Navigation – Plan du site

“They Have Left Us in a Hole”

Democratisation and Political Power in a West African Village
Lars Rudebeck
p. 65-104


Une distinction conceptuelle est introduite entre démocratie minimale et démocratie substantielle. Le dernier concept intègre une démocratie minimaliste constitutionnellement définie, et une notion d’égalité politique dans la pratique. Selon cette perspective, la ‘democratisation substantielle’ et le ‘développement’ sont liés. L’argument est illustré avec une étude de cas diachronique d’un village au nord de la Guinée-Bissau. Pour le village, la démocratisation minimale manifestée à travers des élections multi-partistes depuis le début des années 90 n’a abouti ni à la démocratie substantielle, ni au développement. Il y a néanmoins des limites à la patience des villageois, car en 2006 l’école du village, fermée depuis 1989, a été rouverte par les parents d’élèves. Cette initiative pourrait, en effet, être vue comme une tentative de prise de pouvoir (empowerment) par une communauté locale en quête de développement. Durable ou non, cette initiative concrétise la question-clé de l’organisation horizontale par le bas.

Haut de page

Texte intégral

Decentralisation and democratisation

  • 1  This is a new version bringing in additional material and observations, as well as being marked by (...)

1The conceptual focus of this article is on democratisation. More concretely, its thematical focus is on power, livelihoods and cultural meaning, which were the interdisciplinary catchwords of the conference on decentralisation where an early version was presented in May 2004 (Rudebeck 2004).1 The confluence of concepts and themes becomes logical and possible, as the overlap is considerable between the concepts of democratisation and decentralisation. But, the two are far from identical. All decentralisation is certainly not democratic, and some democratisation may well involve centralisation of democratic power in order to overcome local non-democratic power. Still, as said, the two are close.

  • 2  For the conceptual framework employed, see my previous work on democracy, democratisation and deve (...)

2We are concerned primarily with ‘substantial democracy’ and its links to ‘development’.2 The latter is conceptualised as a process of structural change whereby the gap is bridged between existing possibilities and general needs recognised as legitimate in a given society. In the analytical language employed here, ‘substantial democracy’ is a two-dimensional concept aiming to integrate constitutional minimalistically defined democracy with a notion of political equality in actual practice. ‘Substantial democratisation’, thus, is about political power becoming more equally/less unequally distributed between citizens in a process where constitutional rules and actual practice support each other. This is to say that more people than before begin to take charge as citizens, sharing in decision-making on issues of common concern at various levels of society, including the local level. The more of this, the more ‘substantial democracy’.

  • 3  The argument is in line for instance with Törnquist’s theoretical analysis of post-tsunami develop (...)
  • 4  The writing of this article had largely been concluded by early November 2008, at the very moment (...)

3In this perspective, substantial democratisation and development are linked to each other, at all levels, including the local level.3The following case study of a village in Northern Guinea-Bissau provides a mainly negative illustration of the linkage, in the sense that non-development stands out as linked to non-substantial democratisation. At the end, however, the possibility of a positive linkage is also suggested. First, a brief background is given on overall democratisation and development in the specific country where the case is situated.4

Brief background on democratisation and development in Guinea-Bissau

4Guinea-Bissau is a West-African country, located between Senegal in the north and Guinea-Conakry in the south and east, living largely off agriculture and international assistance. Its surface covers 36,125 square kilometres and the population is a bit over 1.5 million. It is divided into nine administrative regions.

5Like many other countries in Africa and the world, Guinea-Bissau has been going through a process of democratisation since around 1990. This process is shaky and vulnerable for being limited to the constitutional aspects of democracy at the expense of decentralised citizen sovereignty or autonomy in a substantial sense. Constitutionally and election-wise, this democratisation is nevertheless real.

6Guinea-Bissau became a juridically sovereign state in 1974, after eleven years of decolonisation war against Portugal. The autocratic single-party system was legally abolished in 1991, and democratic parliamentary and presidential elections were held in 1994. The ruling party since independence and former liberation movement, PAIGC (Partido Africano da Independência da Guiné e Cabo Verde), managed to win both elections and remained in power. Still, a political multi-party system began to function, and some economic and civic progress began to take place.

7The last few years of the 1990s turned violent, however. Although the regime had been democratically legal since 1994, it was hardly legitimate within its own society. A destructive civil and regional war erupted in June 1998, breaking the peace that had lasted since 1974 (Rudebeck 2001). In spite of being supported by several thousand neighbouring country soldiers sent in mainly by Senegal but also by Guinea-Conakry, the president João Bernardo ‘Nino’ Vieira, legendary fighter in the independence war turned autocratic ruler, was forced to abdicate in May 1999. He had by then been president of the country without interruption since 1980. His humiliating defeat marked the end of the civil war.

8Under the leadership of a transitional government, Guinea-Bissau returned to constitutional democracy by way of general elections in November 1999 and January 2000. The opposition became the largest party in parliament and ousted PAIGC from the government. The presidential election was won by the populist opposition candidate Kumba Yalá who scored 69 per cent of the national vote in the second round, against the PAIGC candidate Malam Bacai Sanhá, interim president under the preceding transitional period. Kumba Yalá was installed as new president for a five-year period in February 2000, supported in parliament by his party PRS.

  • 5  The 2003 putsch and subsequent developments were given good coverage by Lusa.

9Because of its flagrant inability to deal with the country’s developmental problems, Kumba Yalá’s regime was deposed by way of a bloodless military putsch on 14 September 2003 – gradually and somewhat reluctantly accepted by the ‘international community’ as inevitable under the circumstances.5 A transitional regime was set up, and after a period of quite peaceful constitutional transition, parliamentary elections were held on 28 March 2004, resulting in the return to parliamentary majority power of PAIGC, with Kumba Yalá’s party PRS (Partido da Renovação Social) in the second place. The election of a new president was delayed but finally took place in two rounds, on 19 June and 24 July 2005. It resulted, paradoxically, in the return to presidential power of the very same ‘Nino’ Vieira who in 1999 had been chased into political asylum in Portugal by the rebellious forces of his own army and by popular revolt. In the second round of the 2005 presidential election, ‘Nino’ defeated Malam Bacai Sanhá, candidate for a second time of ‘Nino’s’ earlier own party PAIGC, by a very narrow margin. The victory was juridically contested by the loser and by the ruling party that accused Vieira of having cheated. In a heated and paralysed political atmosphere, João Bernardo ‘Nino’ Vieira was nevertheless reinstated as President of the Republic on 1 October 2005.

10Before, during and after the 2003-2005 interlude, governments continued to move in and out of office. Strong dissatisfaction among the military, who often do not receive their pay, is an important factor fostering political instability. Furthermore, the weakness and consequent vulnerability of state institutions in Guinea-Bissau is currently turning the country into a favourite place of transit for South American drug-trafficking to Europe (UN 2008). Large amounts of Colombian cocaine are re-loaded in Guinea-Bissau for further transport to European markets.

11True enough, democratic institutions are in place, but civic and human rights are far from guaranteed and the level of life of the people is deteriorating in rural as well as urban areas. Power in society seems linked more closely to the possession of arms and illicit drug money than to popular support. Thus, the gap between state and society appears as wide as ever and perhaps even widening, in spite of constitutional democracy.

12On 16 November 2008, parliamentary elections were held again in Guinea-Bissau. Initial reports were that election-day had passed off in peaceful and orderly fashion, resulting in a convincing two-thirds majority for PAIGC (Lusa 16-21.2008). At first glance, this kind of outcome might have seemed to forebode a measure of political stability. But less than one week after the election, during the early hours of the night of 23 November 2008, the President of the Republic barely survived an armed attack on his life mounted at his residence by a group of nine soldiers from his own navy (Lusa, 23-26.2008).

  • 6  For facts and analyses, see Lusa’s continuous and detailed reporting.
  • 7  At the moment of writing these lines in early August 2009.

