Part III - Institutions and Development in Tanzania

Based on the five preceding analyses of key thematic areas, and the more general description of the political, economic and institutional features of Tanzania’s development in the first part of the volume, this last part tries to synthesise what has been learned. This exercise leads to an articulated view of the main institutional problems hindering progress in various areas, their negative consequences for development, and, most importantly, their causes, proximate or more distant, as well as their susceptibility to reforms. This is the essence of the ‘diagnostic’ this whole volume has endeavoured to produce.

9 An Institutional Diagnostic of Tanzania’s Development

François Bourguignon and Samuel Mwita Wangwe

This final chapter has two goals. First, it aims to synthesise what has been learned from the previous chapters, emphasising the institutional challenges that have been identified. Second, it seeks to identify the main common factors behind those challenges, as well as the policies and reforms most able to weaken their effects. Indeed, it appears that most of the identified instances of institutional challenges in the Tanzanian economy result from a small number of basic institutional weaknesses, which logically form the core of our final diagnostic. The reflection on the ways to remedy them, while taking into account the political economy context, including the structure of political power, leads to a set of recommendations for reform that follow two key principles about the way to approach reform of the legal and administrative apparatus in Tanzania. The first principle is to allow for more competition and market mechanisms both within the economy and in the way it is managed. The second principle emphasises the need for a continuous rigorous, independent, and transparent evaluation of the functioning of the public sector at all administrative levels, of policy outcomes, and of the socioeconomic progress being achieved. Whether these principles may be applied mostly depends on political economy factors.

I The Identified Institutional Constraints on Tanzanian Development

This first part of the chapter summarises the conclusions drawn from the seven preceding chapters. Organising them within an integrated and coherent framework is undertaken in the second part of the chapter.

A The Uncertain Growth Engine of Tanzania’s Economic Development

Tanzania’s development since independence has been deeply affected by major institutional changes that took place during this period – from a market economy and import substitution industrialisation just before and after independence, to a socialist economy after the 1967 Arusha Declaration, to a market economy again after the economic crisis of the late 1970s and early 1980s that led to the Economic Survival Strategy (1981–2), a home-grown Adjustment Programme (1982–5), and the Structural Adjustment Programme (SAP) driven by the International Monetary Fund (IMF) and the World Bank after 1985. Such a major institutional oscillation is bound to have left some deep imprints on the way today’s economy works.

If the severe slow-down that characterised the difficult shift from a socialist to a market economy during the 1980s and the first half of the 1990s has now given way to solid growth, several concerns can be expressed about its long-run sustainability.

First of all, there is uncertainty about what is and what could be the growth engine of Tanzania in the future. Such an engine must necessarily relate to the tradable sector. This is because final demand constraints related to domestic income are absent in foreign trade and also because production for the domestic economy does not produce foreign currency resources. Relying on non-tradable activities thus requires some growing exogenous source of outside revenues. Yet agricultural output has grown less rapidly than gross domestic product (GDP) since the mid-2000s, and manufacturing has been roughly on a par growth with GDP. It is therefore unlikely they played a leading role in the overall growth of the economy, especially in view of the relatively modest GDP share of manufacturing – that is, 9 per cent. Favourable terms of trade as well as foreign capital and aid flows were more important explanatory factors of the satisfactory growth performances observed since the turn of the millennium. They raised aggregate domestic demand, fostering production and investment in goods and services that could not be imported. This may not last. The emphasis put on industrialisation, including the agroindustry and therefore agriculture, by the present administration and its predecessors is thus most justified. So far, however, results have been limited, and identifying whether institutional constraints are responsible for this lack of dynamism is of great importance.

Three other sources of concern come out of our review of Tanzania’s development. The first is the slow, and in some cases negative negative growth of productivity throughout most sectors of the economy, except in agriculture, where GDP share grew despite a falling share of employment. This seems in contradiction to the rather high observed rate of investment and suggests the latter may not be well allocated, with possibly a deficit in productive factors complementary to private investment, infrastructure, and power in particular. A second concern is the excessive dependence of the economy on foreign financing, including official development assistance. This means that the high level of investment of the last decade or so may not be sustainable in the long run, unless domestic savings can progressively substitute for foreign fund inflows.

The final concern is about the slow reduction of poverty, despite relatively high aggregate growth per capita performances, indicating that growth has not been particularly inclusive. This may be because of the nature of growth or essentially a lack of integration of agriculture and the informal, predominantly rural sector in the growth process, another key institutional characteristic of Tanzanian society. This bias is reinforced by the low quality of public services delivered to poor people. The low quality of education in particular reduces the future ability of the poor segment of the population to integrate within the modern economy.

