4. The Impact of Devolution of Powers to Different Levels
4. The Impact of Devolution of Powers to Different Levels
4.1. Expenditure patterns and priorities
The 1991 HDR highlighted the excellent record of the Government of Zimbabwe in channelling public resources towards areas of priority for human development. It is difficult, however, to make an assessment of relative priorities at the different levels of government. This is partly due to the limited time period under investigation, partly to inadequate published data available, and partly to the counterfactual problem - since the decentralizing reforms have been adopted by a new black majority government. It is likely that to a considerable extent the shifting priorities are to be attributed to the human development orientation of the new national government, rather than to the impact of decentralization. Decentralization may have been a mechanism for improving performance in human development rather than a cause. Only tentative conclusions can be drawn here.
Central government priorities
An examination of the expenditure priorities at the central government level over the past decade reveals a consistently impressive record. Table 13, based on the 1990 IMF GFS, shows that the social allocation ratio for central government expenditure averaged 34%, and the social priority ratio 56%. Figures for the most recent available year [which year?] are 45%, and 53% respectively. This compares very favourably to other developing countries, where the available evidence (for a sample of 25 countries) shows a weighted average, a social allocation ratio of 28% and a social priority ratio of 38% (HDR 1991).
These achievements can be seen in the context of substantial restructuring of public expenditures after independence. In particular, combined spending on defence and administration was cut back, from 44% in 1979 of the national budget to 27% in 1988. Hence the social allocation ratio rose by over 15 percentage points during the 1980s. Similarly, within the social sectors, expenditures have shifted towards the more basic levels of provision, to serve the majority black population. Thus, for example, the share of primary schools in the education budget rose from 38 to 56%, and real expenditure per pupil doubled.
The sectoral allocation of central government expenditures at the provincial level provides another indicator of central government priorities, although the data excludes those CG expenditures which are non-allocable regionally or located in the capital (e.g. much of administration and defence). Data for four provinces (Table 14) shows that infrastructure is accorded first priority (32.5%), followed by urban and economic services and housing (25.3%). Social development ranks third - although there is a wide range of allocations from 44% down to 7%.
Local priorities
Indications of local priorities can be derived from actual local expenditure patterns by the District Councils, and from development planning documents at district, provincial, ward and village levels. Both have some limitations as accurate indications of local priorities. The problem with looking at expenditure patterns is the overwhelming dependency of District Councils on central grants which are tied to specific purposes. For example, until 1988 the very high proportion of District Council expenditure directed towards education was largely due to the central grants for teachers' salaries. Also, the district councils spend very little on capital investment, as much of the investment in infrastructure (particularly on roads and water supplies) is centrally financed through the District Development Fund (see Table 6 above). The exercise is further complicated by the extreme fragmentation of local accounts, ranging from secondary school construction, to the development levy, to poultry (e.g. Gokwe D.C. has eighteen separate accounts - four just for education!). This further limits the ability of councils to vary priorities in expenditure decisions. 13 Nonetheless Table 15 presents a breakdown of expenditure for selected district councils. From our own small sample it can be seen that the social allocation ratio is of the order of 33%. The earlier figures for 1985/6 show a much higher average at 93.1%; this is basically due to the District Councils' former role in education, which has since been rescinded. Because of the strong constraints on LG expenditure, it may be rather misleading to accept these local expenditure patterns as an expression of local priorities.
Analysis of the relationship between "own" finance as a proportion of total finance of the District Council and the share of social expenditure in total DC expenditure, in sixteen Councils, shows a negative correlation both for 1981/2 and for 1984/514. This is undoubtedly in part due to the fact that a large portion of central government transfers are specifically for health and education.
Development plans may be more accurate indications of local priorities, although of course these are partly determined by the centre, since requests are shaped by what the localities think the centre would be prepared to fund. It has also been seen that the local bodies often produce an unprioritised "shopping list" - thus the fact that the plans are formulated without any firm idea of the available resources should be borne in mind. With these caveats, we proceed to examine local priorities as indicated by the planning exercises.
Detailed examination of a sample of development plans of District Councils reveals that social development comprises, on average, 38.7% of total bids - ranging from a low of under 17% in Marondera, to almost 48% in Kwekwe15. Next in importance is agriculture, with about 30% of the total bid - again with a big range across individual districts. Infrastructure ranks third (at 21%), and housing and employment (13.5) is at the bottom (Table 16). It should be noted, however, that the District Annual Plans are drawn up by the DDC, a committee composed predominantly of sector ministry officials.
Analysis of Provincial Development Plans in three provinces shows a more even distribution of emphasis across the four categories - with variations among the provinces (Table 17). Social Development fares worst in two provinces, and best in one. Agriculture and water receives first priority on average with 28% of the total bid. Here again, it is noted that provincial priorities are formulated by the PDC, which can be seen as an organ of the central government.
At the village level, foremost importance was accorded to water and sanitation, according to a recent report (Table 18). For the country as a whole social welfare and food handouts came next, although there were significant variations across the provinces, from 1% in Mashonaland East and West, to 23% in Matabeleland South. Roads and transport were then identified, again with some variation across provinces.
Limited evidence on Ward priorities is indicated by the efforts being undertaken in Mberenga District. This shows a strong preference towards infrastructure, especially water supply, and measures to support agricultural activities (Figure 3).