13The following months were marked by fear and threats of violence between the various military and political factions. Finally, on 2 March 2009, before sunrise, President João Bernardo ‘Nino’ Vieira was murdered in his home by a group of soldiers, only hours after the death of the chief of the army, General Tagmé Na Waié, in a bomb attack.6 No serious investigation, let alone trial by court of any suspects, has taken place so far.7

14Under pressure from ‘the international community’ in various shapes (the United Nations, the organisation of West-African states, the organisation of Portuguese-speaking nations, the European Union, Portugal, and others), Guinea-Bissau did after the initial shock begin to prepare for the regular election of a new president. The campaign was partly violent. One presidential candidate and one other leading politician were murdered by soldiers in early June 2009 (Lusa, 5-8.6.2009), and others, such as former Prime Minister Francisco Fadul, brutally harassed. Still, presidential elections were held in two rounds on 28 June and 26 July 2009, resulting in the end in the reversal of the January 2000 result. The second round was held between the two of the original eleven candidates who had done best in the first round, none reaching 50 per cent. This time the PAIGC candidate Malam Bacai Sanhá did defeat his opponent Kumba Yalá by getting 63 per cent of the votes (Lusa, 29.7.2009). A great number of observers testified to the peacefulness and fairness of the election process as such. A more telling fact, however, is probably that on both election days as many as about 40 per cent of the registered voters simply did not turn out to vote (Lusa, 2.7.2009 and 30.7.2009). This is a much higher abstention rate than in previous elections in Guinea-Bissau, with for instance 25 per cent of registered voters abstaining in the parliamentary elections of 2004 and 12 and 21 per cent respectively in the first and second rounds of the 2005 presidential elections (Rudebeck 2009).

15All through the period of democratisation now described, welfare conditions for ordinary people have been stagnating or worsening, regardless of some GDP-growth in the nineties as well as more recently in 2004-2007, due to good cashew harvests (World Bank 2008). In 1997, before the eruption of civil war, Guinea-Bissau had occupied place 168 among 174 in the UNDP’s Human Development Index ranking (Human Development Report 1999 :137). Ten years later, after seven years of democratic restauration, the equivalent ranking was 175 out of 177, thus even worse. This means, in the concrete, that in 2007 average life expectancy at birth was estimated to be 46 years (45 in 1997) ; adult literacy 45 per cent (34 in 1997); school attendance 37 per cent (36 in 1997) ; while average per capita purchasing power was estimated to be one sixth of the average for all ‘developing countries’ (one fourth in 1997) (Human Development Report 2007/2008 :293).

  • 8  In December 2007, for instance, the Minister of Finance presented a budget in the National Assembl (...)

16At the same time, in a globalised world system, the country’s dependence upon international aid continues to be extreme. This structural condition is not accurately reflected in aid statistics, as flows of aid have been slower to materialise after 1999 than before, due to low credibility of the regime in the eyes of ‘international community’ donors. Still, in 2005, official development aid received accounted for 26 per cent of gross domestic product (GDP), as compared with one per cent for all developing countries and 9 for the category labelled ‘least developed’ (Human Development Report 2007/2008 :293). In a more qualitative sense, aid dependence as a political and ideological constant is distinctly reflected in the leadership’s recurring references to aid as a key solution to the country’s problems.8 Vulnerability in the face of potential ‘earnings’ from drug trafficking is the other side of the same coin. Substantial dependence for development upon the democratically organised support of the people seems far away, in spite of constitutionally correct elections.

17All of this, obviously, is in glaring contrast to hopes and expectations once raised by the successful 1963-1974 struggle for liberation from colonialism.

A village study

18I have described and analysed the transition now summarised in several books and articles which will be drawn upon here (e.g. Rudebeck 1997, 2001, 2002b, 2008b). The special feature of the present article is that it selectively highlights one aspect of this work over the years by focusing upon the political repercussions in one particular village of the overall democratisation occurring in two phases in Guinea-Bissau : firstly, the introduction of democracy by way of multi-party elections in 1994 ; and, secondly, democracy’s fragile re-birth without concurrent development since 1999-2000, after collapse in 1998.

19As a result of multiparty elections generally, citizens at all levels of society are drawn into the democratisation process, if in no other way than by registration as voters. The practical preparations for elections and their implementation also activate many people at the local level. Whether this is decentralisation of political power or not is a completely different matter. From the point of view of substantial democratisation the decisive point is whether the elections do, in practice, empower citizens to jointly formulate their interests and wishes, appoint their representatives and demand accountability. Or, whether the principal significance of the elections is on the contrary to help the leaders mobilise their supporters to legitimise and promote these same leaders’ own interests. In this respect Guinea-Bissau is not alone in the world in inclining significantly more to the latter than the former. But, in Guinea-Bissau as in other places, there are features of both. These features are difficult to capture in quantitative terms and therefore the method of using an illustrative case study can be of great assistance. Such a case study follows below, taking what I like to call a diachronic perspective.

Kandjadja – a village among thousands

  • 9  In Guinea-Bissau the name of the village is usually spelt Candjadja, which is the Portuguese spell (...)
  • 10  The villages in the section are (with official spelling): Candjadja, Dabocunda, Corinto, Sabalacun (...)
  • 11  The point of departure for my population estimate is the number of persons entitled to vote and th (...)

20There are almost four thousand villages of varying size in Guinea-Bissau. I have had the opportunity to follow one of them ever since 1976. We shall obtain our material from that village. The village is called Kandjadja. It is a normal Guinean village, similar to thousands and tens of thousands of other villages in rural Africa.9 I have described it many times before (e.g. 1992, 1997, 2001, 2002 :b). The directly visible changes take place slowly, but life and history do not stand still. With its thousand or so inhabitants Kandjadja is one of the larger villages in Guinea-Bissau. It is the main village in an administrative section that also includes twelve smaller villages spread out in the forest south of the river Farim, which flows westwards towards the Atlantic sea.10 The whole section had somewhere around 4,500 inhabitants in 1999 and thus perhaps close to 6,000 in 2009, although I have no firm ground for judging population growth and out-migration against each other.11 The area is situated in the administrative sector of Mansabá, in the Oio region, in the northern part of Guinea-Bissau. As the bird flies the distance is not more than twenty-five kilometres north-west to the border with Senegal and its politically unsettled Casamance province. From Kandjadja to the sleepy regional capital of Farim there is a bumpy road, twenty kilometres long, through the forest in a north-easterly direction. Bissau is some one hundred and ten kilometres to the south, including twelve initial kilometres of track road eastwards through the forest, before reaching the main road between Farim and Bissau. Cars make only rare apperances in the village.

21The Muslim Mandinga people dominate in Kandjadja, but ethnicity-wise there are also ‘Animist’ Balanta and Muslim Fula in the section. The main village is entirely Mandinga and thus its religion is Islam. The livelihood and economy of the people are based upon agriculture in a wide sense of the term, pursued with simple techniques on the basis of proven methods and, in addition, certain handicrafts. There is also some inflow of money from villagers employed in various occuptions in Bissau or abroad. The amount of this is unknown but hardly impressive.

  • 12  Just after the civil war, the large family I know best told me that, from time to time, they had h (...)

22In recent years cashew nuts have been the most important source of cash income. Previously it was groundnuts. Sometimes an enterprising villager takes other products, for example a sack of onions, all the way to the large Bandim market in Bissau. During the civil war in 1998-99 the area around Kandjadja was not directly affected by acts of warfare. Nevertheless, the war and the ensuing crisis of subsistence and survival were still felt harshly, not least through refugees coming from Bissau.12

  • 13  I have described this in some detail, including the far-reaching effects of economic liberalisatio (...)

23Since the beginning of the 1980s the population has been left to itself, as far as development is concerned.13 In an interview made in 1986, a villager told me that “they (the state) have left us in a hole”. The same could just as well have been said at the time of my visit in 2007. The forest road leading to the village was then in worse condition than it had been in 1976, over thirty years earlier, at the time of my very first visit. The state-employed nurse of the village had already left before 1981. The former state-owned store passed into private ownership in 1986, as a result of the structural adjustment reforms sweeping over Africa at that time in accordance with World Bank and International Monetary Fund prescriptions. Simple consumer goods are found in the store, but the purchasing power in the village is limited. The state school was closed in 1989, as the government gave up its earlier ambitions in the field of public education.

A paradox

  • 14  59 per cent of the Kandjadja voters still supported PAIGC in 1994 as opposed to 38 per cent in the (...)

24Politics in Kandjadja may seem paradoxical. The village and the whole area are, in practice, untouched by government development policies. Literally nothing has reached there in terms of benefits of national policy since the last school teacher left in 1989. Still, for all that, PAIGC, and state power under PAIGC control, have always been able to count on political support from the area where Kandjadja is located. Even in the multi-party elections of 1994 – the first time ever that it was possible to vote against PAIGC – the outcome was quite distinct.14

25The general explanation provided locally for such paradoxical loyalty is that “this is an ‘antiga zona libertada’/an old liberated area”, i.e. an area controlled by the PAIGC as early as in the anti-colonial war between 1963 and 1974. This, today, is not a real explanation. It is rather an indirect way of saying that the only political contacts existing between the village (meaning: a few of the villagers) and the politicians in the capital go through PAIGC.