B Tanzania Compares Well with Peer Countries in International Governance Rankings

There are a number of governance indicators that aim to summarise in a simple multi-dimensional scale the various dimensions of countries’ governance. It is not clear whether such a complex system as the governance of a country can really be summarised through a few summary indicators. Moreover, the significance of most available indicators in terms of symptoms or causes of institutional challenges is often very ambiguous. However, they convey information that is worth examining.

Two approaches have been used in the present study. First, use has been made of the six aggregate indicators of the well-known Worldwide Governance Indicators (WGI) database. They are themselves based on a large number of sub-indicators, which were classified into six categories, and then a ‘common factor’ was statistically extracted from each of the categories. The second approach was developed for this study, with a statistical procedure being used to classify all individual indicators in the extremely rich Quality of Government (QoG) database according to the proximity of their cross-country variations rather than imposing an arbitrary grouping. Then, a common factor was extracted from all indicators within each of these endogenous categories to characterise how a country performs within a category. This endogenous categorisation yields results slightly different from the WGI. In addition to these two sets of indicators, use was also made of the investment climate surveys conducted by the World Bank based on a sample of firm managers.

The main governance challenges according to WGI are ‘control of corruption’, ‘government effectiveness’, and ‘regulatory quality’, in that order. The procedure based on QoG data points first to ‘administrative capacity’, then to ‘competitiveness’, which is similar to ‘regulatory quality’ in WGI, and finally to the ‘control of corruption and the rule of law’. Within that approach, however, corruption is high in Tanzania but not worse than in neighbouring countries (Kenya, Malawi, Uganda, Mozambique, and Burundi), except Rwanda, or in better performing developing economies such as Bangladesh, Vietnam, and Cambodia.

We draw from this statistical exercise that ‘government effectiveness’ or ‘administrative capacity’ to manage the economy seem to be the most serious weaknesses in Tanzania, even though what these generic titles cover and what is primarily responsible for the lack of effectiveness is not very precise.

‘Regulatory quality’ or ‘competitiveness’ include the impact of taxes, business regulation, and the quality of public infrastructure and services on the ease of doing business in general. ‘Doing business’ and ‘competitiveness index’ rankings released respectively by the World Bank and the World Economic Forum confirm the negative expert appraisal included in the preceding indicators.

Corruption is widespread, or even ubiquitous, in the developing world, whether in sub-Saharan Africa, Asia, or Latin America. This does not necessarily hamper all aspects of development. Some aspects of corruption may even facilitate economic development, as when it permits the relaxing of administrative constraints that divert the allocation of resources from their efficient use. Yet it is noteworthy that, substantially more than with the WGI indicators, Tanzanian business managers find corruption to be particularly detrimental to the economy.

C How Tanzania’s Decision Makers Evaluate Its Institutions

Two surveys were launched to ask decision makers in different positions in society what they felt were the most economically constraining institutional challenges in Tanzania. The first was a questionnaire survey administered to a sample of about 100 decision makers from the public and private sectors, at various hierarchical levels and in different areas of the country. The second survey consisted of open-ended interviews with top decision makers and politicians within different areas of expertise and from different horizons. This second group of respondents comprised sixty people, including two past presidents, the Chief Justice, and the Controller Auditor General, as well as top civil servants, businesspeople, and academics.

For the most part, these surveys and in-depth interviews were conducted mostly in 2016 and early 2017. As this was the first year of the Magufuli administration, one can wonder whether respondents were referring to the previous or the new administration when they were interrogated. The formal survey asked the respondents to refer to the medium- to long-term past (described as the past five to ten years) when answering the questions. This timeline was given in recognition that institutions take time to change, and therefore it would be more informative to understand long-standing patterns than any perceived changes that may come with a new administration. To capture the perceived direction of change with respect to the past, respondents were later asked whether they thought things were changing with the new administration. The same principle applied informally to the open-ended interviews. The evaluation of Tanzanian institutions by decision makers thus essentially focused on the pre-Magufuli era.