Evidence from the special cases of CAMPFIRE and the Midlands show the impact of an increase in untied resources at the local level. In the Midlands, the increment currently represents approximately 20% of the total Council budget. The selected priorities of Gokwe District Council16 show about one-quarter of the allocation for economic activities (including dams), 36% for the social sectors and one quarter for water for domestic use (table 19).
In the CAMPFIRE experiment, in effect WADCOs and VIDCOs are allowed to decide how best to spend substantial sums of money. The distribution of wildlife revenue in Beitbridge District Council shows a preference for household cash (half the revenue in one ward), followed by a productive project (42%) and then education (8%).
Conclusions as to relative priorities
To a considerable extent, government choices at each level between and within sectors are determined by the allocation of functions to them. District Council functions are limited to the provision of basic levels of social and economic services, viz. primary and secondary education, clinics and maternity care, sanitation and refuse collection (Table 4). This means that the priority ratio within each sector's expenditure (i.e. the share going to basic services) is very high (approximately one) at the DC level while it is much lower at the central government level, since a substantial share of the central health budget is spent on the large city hospitals and advanced education. Thus the central government social priority ratio stood at 53% in 1989 - although this represents a much better result than many other developing countries, still it is far below that obtained by the District Councils.
The evidence on priorities at different levels of government suggests that more devolution of decision making power would focus public expenditure on infrastructure (especially water supplies) and priority social services (Table 20). In particular, it is clear that local decision makers accord greater emphasis to economic services and agriculture than the centre has shown to date. Indeed the economic allocation ratio in the plans of District Councils is at least twenty percentage points higher than that of the centre. Although the centre scores well on the social allocation ratio, greater devolution of decision making power and financial authority to district councils would still broadly double the proportion of expenditure going to basic levels of provision in the social sectors.
These conclusions have emerged from an examination of the preferences of District Councils, as revealed by the development planning exercise. To date, however, these expressed local priorities have had only a limited impact on overall patterns of public expenditure. District Councils lack their own financial resources, and the sectoral ministries maintain close control, via tied grants, over government expenditures spent locally. In other words, the effective public expenditure ratio over which the District Councils exercise discretion is very small. It follows that the strategy of increasing expenditure on social and economic priorities in Zimbabwe requires a more thoroughgoing and effective devolution of decision making power to the District Councils.
4.2.1. Context
The serious disparities which characterize contemporary Zimbabwean society have already been stressed. The most striking aspect of the inequalities in Zimbabwe today are between the small white and majority black population. The former generally continue to enjoy very high standards of living in both the towns and the country, whilst the Communal Areas still suffer from land pressure and degradation, less adequate social and economic infrastructure, strong dependence on the commercial core economy in terms of employment and services, and limited productive investment opportunities.
Whilst the most dramatic gaps exist between the whites and the Africans there are also considerable inter- and intra-district inequalities within the Communal Areas. Recent work by Jackson and Collier (1991), based on a survey of 600 household in Communal Lands across the five agro-ecological zones, showed a skewed distribution of incomes, wherein the top 10% of households control 42% of measured income, the lower 50% only 15%, and the bottom 25% less than 5% of income.
Estimates of provincial per capita incomes are show large disparities among the eight provinces. Most striking, at the top end, stands Mashonaland East (which includes Harare) with more than double the national average income per head, and Matabeleland East, with about one-and-a-half times the average (table 21). Indeed, these two provinces raise the mean to such an extent that all of the remaining provinces fall below it, the very poorest being Masvingo and Matabeland South. Note that the figures here are provincial averages, which conceal disparities within each province. Unfortunately no estimates of district per capita incomes are available. 17
Evidence on a provincial basis is also available with respect to other indicators of human development - see Table 21. Overall, no clear picture emerges from the various measures of access to services. Taking the indicators shown in Table 21, three Provinces come in the top half of the ranking in more than half of the indicators - Manicaland, Mashona East and Midlands - while three come in the bottom half in more than half the indicators - Mashonaland Central, Mashonaland West and Masvingo.
Decentralization and Equity
Here we are interested in the extent to which decentralization of expenditure has been equalizing or unequalising, across provinces and districts. Data deficiencies, however, prevent any rigorous conclusions. In very broad terms, policy since 1980 has resulted in a significant redirection of public expenditure towards priority services in the long impoverished Communal Areas. Agriculture has experienced more widespread provision of credit, extension services and marketing services. There has been a very large expansion of basic education opportunities, involving the re/construction of approximately 2000 primary and 1200 secondary schools in the Communal Areas since 1980, increases of 86% and 738% respectively. Rural primary health care related programs expanded dramatically between 1982-88 - e.g. the number of Village Health Workers increased 26-fold between 1981 and 1987, to over 8 200. And there has been a considerable effort to extend the provision of safe water in the Communal Lands - coverage has risen from 33% in 1980 to 55% in 1990. Thus during the post-independence period the relative disadvantage of rural areas has diminished. For example, for 1980-84, the numbers of doctors and hospital beds did not keep pace with urban population rise, but there was a significant improvement in the rural areas (Tax Commission Report, p 271).
In general, this expansion in services was associated with the policy of deconcentration, rather than devolution.
Provincial distribution of expenditure
The Public Sector Investment Program, disaggregated on a provincial basis, excluding large scale projects with national externalities, favours Matabeleland North and to a lesser extent, Mashonaland North (tables 22 and 23). In simple terms, it appears that public resources are being directed toward those provinces which already have much the highest per capita income. This is confirmed by regression analysis showing a positive relationship between central government public expenditure per capita, and provincial income per capita (Table 24).