  • 15  See Rudebeck 1992, in particular pp. 269-270, with full references to earlier works.

26This does not mean that people are satisfied with being “left in a hole”. The demand for change (mudança) is as strong in Kandjadja as in other places in the country. Still, in 1994, 1999, 2000, 2004 and 2005, there were as we shall see more votes in the village cast for PAIGC than for any of the other parties. Surely these outcomes may be traced largely to links established during the war of liberation, but in a more complex manner than indicated by mere reference to political tradition.15

  • 16  The homem grande of Kandjadja village took over more or less formally as the chairman of the secti (...)

27What happened in the area in the 1960s was that the PAIGC political structure was grafted onto the local ‘traditional’ structure of extended families led by homens grandes (great men, singular: homem grande) of varying standing, the most prominent of whom were also recognised as leaders of villages and groups of villages. Through this structure, the area of Kandjadja was largely mobilised on the side of PAIGC during the liberation war. During the immediate post-war years, a centralised attempt was made to impose a ‘revolutionary’ political structure with a party committee headed by an outsider sent in by PAIGC to be in charge of the thirteen-village section of Kandjadja. With the resilience of the traditional political structure and fading revolutionary ardour, this reverted quite soon to a situation where the party/state simply maintained elementary political control by relying on the loyalty of locally more or less influential people – ‘clients’ in relation to regional and national levels. These persons in turn continued their co-existence with the rest of the villagers as weak local ‘patrons’, with nothing concrete and ever less hope to deliver in terms of development.16

28We will soon return to more specific considerations of the Kandjadja paradox. Let us see first how democratic elections worked in the village – after a devastating civil war on top of ordinary hardship. How did the local voters rate the political parties, in particular their old favourite PAIGC? Did the promises made by the opposition seem to influence them? How did the opposition’s leading presidential candidate, Kumba Yalá, fare, as the voters were facing his image on the voting bulletin?

29The following account is based on observations and interviews made in connection with visits in May and December 1999, January 2000 and June 2007. To facilitate use of the tables presented, they have all been assembled in a special section at the end of the text.

Democratic elections at village level

Parliamentary elections 1999 and 2004

  • 17  In the section there were five other polling stations besides the one in Kandjadja (which included (...)
  • 18  Interviews on 4 and 11.12.1999. During the latter interview the chairman carefully consulted the n (...)

30As in several other places in the country, voting was delayed in Kandjadja on Sunday, November 28, 1999.17 According to the chairman of the section committee for Kandjadja, the ballot box had arrived late. He was also the chairman of the section’s committee of support for PAIGC and had kept precise notes. According to those notes, voting in the village had started at 15.52 and, in agreement with the rules, had ended at 17.00. Since this was too little time for everyone to have had a chance to vote, the voting had been continued on Monday between 7.30 and 15.00.18

  • 19  Guinea-Bissau’s various political parties will not be presented here. See my book On Democracy’s S (...)

31The reason for the delay in the arrival of the ballot box was most probably a nation-wide threat of strike by the voting officials. Demands for an increase in pay made by certain voting officials in the region had also caused problems the day before. It had therefore been impossible to have the polling equipment in place in the village on Sunday morning. Once the voting finally started, everything had proceeded very calmly and in orderly fashion. According to the chairman of the section, this was only “to a certain extent” due to civic awareness, since the political parties, in his view, had not taken the trouble to make it clear to the people what they represented.19 Consequently the voters could not be expected to really know what they were voting for. Still the citizens had taken everything calmly and turned out strongly to vote, but apparently without great expectations, according to the section chairman.

32Despite the fact that Kandjadja is situated off the beaten track in the forest, with very poor communications, several political parties had visited the village to mobilise voters before election-day. The constituency consisting of the administrative section of Mansabá, of which Kandjadja is a part, sends three members (deputies) to the People’s National Assembly/parliament (Assembleia Nacional Popular/ANP). PAIGC’s candidates were well known in the area and had as always promised to work for schools, health services and road construction if they were elected. Even “Helder Vaz’s people” from ‘Bafatá’ (RGB/MB), the major former opposition party besides PRS, had visited the village, with their promises of similar kind, as well as representatives of ‘Abubacar’ and ‘Bubacar’, leaders respectively of the two small parties UNDP and LIPE. But besides Malam Bacai Sanhá, PAIGC’s presidential candidate as well as interim President of the Republic, there was only one other party leader and presidential candidate who had visited Kandjadja in person during the election campaign before 28November 2008. This was Victor Mandinga from AD. On the other hand Kumba Yalá’s party, PRS, had not shown up at all in the village. This was generally explained by the fact that Kumba Yalá is ‘Balanta’, while Kandjadja is ‘Mandinga’, which is true although not necessarily the whole explanation.

33The Kandjadja votes were counted on site under the close supervision of representatives of the various parties. They were then taken to Mansabá and included in the votes for the entire constituency. The results for the village, including Kandjadja-Balanta, are shown in tables 1 and 2 at the end of the text. The first table compares the results of the 1999 and 2004 parliamentary elections in Kandjadja with the results for the entire country. As seen, PAIGC was considerably stronger in the village than in the whole country. The difference is greater in 1999 and in 2004 than it had been in 1994 (cf. note 10 above). This is true for both years.

34In 1999, immediately after the civil war, PAIGC was down to an all-time low of 15 per cent at the national level. But in Kandjadja, 40 per cent of the 432 voters still voted PAIGC, in spite of the defeat of the regime. Two specific explanations may be mentioned:

  • PAIGC’s candidates for Mansabá constituency’s three mandates in parliament were well-known in the village.

  • There was a direct personal link to the PAIGC presidential candidate Malam Bacai Sanhá’s campaign office.

35The upward link consisted of the fact that a “son of the village”, who had been a teacher in the village school in the 1980s, worked in 1999/2000 for Malam Bacai Sanhá’s wife in PAIGC’s national campaign secretariat.

36Furthermore, according to information from several sources that cannot be verified due to election secrecy, all, or almost all votes cast in favour of the opposition presidential candidate’s party, PRS, originated from Kandjadja-Balanta, a small neighbouring village, administratively part of the main village and situated a couple of kilometres south of it. In that village, all the inhabitants are Balanta. Thus they are not Muslims, they do eat pork, and they like to drink palm wine. They also speak a completely different language from the Mandinga language spoken in the main village. For all these reasons, and others as well, they were tempted in 1999/2000 by the idea of asserting their cultural identity by voting for PRS and its presidential candidate of Balanta ethnic origin. As opposed to the nation-wide distribution of votes which was clearly influenced by a combination of non-ethnic and ethnic factors differing from place to place and between the national and local levels, we do seem here to have a very clear-cut example of ethnic voting.

  • 20  Resulting in 45 out of 100 seats for PAIGC in parliament, enough to form a national majority coali (...)

37In the 2004 election, PAIGC generally did considerably better than in 1999, due to the disastrous developmental and political performance of the regime that had been voted into power in 1999 and 2000 but forced out through the 2003 putsch. At the overall national level this recuperation of PAIGC reached 31 per cent,20 while soaring to an overwhelming 67 per cent in Kandjadja.

38Table 2 shows in detail how people actually voted in Kandjadja in the parliamentary elections of 1999 and 2004. To facilitate comparison between the parties and years, the table also gives percentages.

39The distribution of party sympathies in Kandjadja was different from the whole country, not only with regard to PAIGC, but also in other ways. For example, UNDP, a new and very small party at the national level (3 per cent of the total vote) was the second largest party in Kandjadja in 1999 with 34 votes of the 432 votes cast – more than PRS and ‘Bafata’ (RGB) which both in that year outscored PAIGC at the national level. I was told that UNDP’s success had been achieved by donating “a little medicine” and a new mat to the mosque.

40It is notable in table 2, among other things, that as many as 11 of 13 national parties competing in 1999, and 12 of 15 in 2004, actually succeeded in obtaining at least a few votes even in an isolated village, such as Kandjadja. Those parties that failed to attract any votes whatsoever in the two parliamentary elections in the village are very small even at the national level. One of them is FCG/SD, the party of the only female presidential candidate, Antonieta Rosa Gomes, a lawyer who ran for president both in 1999–2000 and in 2005. Although her party scored zero in the village, Rosa Gomes still did obtain one (1) of the 436 votes counted in the Kandjadja ballot box for president in the first round of the 1999–2000 presidential election. We can only speculate about the driving forces behind that single, and singular, vote for a woman as President of the Republic.