The most frequently cited institutional challenges that came out of the questionnaire survey and the open-ended interviews were the political institutions, the way politics is done in Tanzania, the public administration, and the business environment.¹ More precisely, the following domains were singled out as particularly problematic:

• land law, whose complexity and slow implementation entail social conflicts, numerous court cases, corruption, and frustration among both smallholders and investors;

• corruption in general, between government or administration and business, between the administration and people, and between government and citizen (patronage);

• a lack of transparency in state operations and the weakness of check-and-balance institutions;

• poor regulation of business activity, referring to the operations of big businesses or to the general investment climate;

• poor regulation of utilities, especially electricity; and

• the inefficiency of the civil service: low capacity, absenteeism, corruption, overlapping responsibilities, and patronage in recruitment practices.

Of course, this list is a mixed bag in the sense that it includes institutions per se (e.g. the land law or the executive’s checks and balances) but also the causes (e.g. lack of capacity in civil service) and the symptoms (e.g. absenteeism of civil servants, corruption), and the consequences of these weaknesses (e.g. land conflicts or investment climate). Yet the general picture these surveys consistently draw of economically relevant institutional challenges in Tanzania is instructive, and parallels the conclusions drawn from governance indicators.

Identifying the causes of institutional weaknesses and differentiating symptoms and consequences was undertaken in the second part of the study, by focusing on specific thematic areas thought to be of critical importance both for the functioning of the economy and for our understanding of the nature of institutional challenges.

D Institutions and Development in Some Critical Areas

Five areas were chosen based on the information collected from decision makers in the preceding stage of the study, which were thought to exemplify the way particular institutions may obstruct or slow down development. Each area gave rise to an in-depth analysis, generally, by a Tanzanian scholar, and was then discussed by a foreign expert so as to put the issues at stake in an international perspective. A summary of each study is given in what follows.

1 Business and Politics: An Ambiguous Relationship

In Tanzania, as in most developing countries, big modern firms, or conglomerates, contribute a sizeable part of GDP growth and are the main employers in the modern private sector. Because of this, there are good reasons for governments to support them, presumably in a fully transparent way. But because of both positive and negative externalities and/or market failures more frequent in developing economies, their activity should also be regulated, again in the most transparent way. The relationship between big business and the government, or the ruling party, is particularly opaque in Tanzania and frequently subject to scandals of corruption uncovering anti-developmental policy decisions. This is contrary to a well-thought policy framework that would establish clear reciprocal behavioural rules for business and government agencies in the interest of the community. This absence of a transparent and effective industrial policy – in the broad sense of regulating, monitoring, and possibly supporting big as well as medium-sized business – appears to be a major institutional challenge with major implications for Tanzania’s development.

This key institutional issue has been analysed in depth in Chapter 4, ‘Politics and Business: An Ambiguous Relationship’ by Samuel Mwita Wangwe, and in Hazel Gray’s comments. Various factors explain the current state of play and the relative inefficiency of industrial policies and strategies. Three of them seem particularly important. In some cases, they point themselves to other structural weaknesses of the politico-economic apparatus.

First, regime changes in recent history may not have left very much time for market rules and industrial policy principles to settle down in the national community. The economy went from markets subject to colonial rule before independence, to more open markets just after independence, to limited and mostly hidden market operations during the socialist period, and back to markets with rules progressively being set but still incomplete or largely ineffective during the liberalisation period and in some cases until the late 2010s.

Going back and forth across regimes also implies that there is no full political consensus in the Tanzanian political elite about what regime to favour, whether full market capitalism or a mixed economy with substantial weight given to the state. This absence of clear consensual principles with respect to the relationship between the state and the market, or private firms, may be considered as an institutional failure per se as it introduces uncertainty among economic actors about the attitude of the state with respect to private firms. This ideological ambiguity may be a second reason why no clear strategy with respect to big business and the private sector in general is yet fully explicit. Almost forty years after the end of the socialist era, the Magufuli administration exhibited such ambivalence with respect to private investment. However, the Suluhu administration now reverted to a more business-friendly approach.

A third reason has to do with the ethnicity of the owners and managers of big business. It turns out that a majority of large firm owners are Tanzanian Asians or Tanzanian Arabs. Several Asian businessmen were present in Tanzania just before or just after independence. Starting from trade businesses at the local or regional level, they invested in manufacturing behind the protection barriers erected, on their own advice, by the young independent state. Behind the scenes, or sometimes as managers of parastatals, they continued to have influence during the socialist era. After liberalisation, they progressively built economic empires, soon joined by Tanzanian Arabs. A majority of big business leaders thus belong to ethnic minorities that have tended to dominate economic life outside agriculture, practically since independence, with less strength, of course, during the socialist era. These ethnic groups are not fully integrated with the indigenous African population, which tends to see them as ‘exploiters’ (Fouéré, 2015), and considers them today as running a large part of the Tanzanian economy. Because of this, any policy that would openly favour business might meet popular opposition if the main beneficiaries were perceived to be Asian and Arab big businesspeople. Successive governments may have refrained from developing this kind of policy, even though it was possibly beneficial for development, explicitly because of this often-unacknowledged ethnic division.