Evidence on the per capita distribution of public financial resources in the health sector (table 25), shows that the two provinces where the four central hospitals are located receive 2-5 times more than the other six provinces. Even when the figures are adjusted to take into account out-of-province patients referred to Harare and Bulawayo, per capita expenditures there are still 1.4 - 2.7 times higher than elsewhere. Comparing disparities in health expenditure to the provincial pattern of child malnutrition, as indicated by the percentage of children (24-59 months) who are underweight, it appears that public health expenditures are not being directed to those areas in greatest need. Mashonaland East (with Harare) has by far the highest health expenditures per capita, but not as high a level of need as indicated by Matabelelands North and South, or Mashonaland Central, where health expenditure is much lower, relative to need. Regression analysis reveals no significant relationship between 'need', as measured by malnutrition or by the ratio of people to clinics, and per capita expenditure on the health services, while there is a significant positive relationship between health expenditure per head and per capita income by province, though not as strong as for PSIP as a whole (Table 24).
Government expenditure per head on water supplies per province shows a negative relationship with some indicators of need. Thus the expenditure is greater the existing access to good water; it is also negatively related to population per borehole (though the statistical significance is low). However, it is slightly inversely related with per capita incomes (in contrast to other expenditures) and with the percentage of households with flush toilets, but in each case the results are not statistically significant.
The limited data on provincial distribution of government expenditure thus indicates that this expenditure has not been progressive, but has shown a positive relationship with average incomes; in neither health nor water has distribution been in accordance with indicators of need. The data relates to central government allocations, and does not indicate the distributional effects of devolution, nor of CG grants to the local authorities.
District evidence:
The Tax Commission Report (1987) noted an evident bias of per capita central government transfers in favour of urban councils, relative to district rural councils, for 1981/2 (Table 26).
In the (non-democratic) rural councils there is evidence of an inegalitarian distribution of social services. In Rural Council areas, the health status of, and services for, farm workers, who are not permitted to vote in local elections, are very poor. One study found that 83% of children on farms had never been immunised, 75% of families had never attended a clinic (Loewenson 1983); this compares unfavourably to the situation for Zimbabwe as a whole, with 75% of one-year old children immunised, and 62% of the rural population reported as having access to health care (HDR 1991).
The criteria used for CG grants among districts suggest some would be progressive and some regressive. Recurrent sectoral allocations for education and health are based upon the number of pupils or clinics in the district, which tends to favour those districts which are better endowed with such facilities. The grant from the Ministry of Local Government for administrative expenses, however, is a flat lump sum amount - which apparently does not distinguish between the administrative needs of Councils of varying population size. Maintenance of roads and boreholes, undertaken by the DDF, is carried out on the basis of technical considerations. Thus the distributional incidence of central transfers to meet current expenses does not indicate any marked progressive intent beyond a basic orientation toward rural areas.
It seems likely that capital grants are more redistributive, according to the criteria adopted for these allocations, which aim to achieve an even spatial distribution of schools and clinics. Nor is there evidence of ethnic bias in the location of rural health centres (Herbst 1990). DDF roads allocations are likely to be redistributive in impact, given that the first priority is to construct a network of primary roads throughout all the districts, before proceeding to secondary roads.
As to own council finances, there are clear differences in the amounts raised by District Councils, related to differing resource endowments. Helmsing (1991) found a significant coefficient of variation in per capita local revenue, of 0.42 (excluding school fees), with the variation had been diminishing over time.
Evidence of distributive effects among DCs
Data on distribution of CG transfers and own revenues for sixteen DCs correlated with various measures of deprivation (Table 27) show that both CG grants and LG revenue were inversely related to need (as measured in 1982) both for 1981/2 and 1984/5. From this perspective there was little difference between CG and LG funds. The one exception was the DDF for 1981/2, which was equalizing, but this had been reversed for 1984/5. While most CG transfers were unequalising in this sense, they were more "equitable" between DCs than LG finance (i.e. more equally distributed among DCs), and the equity of their distribution in this sense increased over time, but because of social inequalities between DCs this meant they accentuated social inequalities. Apart from possible political considerations, there is little point in equalizing revenues per capita among districts, given the initial ties between them.
Unfortunately the data does not permit a detailed analysis of the comparative incidence of taxation at the central and local levels. But it seems likely that central government levied taxes are less regressive than local fees and duties levied by district councils18. The Tax Commission (1986) concluded that central government taxes, for 1982/3, were progressive at the top end of the income distribution, and probably mildly regressive among the lower 70% of incomes. Many elements of local revenue are likely to be regressive. The highly regressive development levy constitutes about 5% of own local revenue. School fees form a very substantial proportion - 60% - of local revenue. Primary school fees are likely to be regressive, while secondary fees may be progressive. Business licences may also be progressive.
Ideally this study would measure the impact of the decentralizing reforms on economic activity. Unfortunately a comprehensive assessment is impossible, given the serious lack of data with respect to economic participation, particularly on a regional basis. So instead we are forced to turn to more scattered indications which bear on these issues. Decentralization has had both beneficial and negative effects on economic participation. On the positive side, we can explore the impact of increased public expenditures in the formerly neglected Communal Lands, through the impact of the "growth points" policy and the effects on agricultural growth. On the negative side, there is the possibility that the planning legislation and controls implemented by local authorities have inhibited economic participation, particularly of disadvantaged groups in the community.
