The online platform for Taylor & Francis Group content

Cookies Notification

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more.

Advanced and citation search

Journal of Contemporary African Studies

Why the CCM won't lose: the roots of single-party dominance in Tanzania

Full text HTML PDF

Free access

DOI:: 10.1080/02589001.2012.669566

Melanie O'Gorman^a^*

pages 313-333

Publishing models and article dates explained

Received: 16 Nov 2010
Accepted: 19 Jul 2011
Published online: 26 Apr 2012

New content email alert | Citation email alert

Alert me

Abstract

This paper provides an empirical investigation of the factors contributing to single-party dominance in Tanzania. Despite the fact that Tanzania has had a multi-party democracy since 1995, the party which governed during single-party rule, the Chama Cha Mapinduzi (CCM), has won the vast majority of seats in the National Assembly in the first four multi-party elections. In order to understand the CCM's grip on power, this article analyses the results of a survey conducted amongst subsistence farmers in Tanzania, which provides information on farmers’ livelihood conditions, access to media and political views, and hence provides insight into the preferences underlying voting behaviour. It discusses the dominant reasons for CCM support, as well as the characteristics of farmers who are more likely to support an opposition party. It concludes by discussing possible policy options for enhancing political competition in Tanzania.

Keywords

View all related articles

Abstract

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

Keywords

Introduction

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

Tanzania, after years of maintaining a highly protected economy, is now in the process of economic restructuring. Former state-owned enterprises are being privatised, fiscal and financial reforms are being instituted, and trade has been liberalised. Economic liberalisation in Tanzania occurred shortly before the opening of the country's political system to multi-party democracy. Many predicted that multi-party democracy would infuse competition into the political system of Tanzania, adding an incentive for politicians to build strong bases of public support by becoming more responsive to majority interests (Luckham 1998; Kelsall 2000). This paper discusses the progress of democratisation amid economic reform in Tanzania. The particular question that it addresses is: what appears to be leading to single-party dominance in Tanzania?

Single-party dominance (SPD) may be defined in a variety of ways. In this paper I adopt the definition of Basedau (2005) to identify Tanzania as a single-party dominant state given that the Chama Cha Mapinduzi (CCM) dominates parliament and has done for a long period of time.1 Analysis of this issue should begin with an understanding of the determinants of support for the CCM among rural inhabitants. This is because rural inhabitants account for upwards of 80% of the Tanzanian population. Identifying the factors contributing to support for the CCM, or alternatively for the opposition, can in turn provide guidance on what initiatives are needed to infuse political competition and hence criticism into Tanzanian rural society, with a view to making Tanzanian government more responsive to the needs of the poorest.

In this paper I undertake a qualitative and quantitative analysis of voting behaviour in Tanzania. The qualitative analysis reveals a sense of nostalgia for the CCM, gratitude for the CCM's maintenance of peace, satisfaction with the CCM's performance during one-party rule and a sense that the CCM is the party that identifies the most with farmers’ concerns. The empirical analysis then asks: what farmer characteristics tend to lead to support for the CCM? The analysis suggests that access to newspapers or radio make a farmer more likely to support an opposition party. Involvement with a farmers’ organisation also increases the likelihood of opposition support. Finally, farmers are more likely to oppose the CCM the lower their subsistence consumption, or the higher the value of their capital equipment.

There is a large literature on the issue of SPD. For example, Kenneth F. Greene's resource theory (2007b) explains that an incumbent political party's resource advantage ensures that opposition parties address only niche concerns and are therefore uncompetitive. His book expands on this theory and discusses its applications in the context of Mexico, Malaysia, Taiwan, Japan and Italy (Greene 2007a). While this explanation most likely has relevance for the case of Tanzania, the empirical results below demonstrate that the majority of rural citizens in Tanzania are loyal to the dominant Tanzanian party, in the absence of any material benefits stemming from that loyalty.

The issue of SPD is an extremely important one for the future of African political and economic development, given its potential impact on political contestation and consequently for responsive states in the region. A number of papers discuss the phenomenon of SPD across Africa (for example, van de Walle and Buttler 1999; Randall and Svåsand 2002; Pottie 2003; van de Walle 2003). A key example is Ishiyama and Quinn (2006). They find that formerly dominant parties in Africa have tended to succeed following political liberalisation if there was political contestation during single-party rule, the lower the degree of ethno-linguistic fractionalisation and if they won the first multi-party election. They also note that the more repressive a regime was prior to multi-party democracy, the more successful the successor party is likely to be as it errected barriers to the establishment of independent interest groups (Ishiyama and Quinn 2006, 324). This is echoed by Widner (1997), who notes the weak links between political parties and civil society in Africa. The present paper will provide evidence of whether these same factors seem to be leading to the success of the CCM in Tanzania.

Further African examples include Mozaffer and Scarritt (2005), who provide evidence of SPD in Africa, finding that roughly half of votes and two-thirds of seats have been won by the largest party in the country. They conclude that the authoritarian legacy of many dominant parties and weak oppositions contribute to one party's dominance. Southall (2005) reviews the debate surrounding SPD in the case of the African National Congress (ANC) in South Africa. He notes that the ANC is a ‘broad church’, encompassing diverse interest groups, so that key policy debates occur within the party, in a similar way to the CCM. Basedau (2005) provides a systematic analysis of the possible causes for one-party dominance in sub-Saharan Africa. He finds that neither institutional variables (e.g. electoral systems, presidential regimes) nor ethnolinguistic diversity are able to explain the prevalence of one-party dominance in Africa. The authoritarian origin of dominant parties is the single variable found to have the strongest connection with contemporary single-party dominance.

Since the ‘third wave’ of democratisation hit Africa in the early 1990s, a large number of papers have been written on the broad impacts of multi-party democracy on African economies and societies (for example, Bratton and van de Walle 1997; Bratton, Mattes, and Gyimbah-Boadi 2005). However only a handful of studies has endeavoured to collect direct evidence on the preferences underlying voting behaviour (Posner 2005 for Zambia; Wood and Haines 1998 for Mozambique; Smith 1998 for Mali; Harris 1995 for South Africa; and Bratton and Mattes 2001 for Zimbabwe, Nigeria, Namibia, Malawi, Ghana and Botswana). This paper contributes to this literature by providing primary evidence on the roots of one-party dominance in Tanzania.

This paper is also related to a number of papers analysing African democracy more generally. Bogaards (2004) identifies dominant party systems in Africa, but does not investigate why they exist or their implications. This paper is similar to Kuenzi and Lambright (2001) in that it highlights the lack of party system institutionalisation in Tanzania, while Kuenzi and Lambright highlight this more generally for 30 African countries. Lindberg (2006) analyses the impact of opposition behaviour on a number of variables thought to indicate SPD, finding that opposition participation in the electoral process does indeed improve the quality of that process across a large number of elections and polls. Finally, although not dealing with African democracies, Seligson and Tucker (2005) address a similar question to this paper, in asking why people vote for former authoritarian rulers in free elections. They find that after taking into account individuals’ opinions about the economy and socio-demographic characteristics, citizens that voted for former authoritarians cited a preference for authoritarian forms of government over democracy.