It is unclear why there are few indigenous Africans among top chief executive officers, especially in industry, even though things may be progressively changing. At the time resentment against Asian business was at its highest, indigenous African entrepreneurs complained that an Asian monopoly prevented them from raising funds from banks or hiring people with the required know-how to get into big business. Such a complaint is not necessarily justified. Yet today, only two indigenous Africans are among the fifteen top business leaders in Tanzania. Younger indigenous African businessmen have recently been in the news, though.²

Another reason for the lower weight of indigenous in comparison with non-indigenous Tanzanians among business leaders might be found in the clear comparative advantage, because of ethnic proximity with the people, of the former’s presence in political or public service careers rather than in business, although this could be compensated for by their much larger population size. On the other hand, it is true that Asian and Arab business leaders have sometimes moved openly into politics to reinforce business’s position in general, rather than acting from behind the scenes. In any case, this ethnic division in Tanzania, which is found with comparable features in other African countries,³ is clearly an important determinant of the functioning of economy and politics.

A fourth and last reason can be invoked to explain the difficulty of establishing clear rules for industrial and regulation policy. This is the fragmentation of political power within the dominant CCM party as well as the existence of groups within the party with diverse economic interests in big business. Divergent interests and views, including ideological, across CCM factions themselves linked to diverse business interests make it impossible to get an agreement on policy decisions involving big business interests, or to prevent party members colluding with business, or even to expel them when caught in corruption scandals.⁴ A centralised power within the ruling party would probably eliminate this obstacle, and this may indeed be the strategy that was initially followed by President Magufuli, formerly a second-tier personality within CCM, who was elected on a strong anti-corruption mandate in 2015.

If the lack of explicit and effective industrial and business regulation policies is to be considered as a key institutional weakness of the Tanzanian economy, caused by, and leading to, corruption, clientelism, and cronyism, two important questions must be considered.

First, if no explicit industrial policy favouring some specific sector through protection, infrastructure development, fiscal exonerations, or subsidies seems possible or desirable, then why not adopt a neutral policy that would favour business in general – that is, the World Bank ‘doing business’ approach? Why wouldn’t big businesspeople themselves be asking for such an industrial policy, sometimes referred to as ‘horizontal’ industrial policy? After all, a better investment climate would dynamise investment, foster the development of small and medium-sized enterprises, accelerate growth, and finally benefit big businesspeople too. As long as such a policy would not change their own arrangements and the support they get from government people, they should not oppose such an improvement of the investment climate. The difficulty here is that it would be difficult to improve the business climate by fighting civil servants who are taking bribes for delivering a licence or a permit to new investors or operating firms, when bribes from big business are known to be cashed higher up in their hierarchy. The same would be true when deciding about public infrastructure investments, big firms managing to have them serve their own interest rather than that of the collective. Corruption is contagious. It cannot be stopped for some operations when it continues elsewhere. Without powerful control measures, big business’s grip on parts of the political elite thus trickles down to a deleterious general business climate.

Second, isn’t it the case that the Tanzanian economy does well, so that business–politics collusion may not be a big problem after all? Maybe, but is it doing well sustainably? The review of Tanzanian development in this volume points to the relative weakness of growth in tradable sectors, especially agriculture and manufacturing. Could it be that the big business–politics connection has some responsibility for this? Wangwe indeed suggests that imports are often given some advantage over domestic production through supporting the ‘traders’. Sugar and rice producers have complained that imports make domestic production barely profitable. Is such an industrial policy in the interest of the country – in other words, the consumers and industrial transformers of these products – or in the private interest of traders? Transparency about the goal pursued with such a policy and its consequences is lacking. Public statements are made about policies to be pursued or reforms to be undertaken, but they are generally weakly motivated and not, or poorly, evaluated ex post. This is what Wangwe refers to as transparency without accountability.

2 State Coordination: Fluctuating Stances about Decentralisation

Development requires an efficient government sector that is able to produce and manage public goods, monitor and regulate the private sector, collect taxes to finance these activities, and redistribute across social, geographic, or income groups. The way the government organises its activity is essential for its efficiency. A key organisational issue in this respect is the allocation of decision power among the various decision and monitoring units of any government, from the centre (president or prime minister), to ministries, to local governments and various types of public agencies.