4.3.1. Growth point policy
"... rural service centres and district rural service centres ... form the linchpin of our strategy ... to redress the imbalance between the rich and poor areas and bring about a better life for all."
(Zvobgo, then Minister of Local Government and Town Planning, 1982 - as quoted in Gasper 1989).
At the outset, the lack of widespread economic participation, which was part of the Colonial legacy should be emphasised. The settlers had apportioned the country such that the blacks were left with the worst arable lands, which were subject to increasing population pressure. The Communal Lands were regarded by the settlers simply as labour reserves, which had a "controlled and subordinate role within the dualistic system" (Gasper 1989). The Communal Lands had long been underdeveloped, with almost 60% of the country's people, they generated less than 10% of its income. Off-farm employment and small businesses were relatively unimportant. Manufacturing production in Zimbabwe was highly protected against foreign and informal domestic competition, and concentrated in Harare and Bulawayo (table 28). Thus the prime objectives of the newly-elected black majority government were to overcome the inherited dualism in both agriculture and industry.
A primary instrument for spreading economic participation was the "growth points" policy which aimed to promote growth in specific centres, as a result of an injection of physical infrastructure (Gasper 1986, p.2). Although only a limited number (18) centres were officially designated, the term "growth point" has come to be applied to other centres lower down in the hierarchy of physical planning (see Table 29). Here, following Gasper (1989), we will look at growth points as popularly understood, i.e. to include District Service, and Rural Service, Centres. All have been the sites of major government spending since 1980. The centres were envisaged to become focuses of service provision (education, health, local administration, as well as commercial services, such as marketing, supply of consumer goods and production inputs), thereby triggering local economic growth.
The main focus was on the District Service Centres, which each received a standard package of road improvements, water and sewerage reticulation, new market stands, toilets and bus stands. Gasper (1987) describes these new physical centres as "the visible sign and channel for the restoration and transformation of the Communal Areas, and their rapid transition from labour dustbins into the active mainstream of modern national development" (p. 13). The growth point and District Service Centre infrastructure was built by the Central Government, with some facilities financed by DDF. The DCs are not involved, and thus the approach may be viewed as an example of deconcentration rather than devolution.
The great hope placed in the growth points approach was followed by disappointment from the mid-1980s (Gasper 1989). But the basis for assessment is rather limited; there are only a handful of studies and only a short period of time has elapsed. Nonetheless the following observations may be made:
(i) There has been a much needed expansion of service provision, both public and private. The former ranges across the social sectors, to marketing depots for grain and cotton. Large increases in marketed agricultural output were recorded in the Communal Areas. The provision of marketing facilities is evidently regarded as very useful, according to surveys carried out in Sanyati and Buhera (Gasper 1987). There has been significant private investment in wholesale and retail outlets at many centres; the greatest growth sector was liquor retail outlets (Gasper 1989; Wekwete 1987).
(ii) Expansion of non-farm production of goods has been limited. The main industrial development has been in building materials, which can be traced to the demands created by public sector construction (Wekwete 1987). Overall the traditional areas of repair work and grain-milling are still most prominent (Gasper 1989). Local off-farm employment is relatively insignificant, affecting only about one in ten households in Communal Areas, and accounting for 15% of all measured income (Jackson and Collier 1991).
Between 1965 and 1982 the trend was towards increasing centralisation in manufacturing. Harare alone accounted for 49% of manufacturing output in 1965, and 42% of manufacturing employment (1967); shares which rose to 52% and 46% by 1982. The output share was 53% in the late 1980s. This can be seen in the context of highly centralised decision making (e.g. procurement procedures) and allocation procedures (as prevailed for foreign exchange) which has characterised economic management in Zimbabwe to date. This encourages central business location, especially of large firms, in order to facilitate dealings with central government officials.
(iii) Public employment in the District Service Centres is a significant local source of income (table 30). This would be expected to have multiplier effects expanding local market demand. Helmsing (1986) concluded that the single most important effect of the development of DSCs on rural industries is of a demand-pull nature resulting from public sector wages. The public service is also an important source of future rural entrepreneurs. One survey of District Service Centres found that 45% of local entrepreneurs were former public servants (Gottlicher 1984).
The overall picture in the late 1980s was of significant increases in mainly commercial activity at a few District Service Centres at Growth Points, but no notable expansion at other centres. The net impact on rural employment, the government conceded in 1985, was "insignificant" (Minister of Finance Dr Chidzero, quoted in Gasper 1989). The predominant official interpretation of these developments has been that fiscal incentives are required to support physical infrastructure - some improved tax incentives were included in the 1985 budget. However such measures would tend to be more effective in directing the location of large enterprises. The centralising tendency of the network of government controls (on foreign exchange, investment etc.) may also have inhibited rural activities.
4.3.2. Agricultural growth
At independence, the Communal Lands averaged only 6% of marketed agricultural sales. Zimbabwe experienced significant increases in agricultural output since then, particularly in the Communal Lands which now contribute approximately half of the total. This was due to a range of measures, including the extension of physical infrastructure (roads, marketing facilities and irrigation), and services (e.g. through Agritex and the Agricultural Finance Corporation), and the elimination of discrimination against peasant farmers by parastatals. It is not possible to attribute the gains to decentralization as such. There have been a mixture of organisational forms involved in the improved infrastructure and extension - deconcentrated sectoral ministries taking the lead, but with interaction with NGOs and local government.
