I5. 21 Ibid.




performance of the PNP, he staged another coup detat on 3 I December I 98 I, and established the Provisional National Defence Council (PNDC). Between January i982 and November i983 the PNDC was characterised by socialist revolutionary policies and measures, tinged with populism. The government in this era was composed of avowed socialist intellectuals, junior military officers, a Catholic priest, a student activist and a radical trade union leader.22 The government drew its support from the discontented urban masses, disaffected ranks in the military, students, the urban unemployed, lower-level trade unionists, a group of radical left-wing intellectuals and some politicians. Frimpong Ansah notes that 'all these were hard-core constituents of the urbanized vampire state '. The business community, large-scale farmers and professionals were the regime's declared enemies.

Economic policy was interventionist. The PNDC sought to reduce 'the stranglehold of privatization' on the economy, and to increase state control of essential services as a means of protecting people from unscrupulous local and foreign capitalists.24 Price controls, import duties and tariffs were imposed on certain commodities produced in or imported into the country. The PNDC was hostile towards the prescriptions of Western financial institutions such as the International Monetary Fund (IMF) and the World Bank. Indeed, the PNDC initially blamed Ghana's economic woes on its 'neo-colonialist' structure, and an IMF/World Bank stabilization and Structural Adjustment Programme was seen as inimical to the interests of the country.

The government sought the support of workers, students and the rest of the urban population to bring about a radical change in the economy. In an attempt to rally working-class support behind the regime, one of the first enactments of the regime was the Rent Control Law i982 (PNDC 5). This reduced rents on residential properties by 50 per cent and set a C5o.oo limit for rents on single rooms. PNDC Law 7 also stipulated that if any unoccupied room/house was identified in rental premises, and the owner refused to rent or could not be traced, the premises could be let out lawfully by the local community PCD and local rental control unit to persons who had no dwelling places. These

22 Donald I. Ray, Ghana: politics, economics and society (London, i986), p. 6i. 23 J. H. Frimpong-Ansah, The Vampire State in Africa: the political economy of decline in Ghana

(London, i99i), p. I I 2.

24 Kenneth A. Atafuah, Criminal Justice Policy, Public Tribunals, and the Administration of Justice in Rawlings' Ghana (i98s-1992): a study in the political economy of revolutionary social change and criminal law reform. Ph.D. dissertation, Simon Fraser University, I993, p. I57.



56 MAHAMUDU BAWUMIA

laws were music to the ears of many urban workers who generally faced high rents because of housing shortages. Workers' Defence Committees (WDCs) and Peoples' Defence Committees (PDCs) were established to mobilise the population. Herbst notes that while helping the urban workers, the PNDC seemed to accentuate the urban bias of previous regimes by imposing controls on food, the major source of income for the 70 per cent of the population that live in the rural areas.25

By late i983, a combination of frequent coup attempts against the regime, a severe Sahelian drought, sporadic bush fires, the flight of capital from the country and the continuing miserable performance of the economy threatened the very existence of the regime. This precarious situation was compounded by the mass expulsion of over I

million Ghanaians from Nigeria in i983. For the first time in Ghana's history, the threat of hunger began to loom large for a significant proportion of the population. In the absence of a realistic alternative from the intellectual left (and the Eastern Bloc countries), the Rawlings regime was convinced by August I 983 that the best possible solution to the challenge posed to its survival by subversive political activities and the desperate economic situation was to seek help from the Bretton Wood Sisters.26

On 28 August i983, Chairman Rawlings delivered an address on national radio and television to the people of Ghana. The speech was essentially a close-up analysis and critique of the populist and economically unproductive elements in the ' Ist December Rev- olution' as it had evolved to date:

We can no longer postpone the time for halting the populist nonsense and for consolidating the gains of the past 20 months and making a noticeable leap forward ... Production and efficiency must be our watchwords. Populist nonsense must give way to popular or unpopular sense... to scientific sense, whether it is popular or not. Many of us have spent too much time worrying about who owns what. But there can be no ownership without production first. The only resources which do not have to be produced are those given to us by nature, and these must be used for the benefit of all the people of today and tomorrow. Everything else has to be produced, and until we all fully recognise and act upon this fact, we shall be deceiving ourselves with empty theories.27 (emphasis added)