It would have been too unwieldy to analyse the efficiency of the Tanzanian government apparatus by reviewing all its dimensions. Cases of responsibility overlaps, competition across units, and rent-seeking seem to be present in many areas and at various levels of the public sector’s vertical structure. Instead, the approach to the issue of state coordination in this volume has focused on the relationship between two levels of government, the central and the local, on the way it has evolved over time and a practical aspect of it, namely the collection of local taxes.

The dominant fact in the fascinating account that Servacius Likwelile and Paschal Assey give of decentralisation in Tanzania in Chapter 6 is the constant vacillation of the government since independence about the decision power that should be granted to local government authorities (LGAs), both in terms of local service delivery and tax collection, that is between decentralisation taken as deconcentration, with centrally appointed civil servants in charge of local affairs, or devolution of decision-making power to local governments and their democratically elected decision-making entities. The central–local government relationship went from decentralisation, inherited from the colonial period, in the first decade after independence, to full centralisation during the early socialist period, back to weak decentralisation in 1972, and then to centralisation again whereas an ambitious local government reform was voted in 1998 that had to be progressively implemented over a rather long period. As of 2014, the end of that period, however, neither a comprehensive local government law nor a harmonised central and sector legislation was in place. This is presumably still the case today, although the Suluhu administration has recently shown signs of willingness to implement the National Decentralisation Policy, which has been on hold for two decades or so.

In the field of fiscal decentralisation, the oscillation was still more pronounced. The responsibility for the collection of the property tax changed three times in the last decade: decentralised before 2008, centralised, and then re-decentralised in 2014, and apparently centralised again in 2016.

Such back-and-forth movement would not be a concern if it simply corresponded to a trial-and-error process, the aim of which was to find the most efficient arrangement. This does not seem the case, however, as the periodicity of changes does not permit rigorous evaluation of the various regimes. As well acknowledged in Chapter 6 and in the comments by Jan Willem Gunning, capacity is clearly missing at the local level for the assessment, the monitoring, and the enforcement of taxation. For the time being, it is thus the case that tax collection should optimally be in the hands of the central government, while local governments rely on conditional central government grants to finance local operations not centrally managed. On the other hand, local tax revenues account for a very minor part of the total budget of local governments. Together, these two arguments logically lead to the hypothesis that oscillating responsibilities reflect either some hidden rent-seeking competition between local or regional politicians, central government tax collectors, and possibly ruling party members, or the views of some external decision maker, such as donors or a new leader.

Likwelile and Assey recognise that corruption was rampant both with the centralised and the decentralised tax collection arrangement. They also mention that the World Bank had apparently been pressing for decentralisation to be re-established in 2014, a decision that was cancelled in 2016 under the Magufuli presidency, thus showing a consistent, and probably justified, opposition of the central government to decentralising tax collection. Even in that case, however, the issue of corruption remains open.

Fiscal decentralisation is often justified by resorting to the argument that it is an incentive for the accountability of local governments with respect to their constituency. Gunning is right in stressing that fiscal decentralisation is in fact neither necessary nor sufficient for local government accountability. Sharing information about the actual grants received from the central government and the accompanying conditionality with local constituencies could be effective too, as well illustrated by the positive effect of advertising central government grants to local schools on the management of schools in Uganda (Reinikka and Svensson, 2005, 2011). Existing studies in Tanzania suggest, however, that not much of the publicising of local financial information reaches the public – see Chapter 6 by Likwelile and Assey.⁵ Studies also show that there is an overlap or confusion of responsibilities between local government executives and centrally appointed civil servants working with them or monitoring them at higher district or regional level, but often applying the central government’s agenda (Venugopal and Yilmaz, 2010; Hulst et al., 2015).⁶ None of this is contributing to local accountability.

The preceding mostly anecdotal evidence indicates that decentralisation is not very effective in Tanzania. This means that local preferences for local public goods and constraints in creating and managing them may not be satisfactorily taken into account in the overall public decision-making process, thus reducing the overall efficiency of the public sector.