The broader goal of this paper is to explore the puzzle of rural neglect in Tanzania. In Tanzania, rural neglect manifests itself in a relative bias towards urban areas in terms of public services (for example, in education and health care facilities and access to clean water (National Bureau of Statistics 1991/92 and 2000/01), in urban–rural disparity of infrastructure and in a relative lack of budgetary support for the agricultural sector. Such neglect is a puzzle, because rural inhabitants form the numerical majority in Tanzania, and Tanzania has had a multi-party democracy with strong voter participation in all areas since 1995. Furthermore, support for the CCM is strong, despite policies which continue to prioritise sectors other than agriculture. This paper identifies the reasons behind rural support for the CCM amid rural neglect, and therefore sheds light on factors contributing to a lack of responsiveness to the rural sector on the part of the Tanzanian government.

The next section provides a brief overview of political and economic liberalisation in Tanzania. The qualitative results of the survey are discussed in the third section, as are the hypotheses to be tested. Section four then presents the results of the empirical analysis. The final section concludes and discusses policy implications stemming from the analysis.

Political and economic liberalisation in Tanzania

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

Tanzania began reforming its economy in 1986 from a largely state-controlled economy to one more open to market forces. Reforms affecting the rural sector directly were the dismantling of parastatals, price deregulation, an end to pan-territorial pricing, currency devaluation, the removal of input and food price subsidies, and the liberalisation of trade. As has been the case in many transition economies, agricultural production per person dropped sharply in Tanzania following economic liberalisation. However, per capita production has not rebounded since that time (see Figure 1).

Figure 1. Index of agricultural production per capita in Tanzania (1999 = 100).

PowerPoint slide Original jpg (51.00KB)Display full size

Agriculture's role in the Tanzanian economy is large in many respects. For example, in 2005 agriculture contributed roughly half of the economy's aggregate output and employed approximately 75% of the country's population (World Bank 2006; United Republic of Tanzania 2005). Despite the importance of the sector, budgetary support for agriculture has been very low (see ). This is despite the fact that the Tanzanian government, along with other African governments, committed in 2005 to spend at least 10% of its budget on agriculture through the Comprehensive Africa Agriculture Development Programme (CAADP). Perhaps even more worrying is the declining share of total government expenditure invested in agricultural research and development which is crucial to expanding crop yields. Such expenditure fell from 25 to 30% in the early 1990s to under 1% from 2007 to 2009 (World Bank 2000; United Republic of Tanzania 2006).

Table 1. Government expenditure on support for agriculture (% of total expenditure).

CSV PDF Display Table

provides a summary of input use and the progress of labour productivity in Tanzanian agriculture. This table highlights that fertiliser and machinery use declined through the 1990s, while agricultural land use remained constant. A number of researchers have found that such inputs are important in contributing to agricultural productivity (for example, Lau and Yotopoulos 1989; Hu and Antle 1993; Craig, Pardey, and Roseboom 1997), indicating that declining input use in the 1990s in Tanzania may be responsible for declining labour productivity (shown in column 5).

Table 2. Growth of input use and productivity in Tanzanian agriculture (%).

CSV PDF Display Table

Growth rates of the major food crop yields in Tanzania between 1960 and 2000 are provided in . This table also reveals a sharp downturn in productivity in the 1990s for the very staples that Tanzanian households depend on for subsistence (Evenson and Gollin 2001). Whereas early claims suggested that poor agricultural performance was the result of a non-existent supply response on the part of farmers to price incentives, two more recent papers (Danielson 2002; Thiele 2003) suggest that farmers in Tanzania have in fact responded rationally to price incentives. However, structural features of the Tanzanian economy have muted this supply response. Such features include poor infrastructure and low population densities, implying thin markets (see for example, Winter-Nelson and Temu 2002, 2005).

Table 3. Crop yields in Tanzania (%).

CSV PDF Display Table

Many studies suggest a negative impact of these economic reforms on rural Tanzanians. Kelsall (2003) notes that the Tanzanian government now sees its role as confined to provision of market information and monitoring performance, or rather solely as a facilitator or regulator of the market. This has implied a reduction in credit provision to smallholder farmers, the removal of input subsidies, the withdrawal of input supply from remote areas and reduced support for agricultural research (Ponte 2001, 2002). Prior to economic liberalisation, the Tanzanian economy was regulated to a much larger extent relative to those of other developing countries (Havnevik 1993).2 Given the range of services the Tanzanian government provided in rural areas (marketing, storage, crop transport, crop finance), the vacuum provided by state withdrawal of such services is most likely responsible for the poor performance of the agricultural sector over recent decades (see for example Kherallah et al. 2000).

As noted above, moves towards a more market-oriented economy began in 1986 upon Tanzania's entrance into a structural adjustment programme with the Bretton Woods institutions. The process of political liberalisation began roughly five years after the beginning of economic liberalisation in Tanzania. In 1992, under the advice of ex-President Julius Nyerere, Tanzania amended its constitution to allow the formation of a multi-party democracy. A Political Parties Act was passed in June 1992, repealing the single-party clause in the Tanzanian constitution. The past four multi-party elections have confirmed the dominance of the Chama Cha Mapinduzi, the party that has ruled Tanzania since independence (as illustrated in ). Prior to the October 2010 election, only the Civic United Front (CUF) posed any substantive threat to the CCM's success, and the CUF's support tends to be clustered around Zanzibar. However Chama cha Demokrasia na Maendeleo (CHADEMA) (which in English stands for ‘Party for Democracy and Development’) managed to secure 13% of the seats in Parliament in the last election. The CHADEMA leader, Willibrod Slaa, also secured 27% of the vote in the presidential election, despite the fact that it was his first time running for president.3 Other parties have been unable to attract a significant proportion of votes.

Table 4. Multi-party election results.

CSV PDF Display Table

The continuing strength of the CCM is puzzling. One would expect to see scant support for a party which appears to neglect the vast majority of its citizens, unless of course subsistence farmers do not vote. However, in the 2000 and 2005 elections voter turnout was upwards of 70% in most rural areas (National Electoral Commission of Tanzania 2000; TEMCO 2006). The dominance of the CCM therefore remains a puzzle. This paper provides insight into the phenomenon of single-party dominance in Tanzania through the direct analysis of political opinions among Tanzanian subsistence farmers.

Qualitative analysis and hypotheses

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

The data used to investigate the issue of single-party dominance is from a survey conducted in rural Tanzania in the summer of 2008. This survey was conducted with assistance from the Sokoine University of Agriculture. A total of 286 farmers were interviewed from seven regions – Mtwara (south-east), Iringa (central), Kilimanjaro (north-east), Tanga (east), Morogoro (central), Mwanza (north-west) and Shinyanga (north). The locations of the survey sites are noted in Figure 2.

Figure 2. Location of the survey sites.

Source: http://www.mapsofworld.com/tanzania/tanzania-political-map.html.

PowerPoint slide Original jpg (234.00KB)Display full size

The purpose of this survey was to learn about voting behaviour in Tanzania among rural farmers, and hence farmers were asked whether they voted in the last election, which party they supported if they did vote, and why. The survey also asked for information on agricultural production, educational attainment, family size, access to media, group membership and attitudes toward economic reform.4 The share of support attributed to political parties in the dataset is given in .

Table 5. Vote shares in the sample (taken June/July 2008).