Extension services in the Communal Lands involved various government agencies, parastatals and NGOs. Agricultural extension services are organised under the Ministry of Agriculture, through AGRITEX. Agricultural Extension Officers play a critical role; on average there is one Officer per 850 farmers, 19 with a high level of interaction between the Village Community Workers and AGRITEX (MC&CD 1991).
Growth of agriculture has been rapid, but unbalanced spatially, being concentrated in the areas with the best agro-ecological conditions (i.e. Mashonaland and Manicaland - see World Bank Agricultural Sector Memorandum, May 1991, pp 24-39).
4.3.3. Local regulations and by-laws
The legal and regulatory environment has had a substantial impact on patterns of economic participation in Zimbabwe. Whilst some principal regulatory instruments are nationally enacted and implemented (e.g. the Companies Act), local authorities are empowered to regulate economic activities within their jurisdiction. This control is exerted in two basic ways, both of which have inhibited economic activity. It can be seen as an important factor behind the relatively low share of the non-farm informal sector activities in the Zimbabwean labour force (of the order of only 3% of the labour force: 1986 Labour Force Survey).
First there is the regulation of goods and services, through the health and building standards required for production and conditions of sale. The by-laws vary among councils. Masvingo, for example, has particularly strict rules pertaining to the health standards of premises where food is sold. These regulations have been criticised as stifling economic participation, particularly on the part of those who are disadvantaged in terms of capital, resources and access to credit (World Bank 1991). The extensive requirements are often beyond the capacity of, or a serious obstacle to, informal sector entrepreneurship. This was the finding of a survey of the problems faced by women in the informal sector. In Murehwa, for example, the district council prohibits hawkers from operating within 5kms. of Murehwa Business Centre, thereby separating them from the main source of potential customers.
Second, there are controls over land use exercised by district councils under 1976 legislation which are designed to ensure "orderly" development. 20 Most councils, other than Harare, have considerable land under their jurisdiction. They have the power, for example, to set aside land for markets and issue annual permits upon payment of a fee. Market stalls are said to be in short supply, especially in Harare, which forces vendors to illegal street hawking (World Bank 1991, p.39). In addition, each local jurisdiction has imposed conditions on the purchase of land for business purposes, effectively protecting existing (often non-African) shop owners against new competition. For example, the Murehwa District Council requires a minimum capital of Z$15 000 before it will lease land to an applicant. Moreover, entrepreneurs in the Communal Lands still face the colonial system of leases. Land tenure in the small centres fall under the Communal Lands Act, under which only leasehold title is provided for. There are annual leases which can be withdrawn at three months notice, without reason nor compensation for improvements made. As Gasper (1989) points out, lessors cannot then use their leases as collateral for credit, and are discouraged from making investments on the site.
The foregoing discussion has shown how the colonial policies and rules which were expressly designed to discourage informal trading and small scale manufacturing have been maintained and enforced since independence. They appear to work well in achieving their objective of protecting established business operators against competition, and are actively supported by such groups as the Zimbabwe National Chamber of Commerce. The net impact has been to restrict economic participation in Zimbabwe to those with the capital and resources to overcome the multitude of legal barriers. 21
Formal structures to facilitate local participation and representation have emerged from the various enactments, directives and practices adopted over the past decade. In principle, the possibility of local involvement in decision making exists on several levels, from the village, through to the province. This section seeks to establish the extent to which broad based participation occurs in practice. (The discussion elaborates on the structures outlined earlier (Section 2.1).
Participation at the Village Level
At the village level, the VIDCO is the main focus of institutional decision making. This committee has six members - four elected, and two nominated by the two mass organisations for youth and women - whose "functions can be summarised as facilitating development, decentralised planning, improving communication, liaising between the WADCO and the villagers, and the co-ordination of development activities in the village" (MC&CD 1991). It is assisted by a multi-disciplinary extension agent known as a Village Community Worker (VCW), who works with the VIDCO and community to identify problems and development needs, co-ordinate development activities and collect, record, analyse and submit accurate information about the village. Most VCWs attend VIDCO meetings. VCWs frequently identify local needs through organising community meetings, thereby facilitating local participation (MC & CD 1991).
In practice decision making at the village level does not proceed as smoothly and effectively as the official description envisages. The villages identified for administrative purposes often bore little relation to pre-existing communities (Brand 1991). In many villages the planning exercise, described in 3.2.1 above, is weak, if not non-existent. (In the above-mentioned MC &CD survey, village plans were available for only one third of the sample.) Moreover, the traditional leaders, who theoretically lost their powers and privileges at independence, remain very influential.
Local decision making may be dominated by central government employees. The VCWs who are supposed to work with and advise the VIDCO were originally selected by the local community, but have become government employees paid by, and responsible to the Ministry of Community Development. Recent evidence suggests that an "appreciably large percentage of VCWs in five provinces are still 'telling' the community what needs it has", rather than adopting more participatory approaches to needs identification (table 31; MC&CD 1991).
In contrast, a detailed case study of VIDCO decision making in Chikwarakwara, Beitbridge revealed a substantial degree of local participation and leadership, especially on the part of the VIDCO chairman, and a high level of debate in committee and community meetings (Child and Peterson 1991). Outsiders from the District Council and NGOs who attended the meetings (to determine the allocation of CAMPFIRE revenues) provided assistance as required, but did not dominate proceedings; although it was noted that this stood in contrast to the typical CAMPFIRE meeting, chaired by a district official, where sectoral officials and NGO representatives played an active role (Child 1991). There was considered debate about a number of issues which involved, inter alia, project appraisals and the weighing up of relative costs and benefits.