Should decentralisation be made more effective in Tanzania, and more efficient in view of its potentially high current cost? As noted by Gunning, this is an empirical question, but there is unfortunately little evidence to answer it, in Tanzania or in developing countries in general.⁷

Yet the chapter by Likwelile and Assey, as well as Gunning’s comments, raise an important question for our reflection. It concerns the issue of capacity, and consequently of capacity building. An obvious limit to an enhanced decentralisation strategy in Tanzania is the capacity of local executives to produce local public goods and manage budgetary issues as well as the relationship with their constituency and with central government. This is very clear in the issue of tax collection but probably applies to other dimensions of local governance, even though centrally appointed local administrators or monitoring agents may not always have more capacity themselves. Under these conditions, would a central government resist enhancing decentralisation because it is not in its political interest, or simply because local executives don’t have the necessary management capacity? If decentralisation is essentially a good policy, with adequate local capacity, then it would be important to facilitate capacity building at the local level, a point stressed by Likwelile and Assey. This would probably be the most efficient first step towards a more ambitious decentralisation policy.

In summary, it is difficult to say whether limited decentralisation has been a handicap for Tanzanian development. That it is a desirable long-run goal is not debatable. Yet conditions and, most importantly, the local capacity for a strong effective decentralisation are presently absent in a substantial part of the country. The real handicap, however, lies in the succession of contradictory policy regimes, which has prevented learning taking place over time. Of concern is also the fact that corruption seems to have been present under the various regimes.

3 The Civil Service

The smooth and transparent functioning of modern states relies on the quality of the civil service. Several constraints interfered with the satisfactory working of the civil service in Tanzania in the past and hampered its economic development. Some of them tend to be less severe today, but serious challenges remain.

In Chapter 5, Rwekaza Mukandala lists four such weaknesses: capacity, motivation and conduct, politicisation – or political interference – and resources and tools. Capacity and resources clearly depend on the level of development and the availability of human capital and public funds. From an institutional point of view, motivation and political interference play the most important role because they ultimately define the rules according to which the civil service, the political apparatus – that is, government, parties, parliament – and the private sector interact in the economic development process, and which possibly deviate from official rules.

Initially extremely limited but fast growing after independence, the Tanzanian civil service expanded still faster but, in a first stage, in an orderly way during the socialist era. Needs were enormous because of the expansion of services such as health and education, the centrally controlled management of local communities, and the need to coordinate the multiple nationalised and newly created public companies. Collective motivation was great, and strict rules preventing high-level civil servants doing business or enjoying hidden material advantages were enforced. Unwise decisions at political level, the multiplication of parastatals, the quasi-impossibility of coordinating them, and external events, rather than a dysfunctional civil service, were responsible for the chaos and the severe crisis of the early 1980s.

Before then, however, and still worse during the crisis, the functioning of the civil service had severely deteriorated. Not only had it been completely disorganised, but also initial rules of conduct had been dismantled, partly to compensate for the effects of the crisis. High-level civil servants and parastatal executives obtained huge non-monetary compensation for declining real wages, whereas they were making deals with the private sector to further increase their income. A huge wave of corruption settled in.

Things became worse at the time of the transition back to a market economy. The process of privatisation and the lack of a civil service able to efficiently manage this new economic system opened the door still more widely to all sorts of deals involving civil servants, politicians, and businesspeople. A culture of corruption permeated the whole economy and the public sector in particular. At the same time, the civil service was decimated by the SAP and its focus on market mechanisms as a substitute for state intervention.

It took time before the situation came back to a more normal position. At the turn of the millennium, when donors started to weaken their market bias, the civil service could at last fill its role as a policymaking facilitator and the manager of public goods and services. Yet corruption, political patronage, and the pre-eminence of politics in policymaking were entrenched. This is still the situation today, even though President Magufuli broke away from his predecessors in showing political will and energy to reduce corruption.⁸

Mukandala gives many examples of the way in which corruption and the lack of motivation of civil servants are costly for the economy and the society. For instance, the cost of teacher absenteeism has been evaluated at 1 per cent of GDP. Assuming that other civil servants display equivalent behaviour, showing up at work but not working the legal number of hours or shirking, the cost would rise to 2.5 per cent of GDP. Add to this tax evasion, over-invoicing of contracts, and other corrupt practices, and the cost to the economy becomes sizeable. Patronage is another issue. The size of the public sector increased by 50 per cent between 2012 and 2015, with a net addition of 70,000 employees every year. Questions arise whether all these employees fill real needs; whether the newly recruited employees are friends or relatives of politicians – or civil servants; whether it is possible to expand at such a pace while maintaining the same level of productivity; and whether recruitment criteria had to be substantially lowered for these new employees to be appointed. Likewise, it turns out that civil servants are better paid – a 30 to 40 per cent premium on average – than private sector workers, even controlling for educational level and age.⁹ Although the difference may be smaller for top positions, is this wage premium justified by efficiency arguments or is it a political gift to a pivotal part of the electorate?