Rawlings's rejection of' populist nonsense' reflects a timely recognition of the simple truth, that revolutionary rhetoric and 'mobilisation' are by themselves inherently unproductive, and are inadequate vehicles for sustaining a revolution. There was an urgent need to achieve some

25 Jeffrey Herbst, The Politics of Reform in Ghana, I982-i99i (Berkeley, I993), p. 28. 26 Ray, Ghana, p. io8. 27 West Africa, I 2 Sept. i983.




economic victories in the short term in order to contain the people's mounting disillusionment with the revolution. Economic productivity was subsequently moved to the top of the PNDC's revolutionary agenda.

The i983 Budget, announced by Dr Kwesi Botchwey, the finance secretary, signalled the government's change of course. This Budget contained a significant devaluation of the cedi (an act which had been anathema to Rawlings in I979), and an increase in the prices of basic foodstuffs. This marked the beginning of Ghana's Structural Ad- justment Programme.

STABILISATION AND STRUCTURAL ADJUSTMENT IN GHANA

(I983-I992)

The Structural Adjustment Programme in Ghana was implemented in the context of the government's Economic Recovery Programme (ERP). Between i984 and I992, we can identify two phases of the implementation of the ERP. The first phase (ERP I, i984-6) was designed to control inflation, restore overseas confidence, arrest and reverse the decline in production, particularly in agriculture, rehabilitate the decayed productive and social infrastructure, stimulate exports, curb the consumption of luxury imports, and mobilise domestic and external resources to restore living standards. The goals for ERP II (I987-92) were to ensure economic growth at around 5 per cent per annum in real terms, stimulate significant increases in savings and investment, improve public-sector management, and place the external sector on a sound footing through trade and exchange rate liberalisation, public-sector reform and public investment.28 Ghana's adjustment efforts have been supported by successive arrangements with the International Monetary Fund (IMF) and the World Bank. Overall, the total amount of IMF financial resources committed to Ghana during i983-9i amounted to SDR I,208 million.29

The major themes of policy conditionality in Ghana between i983 and i99i were the following :30 I, increasing the producer price of cocoa; 2, reducing the cocoa marketing costs of the Ghana Cocoa

28 Government of Ghana, Economic Recovery Programme (Report prepared by the government of Ghana for the meeting of the Consultative Group for Ghana, Paris, i987), vols. I and II.

29 International Monetary Fund, Ghana: adjustment and growth, i983 -i, Occasional Paper No. 86 (Washington, I 99 I).

3 J. F. Toye, P. Mosley and Jane Harrigan, Aid and Power: the World Bank and policy-based lending (London, I99I), p. I7.



58 MAHAMUDU BAWUMIA

cm O NO O LO I-tI- r- -

O I-o

NN co w N

0 I- t LO

LO ' o 00 ~.o -. .0 b-) c 0 0 m M OG

t- -Lo ' IO 00 00 t -

a) dzt cl r-N)

C ? - n 00 d 1tcm co r - 00 O z , 04 00 # N M - - C?- t

co 0 N N N Chco

0 t-, N r'- o Lo O 0 t o ? -0 cm o 0 && Ca) L o t-

-, c0 oo - - C o n 00 N c0

(0 <0 (0

?? O

n < 0

C co ( o b. (0( & ~- 4 0) - . 0

O0 i4 Lo (Jo CO N 00 C s O Q 0 GO 04 -4 -n

- o o G i O N H i c0c

00

C.) C~~u t30 N ~~N O co 0 LO?

o1~~~~~~ ~ ~~ cb bCh r- CT ) ~40 of co Lo x) O LO a) CM - N "t . a N N C.) C GO O O N I0_OCOL _

Co Q Y QOC ) C' N L t- C.) U:0 t t- N C o ;I) ) c C 7) r- m

GO 'IC O N 4 00 N 0 N 0 O

Ln 4--) C~ 0 CO4N CGO-

(0LO 041C Q * _{;~~~~~~~~~~~~>

._4~~~~~ ~~~~ L

oc soi s o .r.: ?. 00 t-? ~C ? CONL t - O O H.E~~~~~~~ co C- C CO 0)-

LO coOn

o ~ LO o b OO C O (0dO ( 0 C d L mo tb b do O N 0 b .