Participation at the Ward Level
The Ward Development Committee consists of the chairperson and secretary of each VIDCO in the ward, plus one representative of the youth and women's mass movements. It functions as a liaison body between the VIDCO and the District Council, coordinating village plans and general development activities throughout the ward. It is also important insofar as it is an electoral constituency of the councillor, each ward being represented in the DC by one councillor. There have been efforts to improve planning and institutional capacity at the lower levels (e.g. in the Midlands). There are some very competent examples of development planning being undertaken by WADCOs, although these may not be typical. This was seen above, with respect to Mberengwa District, where Ward Profiles and Strategies have been developed (see figure 3 above).
From the limited evidence which is available on the role and operations of WADCOs (Brand 1991), it appears that ward submissions to council are often prepared by the ward chairmen without consulting the local community (Conyers and Mellors 1989). Recent local newspaper articles in Gokwe reported demonstrations against the "Big Five" councillors who are now permanently based in Gokwe, calling for the men "go back to (their) respective wards and entertain the electorate's problems... and present the residents' problems to the council" (Gokwe Post, 10/ 1991, p.1).
We now turn to assess such indicators of participation as public awareness of, and attitudes toward, VIDCOs, WADCOs and the local leadership, as well as the extent to which representation is broad based.
One recent survey found that many people did not yet know what a VIDCO was, could not tell the difference between the VIDCO and the party, nor could they identify their own councillor (RUP/UZ survey, Mutoko District). In 1985 a sample survey found that less than half of the respondents knew about the existence of a VIDCO in their area, and even fewer about the WADCO. 56% could correctly identify their district councillor, but only 20% their VIDCO chairman (Brand 1991). This is despite the appointment of Local Government Promotion Officers who are responsible for explaining the new system. (But it should be noted that lack of knowledge about local government is a pervasive world-wide phenomenon).
There appears to be considerable variation in the pattern of attendance at village and ward meetings, depending upon the issues to be discussed, the season, the weather and so on. Where there are controversial items on the agenda, local participation can be very high. Nearly 80% turnout was reported from a VIDCO in Bikita involved in a land use planning exercise (Brand 1991). Resettlement proposals arouse a lot of interest. Encouraged by strong financial incentives, the distribution of CAMPFIRE proceeds also attracts high attendance (Child and Peterson 1991). But generally, it appears that participation is quite limited. In Mutoko Communal Area, only a quarter of the sample had ever attended a meeting called by the VIDCO. It is reported from Gutu that some VIDCO leaders complained of poor attendance at meetings (Brand 1991). There is also evidence that few committees meet as often as officially required (fortnightly), and that in some places they do not function at all (Brand 1991).
A survey in Rushinga found that 79% of all respondents felt that VIDCOs had done nothing. Some of the comments on leadership included: "Local leadership arrogant" and "Leaders of the community appear to be those who 'have' and they fail to represent the needs and interests of the poor". Further, the majority of interviewees were unclear about the role and functions of the WADCOs.
There does not appear to be fully representative community involvement in the composition of the local committees. It has been observed that very few VIDCOs had more than the one mandatory woman on the committee, and that it was not uncommon for the "youth" representative to be over thirty, or even forty years of age (Brand 1991). In the contests for elected positions in local government, richer farmers, traders, chiefs and headmen have been able to use their status and ability to utilise clientele networks to reassert local dominance. The government has encouraged better educated members of the local community, such as teachers, to sit on the committees. In contrast, the participation in the village meetings in Chikwarakwara, on the wildlife project, was evidently broad based and involved roughly equal numbers of both sexes. Women contributed actively and influentially to the discussion, did not hesitate to disagree, and voted in equal numbers as the men. However women still sat on the ground and behind the men, and always showed "traditionally correct behaviour when meeting men". One outside observer noted that "there was no sign of (gender) equality, but the meetings were much more open than any I have seen in traditional settings" (Peterson 1991).
Participation at the District Level
At the district level, the elected District Council (and officials) are responsible for District functions of revenue raising and expenditure while the District Development Committee participates in plan formulation and implementation. The DC has to approve the district plan. The specific composition was described in 2.1 above. Here we examine the processes of decision making within, and between, these two institutions.
There are sub-committees of the DDC - in such areas as social development, infrastructure and agriculture - where most of the work on development proposals takes place. It appears that the DDC as a whole usually ratifies decisions made by its sub-committees (as in Maungwe D.C.). The following problems are generally acknowledged concerning decision making processes within the DDC. First and most obviously, central government tends to dominate the sub/committees. Second, the large size of the DDC - involving 20 - 30 members - can make it unwieldy, especially where up to four representatives from each department attend (as reported from Bikita). Third, some members do not attend DDC and subcommittee meetings, with considerable delegation to lower grade officers who lack authority and therefore prolong the decision making process (Mutizwa-Mangizo, 1991).
However, in Gokwe, attendance at DDC and subcommittee meetings is reported to be improving as there are fewer meetings, and these are called for specific purposes rather than merely routine affairs, so that members find that there is a real reason to attend and the quality of debate is higher. Further, the DA (Chairman of the DDC) has established an attendance roll, and reports consistent absenteeism to relevant Provincial heads (PDSP 1991).