  • 21  See Rudebeck (1997 :38-44), on positive recognition in the village of greater civic freedom couple (...)

41It is, in any case, a significant observation about political democratisation in Kandjadja that several citizens do seem to make use of their new-won right to cast a vote of their own, freely and secretly. They value the relative civic freedom coming with democratisation. The same persons are sadly aware, though, that just spreading their individual votes among a dozen political parties quite alien to the village will hardly bring them any political influence, let alone power.21 The majority who did continue to vote for PAIGC are also likely to have been disillusioned about the developmental outcome of their support, while a few of them may still have been hoping for some personal benefit in exchange for the vote delivered.

Presidential elections 1999/2000 and 2005

  • 22  For the 1994 presidential election in Kandjadja, I only have exact data on the second round. In th (...)

42The pattern of the first round of presidential elections is similar to that of the parliamentary elections. On 28 November 1999, Kumba Yalá even received exactly the same number of votes for president as his party PRS did in the parliamentary election taking place simultaneously. On the other hand, Malam Bacai Sanhá received considerably more votes than PAIGC – as many as 313, which corresponded to 72 per cent of the total. This is much more than the 20 per cent he reached in the whole country. It is also the same pattern as in 1994, in the first multi-party election ever to have been held in independent Guinea-Bissau.22 Thus, the impression that Kandjadja in 1999 stuck to old political habits, as opposed to the rest of the country, was even stronger in the presidential than in the parliamentary election. Table 3 at the end gives further detail.

43The ethnic, cultural factor that seems, according to my sources, to explain why most voters from Kandjadja-Balanta voted for PRS and Kumba Yalá worked also in the opposite direction. Only very few people voting in the main village cast their votes for the “Balanta party” and for its leader in 1999–2000, according to the same sources.

44A locally acute ethnic conflict, furthermore, influenced electoral behaviour in 1999/2000. Robbing cattle from neighbouring villages is a traditional test of manhood of young Balanta. In recent years several households in Kandjadja-Mandinga had suffered from that cultural tradition – or criminal habit, as some would say. Things had even gone far enough to cause some villagers to keep their cows locked up at night, for fear of cattle thieves from the neighbouring village. Anger at this insecurity fostered negative feelings for PRS and scepticism about its presidential candidate with his red Balanta cap. Regardless of how much they agreed with him on the need for mudança/change, many villagers feared a hidden ethnic agenda: ”Perhaps Kumba Yalá doesn’t like us ?”

45However, the conflicts over cattle and other material and cultural differences did not keep the inhabitants of Kandjadja-Mandinga and Kandjadja-Balanta from having normal social relations, which did, on the surface, appear uncomplicated and even friendly. For instance, people from the Balanta village do their shopping in the store in the main village ; they vote there ; and until 1989 their children went to school there with the Mandinga children.

46The second rounds of the two presidential elections were straightforward. The choice was either for or against : Kumba versus Malam in 2000 ; ‘Nino’ versus the same in 2005. The inviting images of the candidates faced the voter from the ballot paper. In 1999 Kumba Yalá was wearing a red Balanta cap and Bacai Sanhá the same type of patterned West Africa knitted woollen cap as the historic leader of the anti-colonial liberation struggle, Amílcar Cabral, used to wear in his time, before he was murdered by a member of his own party in 1973. All that was required of the citizen was to step behind the screen and mark her or his preferred candidate. Table 4 shows the outcomes for both years, in absolute as well as relative terms. Table 5 adds the overall national comparison.

47In statistical terms, as seen in the tables, the villagers’ support for the PAIGC candidate Malam Bacai Sanhá was overwhelming in both years, around 90 per cent. This was very much above his national level scores. But in fact, as we know, he lost his bid for the presidency on both occasions : resoundingly in 1999 against the PRS candidate Kumba Yalá and narrowly in 2005 against his former party comrade João Bernardo ‘Nino’ Vieira. In Kandjadja, though, PAIGC prevailed as ever. Even ‘Nino’ himself running as a non-partisan ‘independent’ only got 33 votes in the village in 2005, as opposed to 328 eleven years earlier, in 1994, when he was still  the leader of the country and of PAIGC (see note 18).

Electoral democratisation in the village

48We have seen that the voters of Kandjadja do make use of their freedom to vote in the simple sense that they distribute their votes among many parties. But even after eleven years of electoral democracy, a large majority continued to vote the old (single-) party line in spite of being “left in a hole” as far as development is concerned. In this they are considerably more ‘conservative’ than the average voter in the country. A number of mutually reinforcing explanations have emerged from the foregoing presentation. Let us summarise them :

  • Kandjadja is an “old liberated area”, indicating a specific kind of historical tradition.

  • There are few obvious differences between the parties’ programmes, platforms, ideas or proposals, and therefore most voters have to rely on personal impressions and authoritative persons in their local community.

  • The people have very limited experience of any other ‘upward’ political channels than PAIGC.

  • What matters in the PAIGC hierarchy is personal contacts, either directly or through local ‘patrons’, with PAIGC’s representatives at regional and national level, for instance in parliament.

  • There is an ethnic, cultural Mandinga versus Balanta factor working against PRS and Kumba Yalá, reinforced at the time of the studied elections by acute problems with cattle thefts.

  • There was at the time of the studied elections a direct link between the village and Malam Bacai Sanhá by way of a respected “son of the village”.

49A special event had also impressed the villagers. On 10 January 2000, fairly late in the evening, after dark, on the last Monday before election Sunday, Malam Bacai Sanhá, the interim president of the country and presidential candidate for PAIGC had suddenly arrived in Kandjadja. Driving in from the north, from the regional capital Farim, with his campaign team, he had stopped for a public meeting in the middle of the village which was dramatically illuminated by the headlights of his cars. He spoke to the villagers and answered their questions for about half an hour. He had said, not surprisingly, among other things, that if he was elected president, he would help the people of Kandjadja section to get water, schools and health care. The message, I was told, had been received with a “certain amount of trust”. Malam Bacai Sanhá had been wearing his woollen Amilcar Cabral cap, a symbol of ‘revolutionary’ legitimacy. A great deal of interest had been aroused. Once the meeting was over, the national leader had continued his journey southwards through the night to a larger town called Olossato. Darkness had returned to the village, broken up only by small fires in front of the houses where people gathered to discuss the singular event they had just experienced.

50The chairman of the section emphasised to me that this was the first time ever in Kandjadja’s history that a President of the Republic had visited the village. Kumba Yalá for his part, he added, had never visited the village.

51Thus, on second thoughts, perhaps the voting behaviour of the people of Kandjadja is not so difficult to understand. On the contrary, in the concrete context of prevailing conditions, of which the villagers are well aware, there may well be a type of disillusioned rationality in it. If so, what are the implications for democratisation and democracy? And, what are the implications for ‘civil society’?

Civil society

  • 23  The theoretical and theoretically inspired literature on civil society, democratisation and develo (...)

52‘Civil society’ is a much-used concept. But for analytical purposes, it is diffuse and contested and therefore somewhat difficult to use. At a minimum, though, most of those who do try to use the concept for analytical purposes agree that it refers to people acting together in the social space existing in society between the state and the private sphere. This obviously makes it relevant to democracy theory.23

53At the overall level there are in principle two ways of increasing the opportunities for citizens to assume responsibility in society, once single-party authoritarianism is done away with. One is that the state draws back, leaving space for citizens to act more independently of the state than before – in ‘civil society’. The other way is that the institutions of the state are democratised and decentralised. At the end of 1995 the constitution of Guinea-Bissau was revised so as to allow for the establishment of elected local assemblies (Rudebeck 1997 :44-46), but no further steps in that direction have been taken since. There is no reason to assume that such assemblies would have any effects in practice, unless provided with locally controlled financial resources, which are unavailable anyway.

54The strategic issue for democracy is how the two approaches can supplement and support each other. In its political practice so far, as illustrated by the case of Kandjadja, the Guinean government has withdrawn far in favour of ‘civil society’ at the local level, except for the occasions of national elections. The question has been avoided or evaded, how far such practice can be carried without large groups in the population ending up completely without access to common resources. Rural Africans are known to be patient and resilient, but there may still be limits which cannot be transcended without breaking the bonds of society.