- 0GO t- N t- 0 - ' ) LO c

co 04VI

o~~~~~~~~~~~o Q >4

fA G00G(0CO ,O4CO CO0N O )d'

C)I X - 0 H s?t

.0

0 Ln~~~~~

04

0 o o 0

0 O~~0

0

0 n

_ b U P U i m S C) Q 0




TABLE 6

Agricultural GDP annual growth rates, I985-1990 (Mo)

Sub-sector I 985 I 986 I987 I 988 I 989 I9go I983-90

Agriculture and livestock - I-9 0-2 -0-3 6-o 51I -4.3 2-8

Cocoa I3-2 i8-2 3-3 6-3 3 I 3-0 6.7

Forestry and logging o I I-2 I15 3 4 12 40 I -8

Fishing I I9 I4-0 -IOI 2-3 0-9 2'7 41I

Total agriculture o'6 33 2-0 3-6 4'2 - 20 2.7

Source: Ghana Statistical Service, 1992.

Marketing Board; 3, removal of subsidies and price controls; 4, liberalisation of the trade and foreign exchange regime; 5, cost recovery and removal of subsidies in health and education; 6, public expenditure programming; 7, state enterprise divestiture; 8, public sector management; 9, banking reform.

The macroeconomic results of the SAP (I983-I992)

The data in Tables 3 and 5 show a marked contrast in the performance of the Ghanaian economy between the I974-83 and i983-92 periods. The Ghanaian economy recorded a remarkable recovery between I 983 and I 992. Over this period, the GDP growth rate averaged 5 per cent per annum, with the output of cocoa, minerals and timber recording significant increases. Inflation fell from the very high pre-reform levels (I22 per cent in I983) to Io per cent in 1992. However, over the I983-92 period, urban prices increased faster than rural prices, with the ratio of urban to rural consumer prices increasing from o-83 in I 983 to IP02 by i99i (Table 5). Also, real interest rates (i.e. the differences between the nominal interest rate paid on savings, and the rate of inflation) turned positive by 99 I (Table 5) and the government budget recorded surpluses from i986 to i99i.

Agricultural prices and output

The SAP was aimed largely at improving incentives for production (especially of exportables). As a consequence, the food sub-sector benefited only marginally from the programme. Table 6 shows the growth rates of the sub-sectors within agriculture between I985 and I988. The cocoa sub-sector expanded at an average annual rate of 6 7 per cent, which was higher than the rate for real GDP for the same



6o MAHAMUDU BAWUMIA

TABLE 7 Index of real wholesale prices of major food crops and cocoa: I970-86

Maize Rice Corn Cassava Yam Cocoyam Plantain Cocoa

I970 I 00-00 I 00-00 100'00 I 00-00 I 00'00 I 00-00 100'00 I 00'00

i980 I09-49 74-18 122-55 I I0-46 78 55 93I 3 I i8-88 4143

i983 365-96 274-02 401'75 42486 22558 484-o8 3896i 74VI8 i986 8I'02 77o09 69-75 10447 75U 3 9393 I22-50 81a79

Source: IFAD, Report of the Special Programming Mission to Ghana (July i988), table 6, p. i62.

period. This performance contrasts sharply with an average annual growth rate of about 2-7 per cent recorded for the agricultural sector as a whole. We see from Table 6 that the performance of the agricultural and livestock sub-sector (of which food production is the main sub- sector) has been modest, growing at an average rate of 2 8 per cent. Comparing the data in Tables 3 and 6, this growth performance contrasts favourably with the I970-83 period when agriculture declined at a rate of 02 Xper cent.

Agricultural pricing policies, together with good rainfall, resulted in increases in the output of cocoa as well as cereals and starchy staples (Table 5) from the pre-i983 levels. The real producer price of cocoa increased four-fold between i982/83 and i987/88.31 Since cocoa is mainly produced in the southern rural areas of Ghana, the real price increase is likely to have increased the incomes of these rural households, both rich and poor.