CSV PDF Display Table

Farmers surveyed were asked to give reasons for their support of their favoured political party. For those farmers who indicated that they supported the CCM, the main reasons provided are given in . The top reasons for CCM support are: they performed well during one-party rule, they are the only strong party, they have maintained peace, they identify with farmers and they brought us independence. The first two of these responses are in line with the reasons put forward by Ishiyama and Quinn (2006) for why formerly dominant parties often remain successful after the initiation of multi-party competition. Often such parties are successful because the opposition is weak and fragmented. This was echoed by farmers in our survey who stated that no other party was strong enough to rule. Many farmers offered that they were not even aware of parties other than the CCM.

Table 6. Reasons for CCM support – proportion of respondents citing each reason.

CSV PDF Display Table

The latter may indicate that CCM dominance stems from the cost of acquiring information on the opposition. As the CCM is very well-known, learning about opposition parties entails a cost for the average Tanzanian. This cost may involve time or money used to access media or attend school. The first hypothesis that is tested below is that the likelihood of CCM support is lower the higher a farmer's educational attainment. It may be that students acquire information on the political system and parties in school, or that their literacy has enabled them to access media to a greater extent relative to those who lack formal education.

The second hypothesis also relates to the cost of acquring information on political alternatives – that access to media lowers the cost of acquiring such information and therefore makes it more likely that a farmer will vote for an opposition party. This accords with the views of Lansner (Ndulo) that the ‘ills of a country with a free and open media will always be more apparent to their own peoples, and to the wider world – recognition essential to the possibility of their amelioration’ (Ndulo 2006, 199). These latter hypotheses are tested by including in the model an educational attainment variable, variables indicating whether a farmer is literate or numerate and a variable quantifying the frequency with which farmers access the internet, read newspapers, watch television or listen to the radio.

The weakness of opposition parties in Tanzania may also relate to the absence of ready-made social bases on which opposition parties may capitalise. Once the political system opened up to competition in Tanzania, there was no significant trades union, co-operative movement or human rights movement with a history of independence from the state upon which opposition parties could build. Further, with many different ethnic groups in Tanzania, each representing a small proportion of the total, the opposition is unable to attract a large number of votes by appealing to any one group (Kelsall 2003).

Indeed, the Tanzania Social Capital and Poverty Survey (Ellis and Mdoe 2003) revealed a striking absence of rural political mobilisation in Tanzania. Further, before 1995, political expression outside the CCM was stifled and facilitated by an ‘all-class’ ideology. A third hypothesis that is tested in the next section, then, is that the likelihood of support for the CCM is lower the more involved a farmer is with community organisations. This hypothesis asserts that the cost of obtaining information may be lowered through involvement in community organisations, and that with more information on opposition parties, a voter is less likely to support the CCM. An alternative way of framing this hypothesis is that involvement in community organisations sensitises a farmer to the extent of rural neglect, or strengthens their sense of relative deprivation as they realise that the neglect they experience is a general one. This proposition is tested using a variable in the dataset which asked farmers whether they were a member of a religious, labour, business or farmers’ organisation.

Two other dominant responses noted in relate to a sense of nostaglia with regard to one-party rule, in that many farmers still support the CCM due to its role in the independence struggle. The Tanganyika African National Union (TANU) was the predecessor of the CCM, and it was the principal political party involved in the struggle for Tanzania's independence. Kelsall (2003) suggests that this struggle lent a sense of legitimacy to the CCM, and the survey responses validate this claim. The association of the CCM and peace is a reflection of the efforts of the CCM to create a sense of nationhood after independence. There was a conscious effort to promote Kiswahili as a national language, as well as efforts to promote a national culture through the creation of state-affiliated organisations such as the Youth League, choirs, dance groups etc.

The fourth hypothesis tested below concerns farmers’ ideological support or affection for the CCM: the older a farmer, the more likely they are to support the CCM. This hypothesis asserts that older farmers are more likely to support the CCM as they have been influenced for longer by the CCM's monopoly during one-party rule. As noted above, under single-party rule the CCM absorbed many forms of potential opposition into the organs of the state, such as trades unions, women's groups, youth groups etc. Hypothesis 4 asserts that the longer a farme has lived under such a stifling regime, the less likely they are to be critical of the CCM.5

The most common response in is that the CCM performed well during one-party rule. A number of authors of voter choice have found that retrospective economic evaluations influence voter choice (see for example Ndegwa 2001; Duch and Stevenson 2008; Seligson and Tucker 2005). In the Tanzanian context, farmers’ retrospective economic assessment (potentially overall and of their own circumstances) is of a time when the state provided a great deal of support for agriculture (input subsidies, marketing and transport services, credit etc.) and welfare services. Further, many farmers placed the CCM closest to their own needs, similar to the earlier findings of Sivalon (1999).6

The fifth hypothesis tested below, then, relates to the economic benefits derived from CCM rule. That is, that the higher a farmer's non-farming income, the more likely it is that they will support the CCM. This hypothesis tests the proposition that farmers who are diversifying their income sources are more likely to support the CCM's market-driven, non-farm policy platform. In other words, the more a farmer has been able to diversify their income sources, the more enthusiastic they are about the CCM's platform. There is a large literature suggesting that such livelihood diversification is becoming increasingly common in Africa (see for example Ellis 1998; Bryceson 2002; Barrett, Reardon, and Webb 2001). This hypothesis will be tested using the coefficient on non-farm income of farmers in the sample.

The sixth hypothesis tested below also relates to the economic benefits derived from CCM rule. That is, the more social assistance or production/input subsidies a farmer has received over the past year, the higher the likelihood that they support the CCM. This is a straightforward hypothesis – if the CCM is ‘buying votes’ from farmers, they will most likely be successful in elections. This hypothesis coincides with the literature on political parties in Africa which suggests that multi-party democracy in Africa has been a continuation of patronage politics of single-party rule (Bratton and van de Walle 1997; van de Walle 2003). This hypothesis will be tested using the coefficient estimate on the variable ‘subsidies’, which asked farmers whether they had received subsidies for inputs or production, and on the variable ‘social assistance’, which asked farmers how much monetary support they had received from the government in the past year.

The reasons cited for support of an opposition party are illustrated in . Reasons given were mainly based on practical rather than ideological grounds. For example, half of those supporting an opposition party cited a reason related to rural neglect – that the opposition party they support was concerned with poverty, development or farmers’ concerns. Other reasons put forward suggest that there is confidence in the abilities of opposition parties – farmers thought such parties are strong, have good policies and could provide a stable economy. Hence, although support for the opposition was found to be quite weak among the farmers surveyed, the farmers that did support an opposition party appeared to be reacting to rural neglect, and felt an opposition party would be able to deal with such concerns.

Table 7. Reasons for opposition support – proportion of respondents citing each reason.

CSV PDF Display Table

Quantitative analysis

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

In order to model voter choice in Tanzania, I use a random utility model, as described in Liao (1994). This model represents a voter's support for the dominant political party as a function of the costs and benefits of having that party elected. The variable being explained by this model is thus whether a farmer supports the CCM or not. While this ignores differences within the opposition, the question of interest for this paper is what causes the strength of the CCM to persist, rather than the reasons behind support for particular opposition parties per se. Variables which capture the perceived costs and benefits of CCM rule are then proxied by explanatory variables ranging from access to media to a farmer's educational attainment. These variables are described below and the model is discussed in more detail in the Appendix.