Even though Tanzania’s level of corruption seems comparable with its neighbours and even better growth performers, this is certainly not a reason for not doing anything to improve the quality of Tanzania’s civil service. It is unfortunately not possible to give a precise estimate of how much income is being lost or diverted because of a weak civil service. Available evidence is too partial to permit such estimates. Likewise, as emphasised in Jan Willem Gunning’s comments, it is not possible to evaluate how much growth has been missed. Yet absenteeism, the excessive number of civil servants, and their inadequate skill are doubtlessly detrimental to efficiency and growth.

It remains to be seen whether civil service reform can be done and how. As emphasised by Gunning and Mukandala, any reform has its own political economy context. In the present case, the countervailing forces to a reform would be civil servants themselves, the politicians who use the civil service for patronage or as a channel to cash bribes, and those private agents who benefit from bribing civil servants and politicians. All of them have power, including the civil servants themselves. Yet introducing more discipline in the areas where absenteeism and shirking are strong does not seem impossible. A lot of experiments have been run lately in various countries involving the kind of carrot to offer and the kind of stick to use.¹⁰ Improving the quality of recruitment, devoting more resources to capacity building, and avoiding the recent oscillation in the number of new hires cannot do any harm and will not meet very much opposition either.

4 The Tanzanian Land Management System: Complex, Inefficient, and Conflictive

Land has been the subject of intense debate and dispute in Tanzania as it is the main resource of a large population of poor farmers and cattle herders as well as an essential input for commercial investment in rural and urban areas.

Chapter 7 by Sist J. Mramba sets out to analyse the existing system of land occupancy rights and its effectiveness in managing multiple sources of conflicts, and proposes how the system could be reformed. It is shown that there is some continuity in matters of land tenure (such as government ownership of land) between colonial times and post-independence Tanganyika, and later the United Republic of Tanzania, whereby some basic principles of the colonial era were kept after independence. An important modification that was made is in respect of the way the land could be alienated from customary users so as to protect small farmers. Indeed, the whole period after independence has been associated with a permanent debate about the space to be given to customary laws in handling interests of public and private large investors. The creation of the semi-collective Ujamaa villages in Nyerere’s socialist era impacted on the alienation of customary land in the process of reorganising production. The liberalisation period in the mid-1980s reverted to the dual land system, with the development of large-scale plantations and better protected customary land rights. Nevertheless, land conflicts persisted unabated.

The Land Act (1999), a major piece of legislation, reiterates that all land in Tanzania shall continue to be public and remain vested in the president as trustee for and on behalf of all the citizens of Tanzania. This is of course a key institutional principle. A second principle is the recognition of the customary law through a categorisation of land. For the purposes of management, land is categorised into ‘general land’, ‘village land’, and ‘reserved land’ directly managed by the state. On general land, officially agreed users are given an extended lease, or granted right of occupancy, governed by statutory law. On village land, occupancy is governed by local customary law. The corresponding rights can be formalised by Certificates of Customary Rights of Occupancy (CCROs), with the land under the jurisdiction of the village itself being certified. However, these certificates cannot be delivered as long as the village has not had a its land surveyed and issued a land use plan, which leads to the release of a Certificate of Village Land (CVL).

While the Tanzanian Land Act was associated with much optimism in Africa,¹¹ Klaus Deininger points out in his comments that in practice four elements illustrate the gaps in Tanzania’s land registration system for smallholders, which impose costs on the broader economy. These four elements are: (1) low level of computerisation of land administration; (2) lack of integration between spatial and textual records; (3) high fees for formalising transfers; (4) and limited number of new CCROs created annually. In this regard, Deininger proposes action on improving registry efficiency and integration, adjusting regulations for low-cost first-time registration, and completing CCRO issuance in urban areas.

Land for investment is a major challenge. While large-scale land investment by foreigners has continued to be welcomed, in practice acquisition of land for investment has been complex, especially where the land is village land. In this case, village land has first to be transformed into general land, which presently represents only 2 per cent of total land, and the procedure is cumbersome. Consequently, land investments are limited with consequences on economic growth.

Although the Land Commissioner, acting on behalf of the president, has power to grant general land, the use of this power is not common, and the process is long. Challenges of valuation have attracted frequent land conflicts. The initiative to acquire land may also come from the Tanzanian Investment Centre responding to the demand of investors. However, there are also less formal and less centralised procedures used in the acquisition of land, but these are not necessarily less complex.