The nominal prices of all major food crops rose sharply between I 980 and i986. Compared to i980, prices in i986 were six times higher for guinea corn, eight times higher for maize and millet and ten times higher for rice, cassava and plantain. Over the same period, the producer price of cocoa increased more than twenty times.32 On the inputs side, fertiliser supply improved after the launching of the ERP in i983. The subsidy on fertiliser was phased out, and its distribution privatised by i990. The improvement in fertiliser supply and increase in producer prices was likely to have positively affected food producers, but the phasing out of fertiliser subsidies was likely to have affected them adversely.

Table 7 shows that there was a massive increase in real food prices between I970 and i983. This was due partly to the liberalisation of

3' (OECD, o992). 32 Ministry of Agriculture, Agricultural Land Use of Ghana (Accra, 1 99!i).



VOTING PATTERNS IN GHANA 6i

TABLE 8

Central government social welfare expenditure, I975-I989

I975 1980 I983 I984 I985 I986 I987 I988 I989

(in % of GDP) Narrow measure (I) 8-7 59 2-3 33 44 5-2 5 3 5 7 5.8 Broad measure (2) 95 6.5 26 3 7 48 5-7 62 6.7 7I

(in % of central government expenditure) Narrow measure (I) 399 306 29 I 33-0 32-8 37 6 38 6 415 4I-7 Broad measure (2) 43 8 33 7 32-5 37 4 36 3 412 45-2 48 6 512

Source: Kapur et al. in IMF, Ghana: adjustment and growth, I9839i. Occasional Paper No. 86. Washington D.C., i99i, table II, P. 40); (I) Includes expenditure on education, health, and social security and welfare. (2) Includes the narrow measure plus spending on housing and community amenities, other community and social services, and special efficiency.

previously controlled prices, and also to the drought of i983, which caused a food shortage. However, between I983 and I986, there was a dramatic decline in the real prices of food crops, due to the increase in food supply resulting from higher rainfall and a greater supply of agricultural inputs.33 The implication of these price changes for the welfare of rural and urban residents is unclear. The decline in food prices benefited urban food consumers, while at the same time an increase in cocoa prices benefited the rural sector.

Fiscal policy and social welfare

An important aspect of fiscal policy is the extent to which it aids the poorest members of society. Poverty in Ghana is largely a rural phenomenon,34 and public expenditure policies can improve the well- being of the poor either by providing services consumed in large part by the poor, or by channelling goods and services to the rural areas.

The extent to which government expenditure policy has addressed the social needs of the society can be gauged by examining the level of resources directed towards education, health, and social security and welfare. Provision of such services (if accessible) represents a means for many households to escape poverty. Government spending on these three categories increased from 2-3 per cent of GDP in I983 to 5-8 per cent of GDP in I989, and from 29 I per cent of total government expenditure in i983 to 4V7 per cent in i989 (Table 8).

3 Tabatabai, 'Economic stabilization'. 3 E. Oti Boateng, Kodwo Ewusi, Ravi Kanbur and A. Mckay, 'A poverty profile for

Ghana', SDA Working Paper No. 5 (Washington, i987-8).



62 MAHAMUDU BAWUMIA

Measuring the social welfare element of fiscal policy in this way is, however, problematic. First, much spending on education, health, and social security and welfare does not benefit the needy. It would be preferable to consider only the outlays for primary education and basic health care, both in aggregate and by geographical region, but such data are not available.

Despite the note of caution cited above, the figures cited (Table 8) provide evidence of an increase in poverty-reducing government expenditure, assuming that the share of such spending benefiting the poor did not decline significantly during the period. In fact the opposite appears to be the case. Between i987 and i99i, the share of primary education in the recurrent budget of the Ministry of Education increased from 40 per cent to 43 per cent, and primary school enrolment is estimated to have risen from 65 per cent to 72 per cent.35

There is an additional element involved in estimating the social impact of government spending from the data presented above, namely that some spending that benefited the poor or those adversely affected by adjustment is excluded from the categories of health, education, and social security and welfare. For instance, the special efficiency budget for retraining and supporting redeployed public-sector employees can be seen as part of a general effort to provide a safety net.