The estimation results are presented in and . Instead of examining the coefficient estimates of the regression, it is more intuitive to consider the effect of a marginal change in an explanatory variable on the dependent variable, the probability of voting for the CCM. Such marginal effects are computed for each explanatory variable with all other variables held at their mean values. A chi-squared test indicates that the model estimated above is significantly different from the intercept-only model and hence passes the basic goodness of fit test. To test for multicollinearity between the independent variables, variance inflation factors (VIF) were calculated for each variable. Each of the variables included in the regression had a VIF below 2, well below the value of 10 which Chatterjee and Price (1991) suggest may be cause for concern about multicollinearity.

Table 8. Probit estimation.

CSV PDF Display Table

Column 2 of presents the results of the estimation including only basic socio-demographic variables – age, sex, marital status, family size, educational attainment and region. These are referred to as the control variables, and none is significant in determining CCM support. Thus there is no support for hypothesis 4 given that age has no significant impact on the likelihood of support for the CCM. This is despite the fact that there was a good deal of variation in age across the sample, and suggests that, despite the emphasis on length of time reported in the qualitative analysis, the CCM's grip on power extends to young as well as old farmers.

In column 3 of I include as explanatory variables in the model dummy variables indicating whether a farmer is literate or numerate. Neither is significant. This casts doubt on hypothesis 2 that educational attainment has an impact on support for the CCM. This result is largely because most farmers in the sample have exactly seven years of schooling – completion of primary schooling only.7 Given that educational attainment, literacy and numeracy are highly correlated, I include only educational attainment in the regressions that follow.8

The results of the estimation of further models are provided in . These models include the control variables mentioned above, however the marginal effects of such variables are not reported in order to conserve space. Column 2 of presents the results of the estimation after including media variables. Hypothesis 2 regarding access to media is accepted for the case of reading the newspaper, given that this variable reduces the likelihood of support for the CCM. However, frequency of accessing the internet, listening to the radio and watching television are all insignificant in terms of determining support for the CCM. Very few farmers (only 10 out of the sample) had access to the internet and very few understood what the internet was. Evidence of Kalathil and Boas’ (2003) conjecture that access to the internet makes voters more critical therefore cannot be found in this sample. Although not statistically significant, a farmer's frequency of watching television and listening to the radio is negatively correlated with support for the CCM.

Table 9. Probit estimation – control variables plus additional explanatory variables

CSV PDF Display Table

Column 3 presents the results of the estimation after including media variables as well as group membership variables (all futher models include variables from the last column). Newspaper reading remains significant despite the inclusion of these variables. Involvement with a religious organisation makes it more likely that a farmer will support the CCM. This accords with the earlier findings of Sivalon (1999) and suggests that farmers may view allegiance or loyalty to the CCM as a moral imperative. However farmers that are involved in farmer organisations have a lower probability of supporting the CCM relative to farmers that are not involved. This suggests that a critical view of the CCM may be developed once farmers interact with one another.

The results of the estimation after including non-farm income sources is reported in column 4. Newspaper access and membership in a farmers’ and religious organisation remain significant. However hypotheses 5 and 6 are unfounded in the data. That is, neither variation in financial assistance provided by the government, subsidies received nor variation in the level of non-farming income affects the likelihood of supporting the CCM. This is because very few farmers in the sample earned incomes outside of farming or had received financial assistance from the government – only 10% of farmers had received social assistance (which accounted for an average of 15% of their total income) and only 14% had received subsidies (which accounted for on average 50% of their total income). This contradicts the earlier suggestion of Sivalon (1999), at least for subsistence farmers, that the CCM is viewed as the party of the nation because it provides patronage. Rather, the subsistence farmers interviewed supported the CCM for reasons other than patronage. The fact that the variable measuring non-farming income is insignificant also reinforces the view of these farmers as undemanding of the CCM, in that these farmers are not even benefiting from the liberalisation drive of the CCM. The latter coincides with the findings of Ellis and Mdoe (2003).

In column 5 I include variables measuring the value of agricultural production sold on the market and the value of subsistence agricultural production. Market agricultural income is not significant in explaining support for the CCM. However, the negative and statistically significant coefficient on subsistence production suggests that farmers with relatively low subsistence production are less likely to support the CCM. This suggests that farmers who are worse off in terms of their livelihood conditions do tend to question the CCM's authority.

Finally in column 6, the result of the estimation including all of the above-mentioned variables plus a dummy variable indicating whether the farmer had access to irrigation, the number of livestock they owned and the value of capital equipment is included. Interestingly, with the inclusion of such variables, radio access becomes statistically significant in explaining support for an opposition party, while neswpaper access remains significant. Other significant variables include being a member of a farmers’ or business organisation, the value of subsistence consumption and the value of capital equipment. That is, farmers with access to a radio or who regularly read the newspaper, who are members of a farmers’ organisation, whose subsistence consumption is low or whose value of capital equipment is high are more likely to support an opposition party.

The result that farmers with capital equipment are more likely to vote for an opposition party is somewhat surprising. The ability to purchase capital equipment is most likely related to a farmer's access to credit. Very few smallholder farmers in Tanzania have access to credit (Kherallah et al. 2000), as state-run banks in the rural areas have been closed and microfinance has not yet filled the gap left by such banks. Hence farmers who have been forced to borrow from moneylenders or who have acquired credit or capital from NGOs may resent the lack of support provided by the government in this regard, turning their support to an opposition party.

indicates that farmers who are members of a business organisation are more likely to support the CCM. It could be that export-oriented farmers or those hoping to start a non-farm business are in favour of the CCM's platform, which aims to reduce the barriers and costs of doing business and to liberalise trade.9 Overall, the variables that are robust in explaining political affiliation of subsistence farmers are newspaper and radio access, farmers’ organisation membership and the value of subsistence consumption and capital equipment.

Conclusion

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

In this paper I presented the results of an empirical analysis using a survey recently conducted in Tanzania, in order to gauge the reasons behind political choice among rural inhabitants in Tanzania. This analysis in turn allowed me to uncover some of the barriers to the articulation of majority concerns in Tanzania, and therefore to shed light on the potential of multi-party democracy to reflect such concerns in national policy. The results of this empirical analysis suggest that while many rural Tanzanians are indeed shouldering the burden of economic reforms, such a burden is not straining their loyalty to the party enacting the reforms. A lack of a strong alternative among the political parties seems to be a prime reason for this loyalty, as does a sense of nostalgia for the party which brought them independence and which has maintained relative peace. A natural retort to this observation is that it will simply take time for such nostalgia to fade and for a pro-rural challenge to the CCM to emerge. However, age appears to have no significant effect on CCM support – rural Tanzanians old and young are loyal to the CCM.

Given the strong support for the CCM among subsistence farmers generally, it appears that the CCM neglects such farmers because support among them is strong, despite rural neglect. This is because, as demonstrated in the qualitative results section above, their support is based on ideological factors – an inherent loyalty or nostalgia for a party that helped to bring independence to the country, which has maintained peace, and which historically supported farmers. If rural neglect is being unchallenged, it is likely to continue. This has dire implications for agricultural productivity, poverty and aggregate economic growth, as the largest segement of the Tanzanian economy and society will continue to struggle to build human capital, access important social services such as health care and water and sanitation, and as farmers face input prices that are simply out of reach. An interesting result of the analysis is that farmers who are worse off in terms of their subsistence production do tend to question the CCM's authority. Hence in terms of waiting for subsistence farmers to demand more from the CCM, things may get worse before they get better.