‘Civil’ and ‘political’ society

55A distinction is sometimes made between ‘civil’ and ‘political’ society. It is a question of definition for political and other social scientists whether, for instance, political opposition parties should be included in ‘civil’ or in ‘political’ society. The parliament is part of the state, but the political parties also act in society at large, which could give them a place in ‘civil’ society. Törnquist (1999 :50) suggests that ‘political society’ can be seen as a link between purely ‘civil’ society and the state. Then, obviously, the opposition parties will be crucial to that link.

56Using the distinction between ‘civil’ and ‘political society’, it is of particular interest in our context to note that constitutional democratisation in the studied village affected the two in very different ways. Democratisation did indeed result in the politicisation of the villagers in the sense of getting them to participate actively and willingly in national multi-party elections – largely in support of the old regime. Thus, ‘political’ society was clearly activated and also democratised in the constitutional, minimalistically conceptualised sense. What happened in ‘civil’ society is however more difficult to pin-point. As we shall see, it was activated, too, and possibly also politicised, but not in a democratic way and not in a way involving the decentralisation of political power.

‘Civil society’ at village level

  • 24  RADI is an African NGO for culture-based development work in agriculture. The countries that are r (...)

57Before 1990, one-party rule had made it impossible in practice for voluntary organisations to participate in development work in the rural areas of Guinea-Bissau, unless they enjoyed the direct support of the state and the party PAIGC. But in 1990, the first year of democratisation, something very new happened in the village. Representatives of RADI (Réseau Africain pour le Développement Intégré) began to visit Kandjadja and to encourage the population to start programmes to develop local farming based on their own traditional forms of organisation.24

58Since then, RADI has continued to function in the region of Oio where Kandjadja is located. Seventeen kilometres east of Kandjadja, in a neighbouring administrative section, the organisation has put up a training centre for farmers in the forest near the village of Djalicunda.

59The centre is very ambitious. It is built with local materials, has its own source of fresh water, and even computers working with the aid of solar cells. Local varieties of seeds are cultivated in ways that take the environment into ecological consideration. Those working at the centre in Djalicunda are locally recruited Guineans. Various training courses are held there and the intention is that farmers in the area shall be inspired to integrate modern farming techniques with the use of local seeds and due respect paid to tradition and the environment.

  • 25  Translated from the French language original by Lars Rudebeck.

60According to a written report delivered in 1999, the key concept of RADI’s theory is “cooperation with the people” (RADI 1999a:6-725) :

In Guinea-Bissau there is real potential to develop the rural sector, since land is still available for further cultivation, and there are also possibilities to intensify cultivation, to diversify forestry, to extend local fishing and to increase the herds of cattle. But it is important to note that a strategy to exploit all these possibilities requires more than technical investments. It is necessary to make great efforts to stimulate the interest of the rural population, to train, to organise and reorganise, to improve the motivation of the technicians – all with the aim of getting those involved in the rural areas to assume more responsibility and to strengthen the position of women. This is the focus of RADI/Guinea-Bissau’s support for development.

61Another report described food support provided through RADI to 27 villages in the region of Oio that received refugees during the civil war in 1998/99 (RADI 1999b :2, 4).

62Although Mansabá sector was included in the food programme, Kandjadja, was not among the 27 villages listed in the report. This is possibly due to the fact that support for RADI’s work seemed to have declined in Kandjadja during the second half of the 1990s. It is clear that opinions in the village on RADI are quite divided. Some villagers work actively with the organisation and have succeeded in benefitting from it. Others, however, express dissatisfaction with RADI’s work and find that it focuses more on the organisation’s own needs than on the needs of the local people (interviews, December 1999 and January 2000).

63These observations indicate that in 1999 and 2000 the kind of politicisation of civil society in the village that the activities of this particular NGO had resulted in was more patron-client oriented than characterised by the emergence of autonomous grassroots citizenship. Civil society had been activated, but not democratised.

64The shortcomings and even the absence of the state in developmental activities had created a vacuum, as well as deep popular disenchantment in Kandjadja. This in turn had opened up chances for an organisation such as RADI to strike a chord of strong response within civil society. Thanks to democratisation, alternatives had become legal. Still, the democratic vacuum had not been filled.

A ‘civil society’ initiative

65Visiting Kandjadja on 29 June 2007, I was surprised by a novelty. In fact, I was told that in 2006, deep and growing frustration over the lack of any elementary education whatsoever in the village, except the Koranic school, had in the end caused a number of families to join forces to set up a school of their own.

66Ever since 1989, when the state run school was closed, only the Koranic school had been available for the children of Kandjadja. The two finest houses in the village, the only ones with outer walls of white plaster, are in fact the mosque where the Koranic school is held and the house where the imam (the leader of the mosque) lives.

  • 26  On the occasion of my visit, the teacher was on summer vacation.

67As I was told, the idea of the new school is simple. In 2007, it had one teacher and a little over one hundred pupils enrolled, divided between the first, second and third grades of elementary school. The teacher had been hired from outside the village.26 The families pay him 350 CFA francs (0.53 euro) per child and month. They also provide him with food and a simple house. If some family is unable to pay, others help out.

  • 27  Information received in Uppsala by mobile phone call from the chairman of the section committee in (...)

68In June 2009, my only further information on this grassroots educational project was that the school was still working, attended by about 70 children.27 But whether it survives into the future or not, the fact of its existence in 2007, 2008 and 2009 does represent a departure from the patron-client pattern. It demonstrates, in practice, the idea of citizens getting together in ‘civil society’ to solve a problem they have in common. The problem in this case is due to the dismal failure of the very same state that those citizens were once prepared to trust. Their project can be interpreted as an acute emergency attempt at self-empowerment by a local community. The outcome remains of course to be seen.

The wider context of power, resources and culture

69Understanding the politics of development is, at the very least, a three-dimensional task involving the integrated analysis, of not only power and resources, but also of cultural norms and patterns of behaviour. Let us end this article by briefly setting our West African example in that wider context.

70The problem of democratisation resulting in local civil society being activated rather than democratised is not unique to Guinea-Bissau. It has general relevance. When the local political process is focused upon and reaches into family and personal interests of survival and reproduction, it will very easily take the form of ‘patron-client’ relationships rather than democratic organisation. Both citizens and leaders take advantage of the few opportunities available to them to promote their short-term interests. As far as the citizens are concerned, in situations of harsh material poverty, this is principally a question of day-to-day survival; while for the politicians it is more a question of power resources. As long as the citizens cannot discern any concrete reasons to assume that public, collective, action might actually help them survive and improve their lives, the likelihood is great that they will continue to turn to their patrons, even under the guise of the constitutionally democratic system or through various civil society activities. This is probably not for trusting those patrons, but for not seeing any realistic alternatives. Such minimal democracy, thus, is likely to support political practices incapable of giving rise to development.

71This looks like an impasse, at least in the short run. But in the somewhat longer run, such negative dynamics are not likely to be sustainable. Both the development necessary for people to survive in acceptable ways and their readiness, as well as capacity, to support their patrons’ power is undermined.

72The example of Kandjadja is instructive by clearly demonstrating how little the local people get in return for their political support. But it also suggests certain limits to their patience. Such limits are indicated by the decline in voting support for PAIGC between 1994 and 1999, however modest in comparison to national results, and more distinctly by the start of the new school in 2007. Still, so far, the self-perceived short-term interest of the villagers has not incited them to new forms of politicisation, beyond participating in multiparty elections.

  • 28  To illustrate, Leys 1996 offers a critical analysis of such development studies, while Chabal & Da (...)

73It is a common view in modernisation school inspired development studies that particularistic political culture or codes of conduct marked by patron-client relations, as in Kandjadja, and corruption would somehow causally explain the facts of unequal resource control and lack of development in ‘developing countries’, and today most particularly in Africa.28 That kind of argument, however, risks coming close to tautology or question-begging, by tending to explain political clientelism and corruption as caused by culture or codes of conduct marked by – clientelism and corruption. The same argument also tends to overlook that patron-client relations and corruption as such certainly have not prevented economic development in other parts of the world, whether, for instance, historically in the West or in modern Asia.