In addition, expenditure on housing and community amenities, as well as other community and social services, provided substantial benefits to the poor and disadvantaged. Inclusion of these categories provides a broader measure of social welfare spending: such spending was raised from 2-6 per cent of GDP in i983 to 7'I per cent in i989;

as a percentage of total government expenditure, the increase was equally dramatic, from 32'5 per cent to 5I-2 per cent (Table 8).

Rehabilitation and provision of physical infrastructure

A major plank of the SAP was the rehabilitation and provision of physical infrastructure to help improve productivity. This programme was guided through the Public Investment Programme launched in I984 with a focus on roads, highways, water, sanitation and electrification projects.

35 International Monetary Fund, Ghana.




(i) Transport sector: roads, highways and railways

At the inception of the SAP, most of Ghana's roads and highways were in a state of disrepair.36 More than half the serviceable vehicles were off the road for lack of tyres, batteries or spare parts. The railways were in their worst condition since the system was established. The ports, Black Star Line (the nation's shipping line) and Ghana Airways were all saddled with dismal finances. Cocoa and other export products produced in the hinterland could not find their way to the ports for lack of adequate transport facilities.37 It was therefore a priority of the SAP to rehabilitate the nation's infrastructure. After all, what good is a liberalised market if the goods cannot reach their destination?

To improve the accessibility of rural areas, particularly cocoa, timber and food-producing areas, a programme of construction of bailey and steel assembly bridges throughout the country was instituted. The roads and highway rehabilitation programme was successful in opening up many rural areas. However, priority was given to roads linking export-producing regions, and to this extent the rural towns and villages of the Northern, Upper East and Upper West Regions were not major beneficiaries of road rehabilitation projects under the SAP.

(ii) Electricity

Out of the iI o districts established in I 988, 64 were without electricity from the national grid. A major programme of supplying electricity to the Northern, Upper East, Upper West and Brong-Ahafo regions from the National Transmission Grid began in i987. This was completed in I 99 I and established the basic infrastructure for the supply of electricity from the national power system for virtually the whole of Ghana. Under this programme, many rural areas, for the first time since Ghana's independence, had access to electricity. Under the Volta Region Rural Electrification Project, for example, coastal towns and villages from Denu to Keta were electrified. Other areas include Akatsi, Sogakofe, Jasikan, Kadjebi-Likpe, New Abirem and Don- korkrom. The government also instituted a Self-Help Electrification Scheme to assist rural communities which had embarked on self-help electrification and provided labour and funds to purchase local materials, including poles, to complete their electrification projects.

" Huq, Economy of Ghana, p. 48. 3 Government of Ghana, The Economy and People of Ghana (Accra, i99i).



64 MAHAMUDU BAWUMIA

TABLE 9

Population served with water supply facilities, i983

Rural Urban Total

Total population ('ooo) 7,858 4,863 I2,72I

Percentage served with water 43 7 8o0o 57-6 No. of localities served 3,529 I32 3,66i

Source: K. Ewusi, The Dimensions and Characteristics of Rural Poverty in Ghana (Legon: ISSER Technical Publications, series no. 43. University of Ghana, i984), table 8.7, p. 68.

TABLE 10

Percentage of population served with water supply facilities, i989

Regional Rural Urban total %

Ashanti 59 88 68 B-Ahafo 42 96 56 Central 67 95 75 Eastern 47 98 6I G-Accra 45 99 9I Northern 7 88 27

Upper Regions 33 95 39 Volta 29 69 45 Western 59 77 63 Ghana 58 86 65

Source: Ghana Water and Sewage Corporation, 5-Year Rehabilitation and Development Plan (Accra, I990), p. Fi.

Under this scheme, over ioo towns and villages were connected to the national grid by the end of I992.

(iii) Water

At the beginning of the SAP in I 983, I I 7 towns in Ghana were supplied with pipe-borne water (57-6 per cent of the population). There was a clear urban bias in access to water facilities (Table 9). Table 9 shows that the discrepancy between the rural and urban populations was quite significant, with only 43 7 per cent of the rural population having access to water supply facilities, as compared with 8o per cent of the urban population.