A number of other African former single parties are dominant to a similar extent to the dominance of the CCM in Tanzania. This is the case for the Democratic Party of Equatorial Guinea (PDGE), Gabonese Democratic Party (PDG), Front for the Liberation of Mozambique (FRELIMO) and the South West Africa People's Organization (SWAPO) of Namibia. Each of these parties has continually won upwards of three-quarters of seats in the legislature since opening to multi-party competition.10 A number of features contributing to SPD in Tanzania (noted above) are relevant for dominance in these countries as well – for example, the stifling of open debate and repression of dissent during single-party rule, the benefits of incumbency and consequent disorganisation of the opposition, nostalgia for the party which brought independence and weakness of civil society. The latter three factors were highlighted in the analysis above as important contributors of support for the CCM.

However, a number of features of Tanzanian democracy are unique, and hence entail differing impacts on political competition in Tanzania. For one thing, whereas a dominant ethnic group in the case of Namibia most likely strengthens SPD, ethnic diversity may pulverise the ethnic spectrum (Erdmann and Basedau 2008) and therefore reduce party dominance. However the post-independence efforts of Julius Nyerere to create a monolithic Tanzanian culture has entailed that Tanzanian politics is not based simply on tribalism (Hyden 2000). This probably exacerbates SPD given the large number (127) of ethnic groups in the country which could potentially divide the political spectrum were ethnicity a salient political issue.

Second, van de Walle (2003) and Magaloni and Krucheli (2010) point to the clientilistic nature of politics in many African societies, and the importance of patronage to political success. In such systems, citizens (in particular the poor) have a vested interest in maintaining the former single party because of the patronage that the party provides. In other words, whereas in many African countries (Gabon being the obvious case), patronage entails that citizens are better off maintaining their loyalty to the dominant party rather than taking a risk with an oppositon party, the analysis above suggests that in Tanzania this is not the case. Basedau (2005) also finds that of the very dominant systems mentioned above, Tanzania is the only country that has achieved its dominance without a combination of favourable economic, social and governance indicators.

Finally, despite policies that discriminated against farmers during single-party rule, Tanzania had a ‘relative rural bias’ (Ellis 1984), distinct from the ‘urban bias’ that characterises most other developing countries. As noted in the qualitative analysis above, this seems to have strengthened support for the CCM. However given that citizens will often vote based on evaluations of past economic benefits, the urban bias that persisted during single-party rule in most African countries most likely decreases support for the former single party, all else held equal. Hence Tanzania is also distinct from many other African countries due to its ‘sincere egalitarian rhetoric’ (Lipton 1977) which probably contributes to the dominance of the CCM.

A number of policy implications stem from the empirical results reported above. The significant effect of farmers’ organisations on opposition support suggests that greater financial support for such organisations would improve political competition in Tanzania. Although rural organisation is weak in Tanzania, an important farmer's organisation, Mtandao wa Vikundi vya Wakulima ya Tanzania (MVIWATA), is developing.11 As van de Walle (2002, 79) suggests more generally for Africa, ‘the fate of democracy may hinge on whether civil society groups, opposition parties and ordinary citizens can match wits and strengths with resourceful rulers who aim to keep power and are not all that particular about how they do so’. Greater support for independent media, to ensure open media's commercial survival also appears important. In this regard, existing initiatives such as USAID's Centre for Democracy and Governance, the Institute for International Communication and Development (IICD) or the World Bank's Information for Development initiative seem crucial.

Note on contributor

Jump to section

Introduction
Political and economic liberalisation in Tanzania
Qualitative analysis and hypotheses
Quantitative analysis
Conclusion
Note on contributor

Melanie O'Gorman is an Associate Professor in the Department of Economics at the University of Winnipeg in Canada. Her research is in the areas of economic development, political economy and macroeconomics. She has recently published articles on the determinants of agricultural productivity growth in the developing world and education policy for reducing racial earnings inequality in the US and South Africa. She may be reached at m.ogorman@uwinnipeg.ca

Appendix A

A random utility model is a type of index function model where the outcome of a discrete choice (in this case voting) is observed but the underlying utility leading to the discrete choice are unobserved. Let voter i's valuation of a party P's election be denoted . The valuation is modelled as depending on the economic benefits voter i receives from the election of the party net of the costs of acquiring information on that party and on ideological factors:

where B represents the relative economic benefits that voter i receives from the election of party P, C represent the costs of acquiring information on party P, X represents non-economic or ideological benefits that voter i receives from party P, and the error term ϵ _P is symmetrically distributed with zero mean and its cumulative distribution function (CDF) is denoted F. This formulation reflects the fact that voters in Tanzania are concerned with their own economic well-being, but also with non-pecuniary benefits of political parties.

The observed choice between parties reveals which provides the greatest utility, the CCM or an opposition party. In practice, is unobserved; rather, what is observed is a dummy variable V. Taking V=1 as a vote for the CCM, and abbreviating the opposition parties with superscript OPP, then:

The probability that a farmer supports the CCM is then:

It is assumed that the error term above follows the standard normal distribution and inferences are therefore made regarding voter choice in Tanzania using a binary probit analysis.

Notes

1. Given the focus of this paper, whereas Basedau (2005) requires that parties in question have won the last two legislative elections, I require that the party ruled during single party rule and has won all legislative elections since the opening of the political system to multiple parties.

2. For example, data compiled by La Porta et al. (1999) indicates that Tanzania was 10th out of 105 countries in terms of the extent of state ownership in the economy over the period 1975–1995.

3. CHADEMA's support is clustered in the urban constituencies of Dar es Salaam, Mwanza and Arusha. The party ran on a platform of bringing about market-based solutions to poverty and reducing corruption. Recent demonstrations by CHADEMA supporters have caused President Kikwete to warn that CHADEMA ‘will plunge the country into unnecessary conflict’ (Citizen 2011). As in the case of South Africa, the recent decrease in CCM support in urban areas may indicate that the ‘liberation dividend’ (Southall 2005) is waning.

4. The Afrobarometer surveys (available at http://www.afrobarometer.org/index.html) provide a wealth of information on public attitudes towards democracy and economic reform in a number of African countries. My survey differs from these in that it samples only subsistence farmers and provides greater detail on their livelihood conditions as well as the reasons they give for their political affiliations. Such household-level data is necessary for identifying characteristics that drive voting behaviour.

5. Kitschelt (1995) suggests that the suppression of civil society and limits on the extent to which citizens could criticise the government is indeed a legacy of single party rule. He suggests that in post-communist democracies, incumbents have been successful given their ability to entrench themselves in different aspects of civil society and therefore repress opposition.

6. Sivalon (1999) found that farmers actually feared multiparty democracy because of a sense that the opposition parties (mainly centred in Dar es Salaam) did not have their interests at heart.

7. 62% of farmers in the sample had seven years of schooling, 6% had more than seven years of schooling and the remaining 32% had less than seven years of schooling.

8. The correlation coefficient between numeracy and literacy is 0.9, between numeracy and educational attainment is 0.72 and between literacy and educational attainment is 0.75.