74By concentrating – or appearing to concentrate – the explanatory effort to cultural variables, the question of power in society tends to be avoided. True enough, an inherited culture of client deference to patrons obviously supports unequal power and thus does not facilitate democratisation. But in accordance with Max Weber’s classical sociological perspective, culture is more fruitfully analysed as the ‘switchman’ between interest and action than as an overall explanation of the functioning of society (cf. Rudebeck 1994, 1998). Culture provides values, images and frameworks for interpreting reality. Still, when necessary to daily survival, even democratically minded persons often bow to patrons or bosses, however reluctantly. Likewise, people whose cultural values are non-democratic may well accept to share inherited power if necessary to their political or perhaps even physical survival.

  • 29  Cf.the following incisive remark by Mamdani (1996 :295 note 2) on “the mode of domination” as an e (...)

75In describing the politics of Kandjadja and its links to national politics in this text, I have myself made use of the concepts of ‘patrons’ and ‘clients’ to grasp the hierachical structure obstructing substantial democratisation. I have not, however, designated client deference to patrons as a Guinean cultural norm or trait, causing the village to be “left in a hole”.This is not to say that patron/client relationships do not mark human behaviour. They certainly do.But the decisive question on clientelism in regard to development is about the power relations that determine clientelism rather than its existence as such, of which we are usually well aware.29

76In the theoretical perspective applied here, therefore, the core of the conceptual theme of democracy and democratisation in relation to development is about material interests and relations of power. Culture is conceptualised as the Weberian ‘switchman’. What material forces or interests and structures of power do actually obstruct democratisation ? Can they be challenged ? How do cultural norms, codes and patterns of behaviour intervene in this ?

77Horizontal self-organisation from below by people with similar developmental interests – across lines of ethnicity, gender and age – would thus be crucial to gradually breaking up vertical patron-client structures at all levels, including the local one. However simple or modest, and whether sustainable or not, the recent school project in Kandjadja concretises this core issue. Furthermore, the project may well come to contribute to cultural change in the village. Whether such change is likely or not to facilitate self-empowerment and substantial democratisation as referred to in the introductory part of this article, will in turn depend on the kind of teaching provided in the school.

78This may seem like a tiny conclusion for a study aiming to grasp such a lofty abstraction as ‘substantial democracy’. Though possibly tiny in a quantitative sense, it does nevertheless hold a grain of – substance.

Tables referred to in the text

Table 1. Comparison of national assembly election results in Kandjadja and the country as a whole, 28 November 1999 and 28 March 2004 (results rounded to the nearest percentage point)


Whole country










Other parties





Blank + invalid





Total percentages





Sources : The Kandjadja percentages are based on figures noted by the local voting official, filled in by the chairman of the section committee for Kandjadja on the basis of the voting official’s own information received directly from the polling station, 28 and 29.11.1999 (Kandjadja, 1999 national assembly election), and 28.3.2004 (Kandjadja, 2004 national assembly election). The 1999 figures for Kandjadja were furthermore checked against hand-written material from the deputy elected in 1999 for PAIGC in constituency no. 7, Mansabá, to which Kandjadja belongs (Mansabá sector, constituency no. 7, 1999). The 1999 national level percentages are based on figures published by Comissão Nacional de Eleições, Bissau, 9.12.1999 and 25.1.2000. The 2004 national level percentages are based on figures published by Nô Pintcha, 8.4.2004 and Lusa, 4.4.2004.

Table 2. Results of national assembly elections in Kandjadja, 28 November 1999 and 28 March 2004, total number of votes and rounded-off percentages

Number of votes















3 (3.4)










1 (0.8)





3 (3.2)


















1 (0.5)






2 (2.4)

Frente para a Dem.


2 (1.6)



1 (1.1)



1 (0.8)



1 (0.5)



0 (0.3)

blank + invalid











Sources : The figures for 1999 (Kandjadja, 1999 national assembly election) were noted by the local voting official and then filled in by the chairman of the section committee for Kandjadja in a copy of the original form, on the basis of the voting official’s own information received directly from the polling station, 28 and 29.11.1999. I also checked these data against hand-written material shown to me by the deputy elected in 1999 for PAIGC (Mansabá sector, constituency n° 7, 1999). The figures for 2004 were given to me directly by the chairman of the section committee on the occasion of my visit on 29 June 2007 (Kandjadja, 2004 national assembly election). The figures for total number of votes cast (381) do not quite match the sum arrived at (378) when the separate posts are added. I decided to put the latter figure in the table. There is also a small uncertainty in regard to PUSD, which appears twice in the data (2+13=15 = the figure found in the table). “Frente para a Democracía” is not found in the official statistics. It may refer to FDS (Frente Democrática Social). These minor inconsistencies and uncertainties do not in any way affect the credibility and validity of the data.

Table 3. Results of the first round of presidential election in Kandjadja, 28 November 1999, number of votes and rounded-off percentages

Number of votes

% of total

Kumba Yalá (PRS)



Malam Bacai Sanhá (PAIGC)



Faustino Fudut Imbali (independent)


1 (0.7)

Fernando Gomes (independent)


0 (0.46)

João Tatis Sá (independent)

Abubacar Baldé (UNDP)



Bubacar Rachid Djaló (LIPE)


1 (0.9)

Joaquim Baldé (PSD)


0 (0.2)    

Salvador Tchongo (independent)


1 (0.7)

José Catengul Mendes (FLING)


0 (0.46)

Mamadú Uri Baldé (PRP)

Antonieta Rosa Gomes (FCG/SD)


1 (0.7)










Total votes





8 (of 474))

Total registered votes


8 % (of 474)

Sources : The figures for 1999 (Kandjadja, 1999 presidential election, first round) were noted by the local voting official and then filled in by the chairman of the section committee for Kandjadja in a copy of the original form, on the basis of the voting official’s own information received directly from the polling station, 28 and 29.11.1999. I also checked them against hand-written material shown to me by the deputy elected in 1999 for PAIGC (Mansabá sector, constituency n°. 7, 1999). There were a few very minor inconsistencies between the two sets of data, in no way affecting their credibility and validity. The table is based on the local voting official’s figures.

Table 4. Results of the second rounds of presidential elections in Kandjadja, 16 January 2000 and 24 July 2005, number of votes and rounded-off percentages

Number of votes





Malam Bacai Sanhá (PAIGC)





Kumba Yalá (PRS)



‘Nino’ Vieira (independent)



Blank votes





Invalid votes



0 (0.2)


Total votes cast





79* Information not available on blank and invalid votes and thus not on “total votes cast” for the 2005 election in Kandjadja. Thus, the percentages in the table are not quite comparable between the two years. Calculated on the basis of “votes for candidates”, as for 2005, Malam Bacai Sanhá got 90 percent in 2000.

Sources : The figures for 2000 are based on hand-written records shown to me by the deputy elected for PAIGC in 1999 (Mansabá sector, constituency n°. 7, 2000). This was based in turn on information received on election-day by the PAIGC parliamentarian directly from the 51 polling stations. The polling stations were under the supervision of the PAIGC supervisor for the whole constituency, who was also the chairman of the section committee for Kandjadja. The figures for 2005 were given to me in Bissau on 30.6.2007 by mobile phone from the village by the chairman of the section committee for Kandjadja (Kandjadja, 24.7.2005 presidential election, second round).

Table 5. Comparison between Kandjadja and the country as a whole, second rounds of presidential elections in Kandjadja, 16 January 2000 and 24 July 2005, rounded-off percentages of votes for the two remaining candidates

% 2000

% 2005


Whole country


Whole country

Malam Bacai Sanhá





Kumba Yalá



‘Nino’ Vieira



Sources : For Kandjadja, same as table 4 ; for whole country 2000, Comissão Nacional de Eleições, 2000a ; for whole country 2005, Comissão Nacional de Eleições, 2005.


Literature, reports and news material

80Chabal, P. & J.-P. Daloz 1999. Africa Works : Disorder as political instrument, London: James Currey ; Bloomington : Indiana University Press.

81Hellström, E. 1995. Non-Governmental Organizations and Their Role in Democratization. A Minor Field Study of Guinea-Bissau. Unit of Development Studies, Uppsala University, Minor Field Study Series 9.

82Human Development Report 1999. United Nations Development Programme (UNDP), Oxford University Press, Oxford.

83Human Development Report 2007/2008. United Nations Development Programme (UNDP) :​en/​media/​HDR_20072008_EN_Complete.pdf

84Imbali, F., F. Koudawo, P.F. Gomes, C. Cardoso, P.K. Mendy Guiné-Bissau. Uma retrospectiva. Síntese, Instituto Nacional de Estudos e Pesquisa/National Institute for Studies and Research (INEP), Bissau.