About 86 per cent of urban residents across Ghana had access to water facilities in i989, compared with 58 per cent of rural residents.




Greater Accra, Western, Central and Eastern Regions were the best served areas (Table Io). The worst regions included Northern, Upper and Volta regions. The electricity, road networks and water sectors were rehabilitated and the electricity power grid was extended to northern Ghana, including its regional capital, Tamale, and major districts. Again, the extension of electricity and water services and rehabilitation of roads mainly benefited rural areas which had been ignored by previous governments.

Wages and salaries

As argued earlier, the real wages of workers suffered a decline during the pre-adjustment period, with a resultant loss of morale and an increasing propensity to engage in rent-seeking activity. Under the SAP, the PNDC government pursued the objective of increasing the real wages of public-sector employees. This is yet another sense in which the SAP in Ghana was unorthodox. In the civil service, the salaries of senior management were increased more than three-fold between I 983 and I 988, while those for unskilled labour increased only by 50 per cent, increasing the ratio of the highest to the lowest salary in the civil service from 2 5: I in I 984 to 9 5: I by I 900.38 Table 5 shows that real wages and salaries increased between I 983 and I 992. The real minimum wage increased from an index of 54 in I 983 to I I 7 by I 992.

Real public-sector earnings, on the other hand, increased from an index of 48 in I983 to 270 by I992. This increase in real earnings since i983 was also evident in the agricultural, manufacturing and trade sectors (Table II).

With both salaries and agricultural incomes increasing in real terms after I983, one can expect that the well-being of households improved generally at all levels. Nevertheless, the increase in wages and salaries was made possible, in part, from the retrenchments which took place in the public sector. By August I989, the PNDC government had laid off 29,052 public sector employees, to whom full benefits had been paid. Another 34,500 were dismissed from the Cocoa Board. Again, these retrenchments impacted more negatively on urban households.

38 OECD, I992.



66 MAHAMUDU BAWUMIA

TABLE II

Indices of real earnings (I 970 = I oo)

Agriculture Manuf Trade

I970 100-0 I 000 100I0

I971 IIO07 I43'3 I00-2

I972 IOI4 I43'3 100'9

1973 944 99 3 IOIlI 1974 I09-9 I2I19 I08-I

I975 96-6 I00*7 I09-7

I976 65-5 70?7 IOII

I977 65-2 59 7 69-I

I978 4I14 47.8 88-8 I979 28-I 37.2 79.6

i980 30'9 40-7 io6-6

I98I 22.5 231 I30-6

I982 20-5 2I16 I434

I983 I4.I I67 I44-I

I984 26.7 25.5 I48-8 I985 29-9 421 I69-4

Source: IFAD, I988, p. I58.

Cost recovery measures

In the areas of health, education and water, the SAP called for the introduction of user fees. This was to militate against the situation in which the economy found itself pre-reform, with an inability to provide continuing subsidies to these sectors as a result of depleted public finances. Thus, while, in principle, many services provided by these sectors were 'free', in practice, they were unavailable. Thus in I985, the PNDC government introduced fees for hospitals and clinic consultations and laboratory tests. At the same time, government health expenditure as a share of GDP/Budget which had fallen between I 980 and I 983, increased after I983. Ephson reports that between June and August I 985, there was an immediate 25 per cent drop in the visits to the Korle-Bu hospital (Ghana's pre-eminent hospital located in Accra), and a 50 per cent decline in the much more heavily used Korle-Bu polyclinic.39 To the extent that rural residents relied more on traditional sources of medicine, one can argue that the increase in hospital user fees is more likely to have negatively impacted on urban rather than rural households. Kraus reports that 'after an

" Ben Ephson, 'The price of health', West Africa, I4 Oct. I985, p. 2I56.




initial decline in the use of health facilities following the start of the higher fees, there has apparently been an increase again in use '.40 New and increased book user fees were also introduced in the elementary schools and fees for housing and feeding costs were introduced in the universities. Water rates were also raised by hundreds of per cent in i986 and later years, and electricity rates by lesser amounts.41