9. The Ease of Doing Business index of the World Bank indicates literally no change over the past few years in the ease of doing business in Tanzania, with the exception of the ease of trading across borders (http://www.doingbusiness.org/data/exploreeconomies/tanzania?topic=dealing-with-licenses). Hence farmers who are members of a business organisation and with an export orientation may be reacting to the CCM's platform of trade liberalisation.

10. The proportion of seats won by these parties in the last election was 89% for the PDGE, 77% for the PDG, 79% for FRELIMO and 75% for SWAPO. The Botswana Democratic Party (BDP) is also a very dominant political party on the continent, however it was never a one party state.

11. MVIWATA is a national network of local-level farmer's organisations with the purpose of representing and advocating on behalf of farmers. Its membership is still small, with only 600,000 current individual members, but it has grown each year since its inception in 1991. Its membership base consists of individuals whose livelihood depends on agriculture, and it is representative of the agricultural Tanzanian population as a whole (Wennink, Nederlof, and Heemskerk 2007).

References

1. Barrett, C., Reardon, T. and Webb, P. 2001. Nonfarm income diversification and household livelihood strategies in rural Africa: Concepts, dynamics, and policy implications. Food Policy, 26(4): 315–31. [CrossRef], [Web of Science ®]
2. Basedau , M. 2005 September . A preliminary inquiry into the causes of one-party dominance in contemporary sub-Saharan Africa . Paper prepared for the Panel ‘Dominant Parties and Democracy’ at the bi-annual European Consortium of Political Research Conference in Budapest/Hungary .
3. Bogaards, M. 2004. Counting parties and identifying dominant party systems in Africa. European Journal of Political Research, 43: 173–97. [CrossRef], [Web of Science ®]
4. Bratton, M. and Mattes, R. 2001. Africans’ surprising universalism. Journal of Democracy, 12(1): 107–21. [CrossRef], [Web of Science ®], [CSA]
5. Bratton, M. and van de Walle, N. 1997. Democratic experiments in Africa: Regime transitions in comparative perspective, New York: Cambridge University Press. [CrossRef]
6. Bratton, M., Mattes, R. and Gyimah-Boadi, E. 2005. Public opinion, democracy and market reform in Africa, New York: Cambridge University Press.
7. Bryceson, D.F. 2002. The scramble in Africa: Reorienting rural livelihood. World Development, 30(5): 725–39. [CrossRef], [Web of Science ®], [CSA]
8. Chatterjee, S. and Price, B. 1991. Regression analysis by example, Newark, NJ: John Wiley and Sons.
9. Citizen . 2011 February 28 . Chadema demos threaten our peace . The Citizen ,
10. Craig, B.J., Pardey, P.G. and Roseboom, J. 1997. International productivity patterns: Accounting for input quality, infrastructure, and research. American Journal of Agricultural Economics, 79(4): 1064–76. [CrossRef], [Web of Science ®]
11. Danielson, A. 2002. Agricultural supply response in Tanzania: Has adjustment really worked?. African Development Review/Revue Africaine de Developpement, 14(1): 98–112. [CrossRef], [Web of Science ®], [CSA]
12. Duch, R.M. and Stevenson, R.T. 2008. The economic vote: How political and economic institutions condition election results. Political economy of institutions and decisions, Cambridge: Cambridge University Press. [CrossRef]
13. Ellis, F. 1984. Relative agricultural prices and the urban bias model: A comparative analysis of Tanzania and Fiji. Journal of Development Studies, 20(3): 28–51. [Taylor & Francis Online], [Web of Science ®], [CSA]
14. Ellis, F. 1998. Household strategies and rural livelihood diversification. Journal of Development Studies, 35(1): 1–38. [Taylor & Francis Online], [Web of Science ®], [CSA]
15. Ellis, F. and Mdoe, N. 2003. Livelihoods and rural poverty reduction in Tanzania. World Development, 31(8): 1367–84. [CrossRef], [Web of Science ®]
16. Erdmann, G. and Basedau, M. 2008. Party systems in Africa: Problems of categorising and explaining party systems. Journal of Contemporary African Studies, 26(3): 241–58. [Taylor & Francis Online]
17. Evenson , R. , and D. Gollin . 2001 . The green revolution at the end of the twentieth century . Technical Advisory Committee Secretariat , Food and Agriculture Organization of the United Nations .
18. Food and Agriculture Organization (FAO) . 2007 . FAOSTAT data , http://www.fao.org .
19. Greene, K.F. 2007a. Why dominant parties lose – Mexico's democratization in comparative perspective, Cambridge: Cambridge University Press. [CrossRef]
20. Greene , K.F. 2007b . A resource theory of single-party dominance . Comparative Politics Workshop Department of Political Science, Yale University .
21. Havnevik, K.J. 1993. Tanzania: The limits to development from above, Uppsala: Nordic Africa Institute.
22. Harris, M. 1995. Attitudes and behaviour of the ‘new’ South African electorate: An empirical assessment. International Social Science Journal, 47: 567–82. [CSA]
23. Hu, F. and Antle, J.M. 1993. Agricultural policy and productivity: International evidence. Review of Agricultural Economics, 15(3): 495–505. [CrossRef]
24. Hyden, G. 1999. Top-down democratization in Tanzania. Journal of Democracy, 10(4): 142–55. [CrossRef], [Web of Science ®], [CSA]
25. Ishiyama, J. and Quinn, J.J. 2006. African Phoenix? explaining the electoral performance of the formerly dominant parties in Africa. Party Politics, 12(3): 317–40. [CrossRef], [Web of Science ®]
26. Kalathil, S. and Boas, T.C. 2003. Open networks, closed regimes: The impact of the internet on authoritarian rule, Washington, DC: Carnegie Endowment for International Peace.
27. Kelsall, T. 2000. Governance, local politics and districtization in Tanzania: The 1998 Arumeru tax revolt. African Affairs, 99(397): 533–51. [CrossRef], [Web of Science ®], [CSA]
28. Kelsall, T. 2003. Governance, democracy and recent political struggles in mainland Tanzania. Commonwealth and Comparative Politics, 41(2): 55–82. [Taylor & Francis Online], [CSA]
29. Kherallah, M., Delgado, C.L., Gabre-Madhin, E.Z., Minot, N. and Johnson, M. 2000. The road half traveled: Agricultural market reform in sub-Saharan Africa, Washington, DC: International Food Policy Research Institute (IFPRI).
30. Kitschelt, H. 1995. Formation of party cleavages in post-communist democracies – Theoretical propositions. Party Politics, 1(4): 447–72. [CrossRef], [Web of Science ®], [CSA]
31. Kuenzi, M. and Lambright, G. 2001. Party system institutionalization in 30 African countries. Party Politics, 7(4): 437–68. [CrossRef], [Web of Science ®], [CSA]
32. La Porta, R., Lopez-deSilanes, F., Shleifer, A. and Vishny, R. 1999. The quality of government. Journal of Law, Economics and Organization, 15(1): 222–79. [CrossRef], [Web of Science ®]
33. Lau, L.J. and Yotopoulos, P.A. 1989. The meta-production function approach to technological change in world agriculture. Journal of Development Economics, 31: 241–69. [CrossRef], [Web of Science ®]
34. Liao, T.F. 1994. Interpreting probability models: Logit, probit, and other generalized models, Thousand Oaks, CA: Sage.
35. Lindberg, S.I. 2006. Opposition parties and democratisation in sub-Saharan Africa. Journal of Contemporary African Studies, 24(1): 123–36. [Taylor & Francis Online]
36. Lipton, M. 1977. Why poor people stay poor: Urban bias in world development, Cambridge, MA: Harvard University Press.
37. Luckham, R. 1998. Popular versus liberal democracy in Nicaragua and Tanzania?. Democratization, 5(3): 92–126. [Taylor & Francis Online], [CSA]
38. Magaloni, B. and Krucheli, R. 2010. Political order and one-party rule. Annual Review of Political Science, 13: 123–43. [CrossRef], [Web of Science ®]
39. Mozaffar, S. and Scarritt, J.R. 2005. The puzzle of African party systems. Party Politics, 11(4): 399–421. [CrossRef], [Web of Science ®]
40. National Bureau of Statistics . 1991–92 and 2000–01. Tanzania Household Budget Survey. http://www.tanzania.go.tz/hbs/HomePage_HBS.html .
41. National Electoral Commission of Tanzania . 2000 . General Election 2000: Parliamentary election statistical data . Dar Es Salaam : National Electoral Commission of Tanzania .
42. Ndegwa , S.N. 2001 . A decade of democracy in Africa . International Studies in Sociology and Social Anthropology , 81 . Leiden : Brill .
43. Ndulo, M. 2006. Democratic reform in Africa – Its impact on governance and poverty alleviation, Athens, OH: Ohio University Press.
44. Ponte, S. 2001. Policy reforms, market failures and inputs use in African smallholder agriculture. European Journal of Development Research, 13(1): 1–29. [Taylor & Francis Online], [CSA]
45. Ponte, S. 2002. Farmers and markets in Tanzania: How market reforms affect rural livelihoods in Africa, Oxford, Portsmouth: James Currey, Heinemann and Mkuki na Nyota. NH and Dar es Salaam
46. Posner, D. 2005. Institutions and ethnic politics in Africa, New York: Cambridge University Press. [CrossRef]
47. Pottie, D. 2003. Party finance and the politics of money in Southern Africa. Journal of Contemporary African Studies, 21(1): 5–26. [Taylor & Francis Online]
48. Randall, V. and Svåsand, L. 2002. Introduction: The contribution of parties to democracy and democratic consolidation. Democratization, 9(3): 1–10. [Taylor & Francis Online], [CSA]
49. Seligson, A.L. and Tucker, J.A. 2005. Feeding the hand that bit you: Voting for ex-authoritarian rulers in Russia and Bolivia. Demokratizatsiya, 13(1): 11–42. [CrossRef]
50. Sivalon, J. 1999. “Rural Tanzanians and the national election of 1995”. In Agrarian Economy, State and Society in Contemporary Tanzania, Edited by: Forster, P.G. and Maghimbi, S. Brookfield: Ashgate Publishers Ltd.
51. Smith , Z.K. 1998 . Building African democracy: The role of civil society-based groups in strengthening Malian civic community . PhD diss. , The University of North Carolina at Chapel Hill .
52. Southall, R. 2005. The ‘dominant party debate’ and the South African party system. AfricaSpectrum, 40(1): 61–82.
53. Tanzania Election Monitoring Committee (TEMCO) . 2006 . The 2005 elections in Tanzania Mainland . http://www.tz.undp.org/ESP/docs/Observer_Reports/2005/TEMCO_Report_Mainland_2005.pdf .
54. Thiele, R. 2003. Price incentives, non-price factors and agricultural production in sub-saharan Africa: A cointegration analysis. African Development Review, 15(2–3): 425–38. [CrossRef], [Web of Science ®]
55. United Republic of Tanzania . 2005 . Economic survey . http://www.tanzania.go.tz/economicsurveyf.html .
56. United Republic of Tanzania . 2006 . Economic survey . http://www.tanzania.go.tz/economicsurveyf.html .
57. Van de Walle, N. 2002. Africa's range of regimes. Journal of Democracy, 13(2): 66–80. [CrossRef], [Web of Science ®], [CSA]
58. Van de Walle, N. 2003. Presidentialism and clientelism in Africa's emerging party systems. The Journal of Modern African Studies, 41(2): 297–321. [CrossRef], [Web of Science ®], [CSA]
59. Van de Walle, N. and Smiddy Buttler, K. 1999. Political parties and party systems in Africa's illiberal democracies. Cambridge Review of International Affairs, 8(1): 14–28. [Taylor & Francis Online]
60. Wennink , B. , S. Nederlof , and W. Heemskerk 2007 . Access of the poor to agricultural services: The role of farmers’ organizations in social inclusion . Amsterdam : Koninklijk Instituut Voor de Tropen Publishers .
61. Widner, J. 1997. “Political parties and civil societies in sub-Saharan Africa”. In Democracy in Africa: The hard road ahead, Edited by: Ottaway, M. Boulder, CO: Lynne Reinner.
62. Winter-Nelson, A. and Temu, A. 2002. Institutional adjustment and transaction costs: Product and inputs markets in the Tanzanian coffee system. World Development, 30(4): 561–74. [CrossRef], [Web of Science ®], [CSA]
63. Winter-Nelson, A. and Temu, A. 2005. Impacts of prices and transactions costs on input usage in a liberalising economy: Evidence from Tanzanian coffee growers. Agricultural Economics, 33(3): 243–53. [CrossRef], [Web of Science ®]
64. Wood, G. and Haines, R. 1998. Tentative steps towards multi-partyism in Mozambique. Party Politics, 4(1): 107–18. [CrossRef], [Web of Science ®], [CSA]
65. World Bank . 2000 . Agriculture in Tanzania since 1986: Follower or leader of growth? Washington : A World Bank Country Study .
66. World Bank . 2006 . World Development Indicators . http://databank.worldbank.org . [CrossRef]