85Leys, C. 1996. The Crisis in Development Theory. New Political Economy 1 (1), 41-51.

86Lusa: Agência de Notícias de Portugal, various issues, 2003-2009 :

87Mamdani, M. 1996. Citizen and Subject: Contemporary Africa and the legacy of late colonialism. Princeton University Press, Princeton.

88RADI (Réseau Africain pour le Développement Intégré/African network for integrated development), no date. Le RADI. Ce qu’il est et ce qu’il veut, presentation brochure, Dakar.

89–.1995. Projet d’appui aux associations rurales de la Région d’Oio pour leur reconnnaissance légale-juridique. RADI, Farim, April.

90–.1999a. Rapport d'activités et d'exécution financière du projet de renforcement institutionnel GW006031 au profit de RADI/Guinée-Bissau, mai 1997 à avril 1999. R.A.D.I., Farim, April.

91–.1999b. Rapport d’exécution technique et financière du solde de projet GW006011, dans le cadre d’un programme d’appui d’urgence aux populations victimes du conflit armé en Guinée-Bissau, dans la région d’Oio. RADI/Guinée-Bissau, Farim, 30.4.1999.

92Rudebeck, L. 1992. Politics and Structural Adjustment in a West-African Village. In When Democracy Makes Sense (ed.) L. Rudebeck, AKUT, Uppsala.

93–. 1994. Traditional/Modern in Modernised Modernisation Thinking: The development of a Weberian dichotomy. In The Theoretical Heritage from Marx and Weber in Development Studies (ed.) J. Martinussen. Occasional Papers 10, International Development Studies, Roskilde University, Roskilde.

94–.1997. “Buscar a felicidade” (To seek happiness) na Guiné-Bissau, Lala Kema, n°. 1, INEP, Bissau, 1997. A condensed version is found under the title “‘To Seek Happiness’. Development in a West-African Village in the Era of Democratisation”. Review of African Political Economy 71, 1997, 75-86.

95–.1998. Popular Sovereignty and Constitutionalism. A historical and comparative perspective. In Democratization in the Third World (eds.) L. Rudebeck & O. Törnquist, with V. Rojas. London : Macmillan.

96–.2001. On Democracy’s Sustainability. Transition in Guinea-Bissau, Sida Studies 4, Sida, Stockholm.

97–.2002a. Beyond Democratic Constitutionalism : on the twofold meaning of democracy and democratization. African Sociological Review 6 (1), 173-180.

98–.2002b. Multiparty Elections in Guinea-Bissau: viewed against late-colonial non-party and post-colonial single-party elections. In Elections in Africa (eds) M. Cowen & L. Laakso. London : James Currey.

99–.2003. Democracy as Actual Practice. In Democracy as Actual Practice: What Does Democracy Really Bring ?(ed.) L. Rudebeck. Utsikt mot utveckling 20, Collegium for Development Studies, Uppsala University, Uppsala.

100–.2004. Democratisation and ‘Civil Society’ in a West-African Village. In Decentralisation in Practice : Power, Livelihoods and Cultural Meaning in West Africa, International Workshop Highlights, Uppsala, Sweden, May 4-6, 2004, IIED (International Institute for Environment and Development), London and Department of Cultural Anthropology and Ethnology, Uppsala University, Uppsala (brochure + CD-ROM).

101–.2007. Building Civic Trust in Post-Conflict Societies. In Reconciling Winners and Losers in Post-Conflict Elections in West Africa: Political and Policy Imperative (eds) C. Norberg & C. Obi. Uppsala : Nordic Africa Institute :​publications/​books/​book.xml?id=2524

102–.2008. Demokrati, utveckling och globalisering (Democracy, development and globalisation). In Statsvetare ifrågasätter : Uppsalamiljön vid tiden för professorsskiftet den 31 mars 2008 (eds) S. Gustavsson, J. Hermansson & B. Holmström. Uppsala : Acta Universitatis Upsaliensis.

103–.2009, forthcoming. Electoral Democratisation in Post-Civil War Guinea-Bissau 1999–2006. Uppsala : Nordic Africa Institute.

104Sjögren, A. 2001. State, Civil Society and Democratisation: Theoretical Debates Past and Present. In Civil Society and Authoritarianism in the Third World (eds) B. Beckman & A. Sjögren. PODSU (Politics of Development Group), Department of Political Science, Stockholm University, Stockholm.

105Törnquist, O. 1999. Politics and Development: A critical introduction. London : Sage Publications.

106–.2009. Introduction and General Conclusions. In Aceh : The Role of Democracy for Peace and Reconstruction (eds) O. Törnquist, S. Adi Prasetyo & T. Birks. Jakarta : PCD Press and ISAI (Institute for the studies of the free flow of information).

107UN 2008. Security Council warns drug trafficking, other crimes imperil Guinea-Bissau. UN News Service, 3.10.2008.​apps/​news/​story.asp?NewsID=24349&Cr=guinea&Cr1=bissau

108World Bank 2008. Country Brief, Guinea-Bissau.


Election documentation

110Comissão Nacional de Eleições (National Election Commission), 2000a. Official record of national assembly elections held on 28 November 1999, Bissau, 9.12.1999 and 25.1.2000.

111–.2000b. Official record of presidential elections held on 16 January 2000, Bissau, 21.1.2000.

112–.2005. Official record of second round of presidential elections held on 24 July 2005, Bissau, 10.8.2005.

113Kandjadja, 1999 national assembly election. Copy of the form (white) to be completed for the first round of the parliamentary election by the local voting officials at each voting station, completed in this case by the chairman of the section committee for Kandjadja, based on the returning officers’ figures, 28 and 29.11.1999.

114–. 1999 presidential election, first round. Copy of the form (brown-yellow) to be completed for the first round of the presidential election by the local voing officials at each voting station, completed in this case by the chairman of the section committee for Kandjadja, based on the returning officers’ figures, 28.11.1999.

115–.2004 national assembly election. Figures provided 29.6.2007 by the chairman of the section committee for Kandjadja, based on his own preserved notes from 28.3.2004.

116 –24.7.2005 presidential election, second round. Figures received in Bissau 30.6.2007 by mobile phone call from the chairman of the section committee for Kandjadja consulting his notes.

117Mansabá sector, constituency 7, 1999. Detailed record, shown to me in Bissau, of the results of the national assembly election and the first round of the presidential election 28.11.1999 for the voting station in Kandjadja (no. 81), in Mansabá sector/constituency 7, Oio region. The record is hand-written, based on reports from the returning officers and drawn up during the days after election day by the PAIGC deputy (member of paliament) for Mansabá then elected,

118–.2000. Detailed record, shown to me in Bissau, of the results of the second round of the presidential election 16.1.2000 for 51 polling stations, including Kandjadja, in Mansabá sector/constituency 7, Oio region. The record is hand-written, based on reports from the returning officers and drawn up during the days after election day by the PAIGC deputy (member of paliament) for Mansabá elected 28.11.1999,

119Nô Pintcha, 8.4.2004. Local newspaper report on results of national assembly election 28.3.2004, based on data from Comissão Nacional de Eleições (National Election Commission).

120Lusa, 4.4.2004. Portuguese news agency report on results of national assembly election 28.3.2004, based on data from Comissão Nacional de Eleições (National Election Commission).



122AD, Acção Democrática/Democratic Action

123ANP, Assembleia Nacional Popular/National Assembly, i.e. parliament

124APU, Aliança Popular Unida/Popular United Alliance

125CFA, Communautée Financière Africaine

126CNE, Comissão Nacional das Eleicões/National Election Commission

127FCG-SD, Fórum Cívico Guineense - Social Democracia/Guinean Civic Forum - Social Democracy

128FDS, Frente Democrática Social/Democratic Social Front

129FLING, Frente da Libertacão Nacional da Guiné/Guinea’s National Liberation Front

130GDP, Gross Domestic Product

131LIPE, Liga Guinense de Protecção e Desenvolvimento Ecológico/Guinean League for Ecological Protection and Development

132NGO, Non-Governmental Organisation, i.e. voluntary organisation – ONG in Portuguese, French, etc.