In a comprehensive examination of the welfare impact of adjustment on equity in Ghana, the OECD concludes that:

it is not an unreasonable view that equity overall may have been somewhat increased by adjustment. This is the case both for the rural/urban balance and for the general comparison between 'rich' and 'poor' since the richer urban populations have certainly been exposed to more negative effects.42 (emphasis added)

The evidence discussed so far suggests that rural households are likely to have benefited most from the SAP during the I983-92 period, given the higher prices they received for their products as well as the increased access to physical infrastructure. The cost of adjustment, however, was disproportionately borne by urban households. The voting behaviour exhibited by rural and urban voters, with rural voters voting for the incumbent and urban voters doing the opposite, corresponds to the 'rational' allocation of preferences found in more developed Western democracies.43

This article certainly does not seek to suggest that economic considerations were the sole determinants of electoral behaviour in the 1992 Ghanaian presidential election. Ethnic, political, ideological, class, kinship, economic and other factors were all probably influential in accounting for voters' choices. Table I, for example, indicates that ethnicity might have been an important influence on voting behaviour, since it shows that the Ashanti and Volta regions do not appear to have followed the national pattern. The Ashanti region voted overwhelmingly against the incumbent, while the Volta region voted overwhelmingly for him. This doubtless reflects the fact that the PNDC was (and is) largely perceived as being Ewe-based. Herbst quotes a senior community official as saying:

40 Jon Kraus, 'The political economy of stabilization and structural adjustment in Ghana', School of Advanced International Studies conference on the political economy of Ghana. Johns Hopkins University (Baltimore, I99I), p. I42. 41 Ibid.

42 OECD, Adjustment and Equity in Ghana (Paris, i9992). 43 Anthony Downs, An Economic Theory of Democracy (New York, I957), pp. I 5-33.



68 MAHAMUDU BAWUMIA

The region [Ashanti] feels discriminated against... People feel the projects which have been established in Ashanti have been done with wealth from the region - especially the cocoa which pays the East German firms for the roads - but that other regions get aid. People in Ashanti region do not feel they are getting their fair share."

Thus, even though there was an increase in economic activity in the Ashanti region, with a rehabilitation of infrastructure and the cocoa and mining sectors, 'many in Kumasi (the Ashanti regional capital) attribute most of the development to the resources the region itself generates. In some cases, the overall animosity towards the perceived Ewe-based regime causes some citizens to ignore the welfare gains they may have experienced'." The vote in the Volta region, conversely, may be attributed to the fact that Rawlings hails from this region, and thus the voters were making sure that the interests of the region were being protected by returning their 'son' to power. By the same token, Rawlings obtained only I 4-43 per cent of the votes in Sissala constituency (Upper West region), and this can be attributed to the fact that former president Hilla Limann, who obtained 78 per cent of the votes, hails from this constituency.46

The results from the Ashanti and Volta regions suggest that ethnic perceptions sometimes affect considerations of economic improvement. Also, voters might not vote for a candidate if that candidate was perceived to have treated them unfairly, notwithstanding the fact that their welfare may have improved in reality. Thus, it may be more a question of perceived fair treatment rather than economic improvement as such. Table i, which aggregates the share of votes on a regional basis, likewise obscures a considerable level of variance between individual rural constituencies. For example, Rawlings obtained 49 per cent of the vote in Gukpeju-Sabongida constituency (Northern Region), as against 62-5 per cent in the region as a whole; 46 per cent in Akropong (Eastern Region), as against 56-7 per cent; and 47 per cent in Wenchi East (Brong-Ahafo), as against 6i 5 per cent.47 Such differences were doubtless due to factors such as historic party attachments, local factionalism or class composition.