In principle the Council dominates decision making at the district level and gives final approval to the DDP. The lines of responsibility between DDC and Council are sometimes unclear, which can result in considerable conflict - council officials and departments, versus councillors (reported from Gokwe D.C. by PDSP 1991). One area of dispute is day-to-day administration, where some councillors involve themselves in such issues as personnel management. Councillors have expressed frustration that the DDC is not sufficiently responsive to them (PDSP 1991).
More generally, case studies have suggested that most councillors participate freely at council meetings, but that their contributions are "often irrelevant, unfocused and repetitive", and that they attend meetings mainly as an obligation, regarding the meetings as a waste of time (PDSP 1991). It has also been observed that council meetings are unlikely to address broad social and economic strategies or policies - but instead focus on specific, relatively minor projects, or personal benefits, such as travelling and subsistence allowances (Chiwanga 1990).
As to patterns of participation in the District Council, these inevitably reflect the composition of the VIDCOs and WADCOs. In Gokwe District Council, a small group of relatively articulate and powerful councillors (the aforementioned "Big Five") have tended to dominate meetings, with other councillors reluctant to participate (PDSP 1991).
We can briefly compare the picture drawn of the District Councils, to that of the Urban Councils. Under the Urban Councils Act, councillors are elected every four years and operate a committee system for managing the towns. Relative to the District Councils, these councils are more independent, and have sophisticated management and technical staff that match up fairly well to their central government counterparts. Increasingly vocal opposition to rising tariffs is said to have been met through the Minister forcing the councillors and staff to enter into dialogue with the residents in order to explain intended tariffs and tax increases so that "in the end people realise that the delivery of urban services costs them money, money which the urban council can only get from them"(Moyo 1989).
Participation at the Provincial Level
Similar procedures occur at the provincial level, although the centralising forces are stronger in that the Provincial Governor is appointed by the President, provincial officials report directly to Harare, and the Provincial Council is less directly representative of local community views. There is said to be considerable disillusionment amongst provincial officials in the development planning system, given the evidence that the plans are not being used (fully) in the annual budgeting process (PDSP 1991). It is reported from the Midlands that the PDC meets quarterly, where officials from each sector present a list of projects, and there is little discussion or debate. Plan reviews focus on capital development projects. Monitoring and evaluation, undertaken by sectoral ministries, is not presented at the PDC. There is a similar lack of enthusiasm in the sub-committees, attendance is often poor and delegation frequent. Many of the PDC Sub-Committees also show apathy and lack of direction, although the sense of commitment was found to vary across sectors - in the Midlands, the Agriculture Sub-Committee was much more active than, for example, Housing and Employment.
The Provincial Councillors are reported to be generally unsure of their role and likely to rubber-stamp whatever is proposed by sectoral ministries through the PDC (PDSP 1991). Interviewing participants, Mellors and Conyers (1989) found that Provincial Councillors are not fully aware of the potential role of the Council. They are under the impression, for example, that they have to accept the decisions or recommendations of the PDC, and that it is not their responsibility to initiate discussion or act to improve the planning system.
To summarise, central officials often dominate DDC and PDC meetings because of their technical competence, superior resources and (often) numerical dominance on the development committees. Moreover, the councils can only comment on the development plans, but cannot force their decisions upon the development committee. The DDC is made up of all heads of central departments operating in the district, plus a senior executive officer representing the District Council. In one study it was found that in DDC meetings attended by 35 persons, there were only two representatives of the local authority (the council chairman and the senior executive officer), the rest being central government representatives (Mutizwa-Mangiza 1990). The frequent absence of village and ward plans was noted above, which means that the sub-committees of the District Development Committee often end up drafting project proposals. The Central government appointment to the powerful position of District Administrator22 and Provincial Administrator (as well as Provincial Governor) further increases the power of the central government. It is reported from the Midlands, for example, that the Provincial Council feels itself to be in a weak position vis a vis the Provincial Development Committee, and ends up rubber-stamping PDC submissions.
The role of the party
The Party structure operates at several levels broadly following the structures established by ZANU during the war (figure 8). It has been observed that the Party, present from the grass-roots level in the form of cells to the very top (the politbureau) is "obviously" more powerful than local government, "which is strongly represented (only) at district level and below" (de Valk 1990). At the local level, the party network has been active in mobilising such activities as self help, where it organises and collects contributions. One survey carried out in Charewa, Mutoko district in 1984 found that a large proportion of respondents (one-third) claimed to hold some Party office (Brand 1991).
Brand (1991) refers to the "deliberate (con)fusion of the structures of party and state in Zimbabwe", whilst pointing out that this has not proceeded as far as in Zambia or Tanzania. It was noted above that some surveys have found that many people cannot distinguish between party and local government structures. Although the ZIDS study of Rushinga district concluded that most people knew the difference between the ZANU(PF) structure at the village level and that of the VIDCO "in that the former is concerned with political matters, whereas the latter structure addresses developmental issues" (ZIDS 1986), it is not clear from the results whether the two functions were regarded as mutually exclusive, or whether the party monopolised local discussion of a political nature.
The relationships between the Party and local government operate in two ways: through the individual, who may be a member of both simultaneously, and through institutional relationships. The 1984 Congress proclaimed the supremacy of the Party over the government in outlining the main directions of policy, and appointed five standing committees charged with overseeing government policy.