133PAIGC, Partido Africano da Independência da Guiné e Cabo Verde/African Party for the Independence of Guinea and Cape Verde

134PDS, Partido Democrático Socialista/Democratic Socialist Party

135PRGB/MB, Partido da Resistência da Guiné-Bissau - Movimento Bâ-Fatá /Guinea-Bissau’s Resistance Party – Bâ-Fatá-Movement

136PRP,  Partido da Renovação e Progresso/Party for Renewal and Progress

137PRS, Partido da Renovação Social/Party for Social Renewal

138PSD, Partido Social Democrata/Social Democratic Party

139PU, Plataforma Unida/United Platform

140PUN, Partido da União Nacional/Party of National Union

141PUSD, Partido Unido Social Democrata/United Social Democratic Party

142RADI, Résau Africain pour le Développement Intégré/African Network For Integrated Development

143RGB/MB, the same as PRGB/MB, often merely RGB

144UE, União Eleitoral/Electoral Union

145UM, União para a Mudança/Union for Change

146UNDP, União Nacional para o Desenvolvimento e Progresso/National Union for Development and Progress

147UNDP, United Nations Development Programme

Haut de page


1  This is a new version bringing in additional material and observations, as well as being marked by the author’s continuous reflection on the issues under consideration.

2  For the conceptual framework employed, see my previous work on democracy, democratisation and development (e.g. Rudebeck 2002a ; 2003, 2008), with references. Using the term “substantial” for a non-minimalistic and two-dimensional view of democracy is by inspiration from OlleTörnquist : “Substantial democracy ‘only’ means that the conventional rules of the game... are both fair and applied in vital sectors of society...” (2002 :29).

3  The argument is in line for instance with Törnquist’s theoretical analysis of post-tsunami development and democratisation in the autonomous province of Aceh in Indonesia. Recent developments in Aceh lead Törnquist to suggest (2009 :1) : “Peace and development in Aceh is due to more, not less, democracy.”

4  The writing of this article had largely been concluded by early November 2008, at the very moment Guinea-Bissau was entering a period of acute political violence and instability. For the sake of completeness, the dramatic key events of the period until early August 2009 have been added to the background description. Empirically these events serve furthermore to support the analytical argument of the article.

5  The 2003 putsch and subsequent developments were given good coverage by Lusa.

6  For facts and analyses, see Lusa’s continuous and detailed reporting.

7  At the moment of writing these lines in early August 2009.

8  In December 2007, for instance, the Minister of Finance presented a budget in the National Assembly which added up to a deficit of 60 per cent. His comment was that the deficit “would easily be overcome with the approval of the Post-Conflict programme about to be negotiated with the International Monetary Fund and a subsequent donor conference” (Lusa, 6.12.2007).

9  In Guinea-Bissau the name of the village is usually spelt Candjadja, which is the Portuguese spelling. When transcribing from Mandinga, however, which is the predominant language in the area, as well as when writing in Guinean Creole, it is common to use the letter K for the k-sound. I therefore continue to use the K, as I have always done, although this is not the official usage.

10  The villages in the section are (with official spelling): Candjadja, Dabocunda, Corinto, Sabalacunda, Ninjobaia, Colissari, Mandina, Breco-ba, Breco-rim, Salinto, Madina Saladala, Tambato, and Djebacunda.

11  The point of departure for my population estimate is the number of persons entitled to vote and thus over 18 in the section in 1999, which was 2,209 voters, according to a local count made in 1999 (Mansabá, sector, constituency n°. 7, 2000). I have then extrapolated from the proportion over 20 + 1/10 of those under 20 years found in the whole country according to population estimates by a group of Guinean social scientists (Imbali et al. 1995). My figure is thus based on the assumption that the age distribution is the same in Kandjadja as in the whole country. If so (and if Imbali et al. are right), then the population of the thirteen villages of the section would have been 4,536 inhabitants in September 1999, which is not unlikely.

12  Just after the civil war, the large family I know best told me that, from time to time, they had had up to thirty different relatives staying with them as refugees in their simple houses. Now their supplies had been emptied, including the seed for sowing the following year.

13  I have described this in some detail, including the far-reaching effects of economic liberalisation and structural adjustment, in earlier writings (see Rudebeck 1992, with full references to even earlier writings). I have also noted the lack of any change during the 1990s (2001 :72 ff.)

14  59 per cent of the Kandjadja voters still supported PAIGC in 1994 as opposed to 38 per cent in the whole country (Rudebeck 2001 :77).

15  See Rudebeck 1992, in particular pp. 269-270, with full references to earlier works.

16  The homem grande of Kandjadja village took over more or less formally as the chairman of the section until his death in 1992. His son coninued in the same function which, however, seems to have become more and more nominal, except in connection with elections when he, as we shall see, continued to be very much in charge at the local level.

17  In the section there were five other polling stations besides the one in Kandjadja (which included Kandjadja-Balanta), namely in the villages of Mandina, Djebacunda, Corinto, Madina Saladala and Salinto.

18  Interviews on 4 and 11.12.1999. During the latter interview the chairman carefully consulted the notes he had made on election day.

19  Guinea-Bissau’s various political parties will not be presented here. See my book On Democracy’s Sustainability (Rudebeck 2001 :42-56) for a detailed such presentation. The following summarising characteristic is just as valid in 2008 as it was around 2000 : “It is not possible to distinguish between the political parties in Guinea-Bissau by studying their party programmes. The slogans of all the parties are almost identical... all the parties claim to stand for democracy, justice and human rights, as well as a market economy that is socially accountable... Any discernible differences tended to concern social and cultural/ethnic roots, financing, historical ties between leaders and supporters, and the personal qualities of the leaders” (2001 :42.) The full names of the parties are found in the list of abbreviations at the end of this article.

20  Resulting in 45 out of 100 seats for PAIGC in parliament, enough to form a national majority coalition, given the system of distributing seats between the parties applied in Guinea-Bissau (the so-called d’Hondt system) which favours the largest party.

21  See Rudebeck (1997 :38-44), on positive recognition in the village of greater civic freedom coupled with disillusionment over powerlesness in regard to development.

22  For the 1994 presidential election in Kandjadja, I only have exact data on the second round. In that round ‘Nino’ scored 90 per cent in the village as compared with 49 in the country as a whole (Rudebeck 1997 :31, 25).

23  The theoretical and theoretically inspired literature on civil society, democratisation and development is huge. This is not the place to enter into it. For a useful introduction see Törnquist 1999. For an instructive survey of the history of the concept see Sjögren 2001. For a recent comment by myself on the closely related discourse on ‘civic trust’ see Rudebeck (2007 :17-18). I point out there that trust between citizens is often assumed to reign high in ‘civil society’, which clearly could be positive for democracy – when or if it happens and, if so, depending on how. It is a significant research task to study the if, when and how of that possibility. Neither the correlation itself nor its democratising implications can be taken for granted.

24  RADI is an African NGO for culture-based development work in agriculture. The countries that are represented through their national organisations are Burkina Faso, Central African Republic, Guinea-Bissau, Guinea-Conakry, Mali, Mauritania, Senegal and Congo. (RADI, no date; Hellström 1995 :36-37). Another document (RADI 1995) provides information on RADI’s organisational structure in the Oio region. The Guinea-Bissau branch of RADI relies on funding by SWISSAID, a Swiss NGO.

25  Translated from the French language original by Lars Rudebeck.

26  On the occasion of my visit, the teacher was on summer vacation.

27  Information received in Uppsala by mobile phone call from the chairman of the section committee in Kandjadja, 26 October 2008, confirmed in another call in mid-June 2009.

28  To illustrate, Leys 1996 offers a critical analysis of such development studies, while Chabal & Daloz 1999 navigate close to it in their analyses. In an essay on ‘modernisation thinking’ found in a contribution to a volume on the theoretical heritage of Marx and Weber in development studies, I have myself attempted a critical deconstruction (Rudebeck 1994).

29  Cf.the following incisive remark by Mamdani (1996 :295 note 2) on “the mode of domination” as an explanatory variable: “My point about clientelism is that it is more an effect of the form of power than an explanation of it”.

Haut de page

Pour citer cet article

Référence papier

Lars Rudebeck, « “They Have Left Us in a Hole” », Bulletin de l'APAD, 31-32 | 2010, 65-104.

Référence électronique

Lars Rudebeck, « “They Have Left Us in a Hole” », Bulletin de l'APAD [En ligne], 31-32 | 2010, mis en ligne le 24 septembre 2010, consulté le 28 mars 2017. URL :

Haut de page


Lars Rudebeck

Professor of Political Science, Centre for Sustainable Development, Uppsala University :

Haut de page

Droits d’auteur

Bulletin de l'APAD

Haut de page
  • Logo LIT Verlag