Table 1. Government expenditure on support for agriculture (% of total expenditure).
1995	1999	2000	2001	2002	2003	2004	2005	2006	2007	2008
3.6	3.9	3.0	2.8	3	3.5	3.1	3.3	3.3	2.5	4.3
Source: Economic Survey, Government of Tanzania.

Table 2. Growth of input use and productivity in Tanzanian agriculture (%).
Time period	Fertiliser per hectare	Machinery per worker	Agricultural land use	Labour productivity
1960–1970	30	−2	2	−2
1970–1980	8	−5	1	4
1980–1990	3	−5	1	1
1990–2000	−3	−1	0	−2
*Fertiliser per hectare includes inorganic fertiliser use per hectare of agricultural land, while machinery includes tractors, harvesters, threshers and milking machines. For the labour productivity statistics, agricultural output was taken as the price-weighted sum of the quantity of each agricultural commodity produced in a given country in a given year, minus quantitites of output used as seed and feed
Source: FAO (2004).

Table 3. Crop yields in Tanzania (%).
Time period	Rice	Wheat	Maize	Sorghum	Millet	Cassava
1960–1970	−7	−1	−8	−6	1	−1
1970–1980	4	5	16	2	0	12
1980–1990	6	4	2	8	5	2
1990–2000	−2	−4	0	−1	−2	−3
Source: Evenson and Gollin (2001).