More fundamentally, some doubt has been cast on the overall fairness of the elections 48 and it could be argued that tampering with the election returns on behalf of the incumbent might be easiest in rural

4 Herbst, Politics of Reform, p. 87. 45 Ibid., p. 88. 46 Ghana Electoral Commission, The Presidential Election, I993. Regional and constituency results

(Accra, I993). 47 Ibid. 48 Mike Oquaye, 'The Ghanaian elections of 1992: a dissenting view', African Affairs 94, 375

(1995)-




areas, where the electorate was less sophisticated, and the electoral process was subject to less external supervision than in the major towns. While the author of this article is not in a position to pass judgement on particular allegations of malpractice, it is none the less noteworthy, both that there was a considerable level of variance both between and within regions, and that the differences between rural and urban voting figures can be readily associated with the differential impact of the incumbent government's policies on voters. This article therefore concludes that, other factors notwithstanding, the SAP was probably the decisive factor in Rawlings's victory. The SAP was implemented in i983 after a decade of persistent economic decline. This decline resulted in increased rent-seeking by all households, but the rural households generally fared most poorly.

Mikell has argued that the PNDC's adoption of a Structural Adjustment Programme in i983 represented a U-turn in economic policy.49 This U-turn alienated the urban working class and students, on whom Rawlings initially relied for support. The PNDC thus turned to the peasants for support and sought to create a rural constituency for its reforms. The PNDC aimed to break the cycle of rural disengagement by creating a constituency which was integral to Ghana's livelihood, and, hence, acted aggressively on behalf of rural areas by creating militant mobi-squads and self-help labour groups (or nnoboa) in the cocoa-growing areas in i984. These activities were successful, and the rural producers became consistent supporters of PNDC. Mikell writes that 'certainly the farmers I met in Sunyani in I986 were more enthusiastic about Rawlings and the PNDC than many of the bureaucrats and teachers in Accra and the Inland towns ... they were

responding positively to the restoration of their agrarian economy...'50 On I July i987, the PNDC initiated political reform at the local level with proposed elections for District Assemblies in i988. The District Assemblies were presented as the 'bodies exercising state power as the people's local government' and formed the basis for a grassroots democracy. As Chazan notes, the electoral patterns in the district assembly ballot revealed key features of state-society relations by the end of the decade.51 The very high participation rates (59-I per cent) compared with the low turnouts of 32-25 and i8-4 per cent in the

49 Gwendolyn Mikell, 'Peasant politicization and economic recuperation in Ghana: local and national dilemmas', Journal of Modern African Studies 27, 3 (1989), 455-78.

50 Ibid., p. 456. 5 Naomi Chazan, 'The political transformation of Ghana under the P.N.D.C.', in Donald

Rothchild (ed.), Ghana. The political economy of recovery (Boulder, i99i).



70 MAHAMUDU BAWUMIA

parliamentary elections of I979 and the district council elections of I978 respectively, indicate that:

For the first time in post-colonial Ghanaian history, a regime derived its support primarily from rural constituencies.... Indeed voter participation in the large cities was much below national and regional averages ... This distribution was indicative of the relative strength of the PNDC in the countryside and its waning credibility in the cities..."

The balance of power under PNDC had thus shifted in favour of the rural sector.

The SAP policies implemented during the period under con- sideration (i983-92) largely benefited the producers of cash crops like cocoa. Cocoa producers, largely in the rural south, also benefited from increases in producer prices. The terms of trade moved against food producers and in favour of cocoa producers. Producers of cash crops were largely located in the western, eastern, Brong-Ahafo, Ashanti and greater Accra regions.

The cost recovery measures implemented in health care, education, sanitation and other areas, it was argued, were more likely to impact on urban households. For rural households, the existence of free health care in the pre-reform period was not borne out in reality since many rural households did not have access to these services.

The election results indicated that Rawlings derived a greater share of his support from the rural areas. Many rural voters who had been the major beneficiaries of the SAP overwhelmingly voted with their pocketbooks and returned the incumbent to power. Many urban voters on the other hand, who, as a group, bore most of the cost of adjustment, similarly voted with their pocketbooks by voting against the incumbent. A major lesson to be drawn from the Ghanaian experience is that good policies would be rewarded and bad policies will be punished by the electorate. In this regard, the factors influencing voting behaviour in Africa are not fundamentally different from those in Western democracies.

52 Ibid., pp. 36-7.



understanding the rural--urban voting patterns in the 1992 ghanaian presidential election. a closer...

Documents