It has been observed that in most parts of the country, outside Matabeleland, the network of provincial, district, ward and village development committees consist almost entirely of ZANU nominees (Cliffe and Stoneman 1989). ZANU(PF) has asked all civil servants to become Party members (Chikowere 1988). It is also said that the District Administrators outside Matabeleland were regarded as ZANU agents (Libby 1984). Likewise the Local Promotion Officers who were charged with organising rural development projects also assisted in Party activities such as transportation to ZANU meetings and fund raising. In the Mashonaland provinces where the ZANU has been virtually unchallenged, the pattern of local government elections has basically consisted of the nomination of official Party candidates who are then unopposed (Cliffe and Stoneman 1989).
Party dominance is facilitated by the general absence of secret ballots in village, ward and council elections. The typical mode of election requires participating constituents to line up behind their preferred candidate, or a show of hands or voices (as provided for in the District Councils (Elections) Regulations, 1980: S3). In the majority of interviews conducted with District Council officials in Bikita this was regarded as undesirable (Mutizwa-Mangiza 1991). Secret ballot was considered by interviewees to provide a means to enable voters to vote against party candidates and traditional leaders without fear.
Given the dominance of the party, the extent of devolution depends not only on formal governmental structures, but also on how far the party itself is subject to central control. Decentralised structures could be offset in practice by a highly centralised and dominant party. In the Zimbabwe case, the party is not tightly controlled from the centre, but centralising forces (over nominations and finance) undoubtedly further weaken the genuine devolution of powers.
Conclusions on political participation
The process of decentralization in Zimbabwe has been characterised by the multiplication of formal opportunities for citizens to participate in decision making. These exist on several levels: direct elections select village committee members; while the chairpersons and secretaries of each VIDCO act as the representatives at the ward level, together with youth and women's representatives. Then the chair of each ward becomes a member of the District Council. Members (Chairpersons and one other) of the DCs constitute the PC.
However three factors have constrained effective broad-based local participation in the system of local government - the dominance of decision-making structures by central officials; the tendency of a small elite to represent local interests, with limited popular participation; and the strong role of the Party.
Overall, it can be said that the decentralizing reforms have created the framework for broad based representation and widespread consultation with local communities across Zimbabwe. However the evidence available casts doubt upon whether this ideal has been realised. This can be attributed to a range of factors, some which might be addressed relatively easily - such as the introduction of secret ballot at the local level - and others which reflect more deep-seated characteristics of national politics and society, which are more difficult to overcome. Nonetheless developments are still at an early stage, since it is less than a decade ago that the local structures for participation were created. The danger however is that the Councils, WADCOs and VIDCOs will function as "mere conduits" (Brand 1991) between local elites and (or party nominees) on the one hand, and central bureaucrats and technicians on the other.
The general picture of rather limited participation is perhaps surprising, given the high level of popular mobilisation during the struggle for independence. But there was not substantial continuity between these movements and the formal structures of local government created after independence. It has been argued that the "strength of the grassroots organisations lay decisively in the interaction of peasants with guerrillas" (Stoneman and Cliffe 1989), so that the withdrawal of the latter removed much of the momentum. Moreover the restructuring of local government was basically a top-down initiative which did not take existing popular organisations into account. And up until 1985, when specifically development-oriented committees were created at the sub-district level, the lowest level of government (districts and wards) were "far from the peasant universe of the kraal" (Stoneman and Cliffe 1989). Brand (1991) argues that the formal decentralization villages are largely demographic creations based on the desire for round figures23, which are typically several times larger than pre-independence communities. In some places the people's committees remained significant - in one district in Chibi in the south, de facto authority for land allocation was exercised by the committees, rather than by the new District Councils where the formal power rested.
13 The Tax Commission Report described this as "very undesirable" from a policy perspective, and suggested that "nothing short of a complete overhaul of local accounting systems to conform to that of the central government will enable decision makers to judge the relative importance of various revenue and expenditure heads" (p. 331).
14 For 1981/2, s = 105.8 - 2.07o*, r2 = 0.264, and for 1984/5, s = 101.9 - 0.83o*, r2= 0.174, where s = the proportion of total DC expenditure going to the social sectors; and o* = the proportion of total DC income which comes from the D.C.'s own sources.
15 This is classified on the basis of DDC Sub-Committee - so that the definition of "Social Development" includes such items as social welfare, community halls and uniform making, as well as health and education.
16 This is not, however, a "perfect" experiment, insofar as the Council has an obvious incentive to nominate projects which would have been unlikely to succeed through the normal channels. This may explain the absence of requests for clinics and secondary schools, for which funding may be more easily forthcoming through the relevant sectoral ministries.
17 Data is available for some districts for certain indicators. See Appendix 1.
18 The situation for Urban and Rural Councils is different because they rely heavily on a rating system which is likely to be progressive.
19 I don't have any information which could allow a more detailed story to be told about the distribution of extension services, or any assessment of differences in impact across the country (e.g. whether the 40% of households who do not market any surplus also tend to be left out of the extension effort.
20 A vivid example is the story of an owner of shop owner in a very remote part of Gokwe District, who had the sole business in a so-called Business Center, who was advised by a vigilant passing local official that his shop would have to be demolished because it breached planning regulations. The land use plans indicated that a future road was to run behind the shop, in which event it would be "facing the wrong way".
21 Examples can also be given of centrally-imposed controls and regulatory procedures which hamper local development - e.g. applications for water rights to construct dams or draw water.
22 The 1998 legislation will enable the new Rural District Councils to appoint their own professional chief executive - a practice which had long existed in both the urban councils and the former rural councils.
23 That is, from a projected national population of six million, given 6000 people per ward, to obtain 1000 WADCOs and 6000 VIDCOs.