Table 4. Multi-party election results.
Percentage of seats in the National Assembly election
Party	1995	2000	2005	2010
CCM	79	87	85	75
NCCR – Mageuzi	7	0.3	0	1
CHADEMA	1	2	3	13
CUF	10	8	10	10
UDP	1	1	0.3	0.3
TLP	0	2	0.3	0.3
APPT	N/A	N/A	N/A	0

Percentage of votes in the presidential election
Party (candidate)	1995	2000	2005	2010
CCM	62 (Mkapa)	72 (Mkapa)	80 (Kikwete)	63 (Kikwete)
NCCR-Mageuzi	28 (Mrema)	N/A	0.5 (Mvungi)	0.3 (Rungwe)
CUF	6 (Lipumba)	16 (Lipumba)	12 (Lipumba)	8.3 (Lipumba)
UDP	4 (Cheyo)	4 (Cheyo)	N/A	0.2 (Dovutwa)
TLP	N/A	7.8 (Mrema)	0.7 (Mrema)	0.2 (Mgaywa)
CHADEMA	N/A	N/A	6 (Mbowe)	27 (Slaa)
APPT	N/A	N/A	N/A	1.2 (Kuga)
Note: NCCR stands for National Convention for Construction and Reform, CHADEMA stands for Chama cha Demokrasia na Maendeleo, CUF stands for Civic United Front, UDP stands for United Democratic Party, TLP stands for Tanzania Labour Party and APPT for African Progressive Party.

Table 5. Vote shares in the sample (taken June/July 2008).
Party	Percentage
CCM	84.62
CUF	4.2
CHADEMA	2.10
TLP	3.85
NCCR	1.02

Table 6. Reasons for CCM support – proportion of respondents citing each reason.
Reasons for CCM support	%
Performed well during one-party rule	24.1
They are the only strong party	18.6
They are the party that identifies with farmers	7.6
They are the party that identifies with the business community	2.5
They are progressive	4.2
They are conservative	0.4
They identify with my ethnic group	1.3
They will maintain a stable economy	4.2
They have maintained peace	18.2
They brought us independence	8.4
They are the only party I've known	4.6
I love the CCM	3.8
They have brought development to our community	1.3
I support them but plan to change soon	0.8

Table 7. Reasons for opposition support – proportion of respondents citing each reason.
Reasons for opposition support	%
They have the best policies	12.5
They are a strong party	15.6
They are the party that best identifies with farmers	6.3
They identify with my ethnic group	3.1
They are the most progressive party	9.4
They can provide a stable economy	6.3
They are concerned with the development of our community	9.4
They are concerned with poverty	34.4
Change is needed	3.1

Table 8. Probit estimation.
Dependent variable – dummy variable indicating a vote for the CCM
Marginal effects	Controls	Controls, numeracy and literacy
Region	−0.007	−0.007
	(0.010)	(0.010)
Age	0.002	0.002
	(0.002)	(0.002)
Sex	0.039	0.048
	(0.048)	(0.048)
Family size	−0.007	−0.005
	(0.008)	(0.008)
Educational attainment	0.004	−0.003
	(0.009)	(0.012)
Marital status	−0.062	−0.057
	(0.056)	(0.058)
Numeracy		0.211
		(0.163)
Literacy		−0.122
		(0.073)
N	277	277
Adjusted R²	0.020	0.025

Table 9. Probit estimation – control variables plus additional explanatory variables
Dependent variable – dummy variable indicating a vote for the CCM
Marginal effects	Access to media variables	Group membership variables	Non-farm income	Farm income	Agricultural variables
Internet	0.043	0.0155	0.015	0.003	0.006
	(0.054)	(0.048)	(0.046)	(0.031)	(0.020)
Television	−0.010	−0.011	−0.014	−0.014	−0.011
	(0.015)	(0.015)	(0.014)	(0.011)	(0.009)
Radio	−0.046	−0.036	−0.035	−0.094	−0.100
	(0.026)	(0.026)	(0.025)	(0.042)	(0.030)*
Newspaper	−0.048	−0.030	−0.029	−0.021	−0.017
	(0.016)***	(0.016)*	(0.015)*	(0.015)*	(0.013)**
Religious group		0.055	0.054	0.023	0.015
		(0.021)***	(0.020)***	(0.018)	(0.014)
Farmers’ group		−0.085	−0.081	−0.044	−0.032
		(0.021)***	(0.021)***	(0.025)***	(0.022)***
Business group		0.027	0.027	0.029	0.027
		(0.029)	(0.028)	(0.023)	(0.020)*
Labour group		0.032	0.023	0.040	0.033
		(0.038)	(0.038)	(0.033)	(0.026)
Non-farm income			1.54E-07	6.78E-08	4.14E-08
			(1.69 E-07)	(1.24 E-07)	(8.40 E-08)
Subsidies			2.20E-06	6.09E-07	3.37E-07
			(2.10E-06)	(1.36E-06)	(8.95E-07)
Social
Assistance			1.53E-07	−4.67E-07	7.93E-08
			(1.09E-06)	(8.38E-07)	(5.87E-07)
Subsistence
Production				−8.45E-10	−8.85E-10
				(4.31E-10)**	(5.15E-10)***
Market revenue				1.06E-08	7.89E-09
				(9.06E-08)	(7.11E-09)
Irrigation					0.042
					(0.042)
Livestock					0.001
					(0.001)
Capital					−3.70E-07
					(2.63E-07)*
N	279	273	273	215	213
Adjusted R²	0.091	0.172	0.181	0.238	0.272

Article metrics

Views: 222
Citations
Scopus (1)

Article metrics information

Librarians

Librarians' area
Pricing
Institutional account
Access entitlements
Co-branding
IP ranges
Link resolver preferences
Usage reports

Cookies Notification

Journal of Contemporary African Studies

Volume 30, Issue 2, 2012

Why the CCM won't lose: the roots of single-party dominance in Tanzania

Abstract

Keywords

Related articles

Permalink

Abstract

Jump to section

Keywords

Introduction

Jump to section

Political and economic liberalisation in Tanzania

Jump to section

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 1. Government expenditure on support for agriculture (% of total expenditure).

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 2. Growth of input use and productivity in Tanzanian agriculture (%).

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 3. Crop yields in Tanzania (%).

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 4. Multi-party election results.

Qualitative analysis and hypotheses

Jump to section

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 5. Vote shares in the sample (taken June/July 2008).

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 6. Reasons for CCM support – proportion of respondents citing each reason.

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 7. Reasons for opposition support – proportion of respondents citing each reason.

Quantitative analysis

Jump to section

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 8. Probit estimation.

Why the CCM won't lose: the roots of single-party dominance in Tanzania

DOI:

Published online:

Table 9. Probit estimation – control variables plus additional explanatory variables

Conclusion

Jump to section

Note on contributor

Jump to section

Appendix A

Notes

References

Table 1. Government expenditure on support for agriculture (% of total expenditure).

Table 2. Growth of input use and productivity in Tanzanian agriculture (%).

Table 3. Crop yields in Tanzania (%).

Table 4. Multi-party election results.

Table 5. Vote shares in the sample (taken June/July 2008).

Table 6. Reasons for CCM support – proportion of respondents citing each reason.

Table 7. Reasons for opposition support – proportion of respondents citing each reason.

Table 8. Probit estimation.

Table 9. Probit estimation – control variables plus additional explanatory variables

Article metrics

Users also read

Librarians