Encouraging investments is one of the objectives of the present land management system. Another, possibly more important but also potentially conflictive, is to protect indigenous smallholders from acquisition by a large-scale operator. Transforming village land into general land to facilitate large-scale investments, in agriculture as well as in other activities, has been perceived as a threat to smallholder farmers rather than an opportunity to complement their activities with large-scale investors. A further disempowering mechanism has been exhibited in the way the commons in communities under customary tenure are managed. The village communities have been particularly vulnerable, given the argument that commons are unowned or idle, or simply that they belong to the state. On the agricultural side, the issue is clearly that of the trade-off between modern high-productivity large-scale production process and traditional low-productivity farming and herding. No clear answer has been given to this socially hypersensitive question.

Alternative models of empowerment have been used that engage more of the existing producers, such as contract farming and out grower schemes. These do not displace smallholder farmers, but there are no institutional mechanisms to provide them with some capacity over price fixing. As a result, farmers often perceive that they are underpaid for the crops they supply to the large factories. On the other hand, the institutional framework for using smallholders’ land as equity for facilitating participation in large-scale investments is not functional. The fact that this has been functional in other countries such as China suggests that this option deserves further scrutiny to find how it could be an option in Tanzania.¹² The dichotomy between traditional and modern agriculture and the tensions owing to the absence of a well-functioning land market are likely to continue.

Institutional mandates for granting land rights and ruling on disposition and dispute settlement do exist, but these are overshadowed by a multiplicity of actors and entities that overlap, have no effective coordination or collaboration, and often lack administrative capacity. In effect, capacity building at the various governance levels is needed. In practice, land management and administration has been faced with challenges of efficiency and effectiveness as well as shortage of staff and rent-seeking.

The question of the balance between centralisation and decentralisation arises in the case of land in the form of dualism in the land disposition system: centralisation of control and management of general land on the one hand and devolution of control to customary law at the village level. The oft-cited cases of overlap of responsibilities and the complexity of the relationship between the various central and local public entities in the land management system are a sign of struggles for power over resources and over rent-seeking between these levels of government. Some of the conflicts and difficulties in managing land issues have in part been attributed to this dualism, which affects the economic efficiency of the whole land sector.

In this regard, Deininger provides useful insights into the alternative of providing space for village governments to integrate the investment activities of large-scale investors with village-level land use plans. He makes a case for complete issuance of CVLs – which, today, only a fraction of villages possess – using modern low-cost surveying technology, and proposes clarifying the content and status of village land use plans as a basis for attracting potential investors with a profile that would most effectively contribute to local development. This option would provide villages with the incentives to systematically identify investment opportunities that are consistent with their own land use plans. This provides a strong case for empowerment of local governments in land management for investors.

Weak land governance and property rights systems, including limited capacity to generate clear records, provide information on time, and process titles, lead to opaque land deals, and have contributed to facilitating corruption and undercutting responsible actors seeking access to land for productive investment. As very well shown by Mramba, the same is true of the land judiciary system, which does not have the capacity to deal rigorously with the numerous cases put to it. Nowadays, the ongoing awareness campaigns and computerisation through the Information Land Management Integrated System are expected to bring more transparency and minimise the opportunities of corruption. This is an opportunity that remains to be harnessed.

A finding worth emphasising in Tanzania is the complementarity between local governance in land matters and the impact of land titling, even through rather informal CCROs, on agricultural investment and productivity. Thus, Kassa (2018) shows that the effects of titling on agricultural investment is generally positive and sizeable, but the investment return is either negative or non-existent when there is a higher level of disapproval of the local governing units by plot owners. Likewise, Hombrados et al. (2015) find no significant effect of land titling once the characteristics of farm households are taken into account, and suggest that the capacity of local civil servants in charge of land issues and the enforcement of property rights may be determinant.

Over the last fifteen years or so, several major reports have been mandated by parliamentary commissions or the executive to evaluate the present legislative land occupation right system and suggest possible reforms. As of 2018, no legislative change had been made to the 1999 Land Act.

5 The Power Sector: An Example of De Facto Weak Regulation

The power sector has long been singled out as a major obstacle to development in Tanzania. In her in-depth analysis of the institutional evolution of the Tanzanian power sector (Chapter 8), Catrina Godinho addresses this issue. She considers the underlying political economy factors that have made it so difficult for Tanzania to move from an institutional equilibrium that does not bring about the desired sector outcomes in terms of power supply (investment, system expansion, and technical performance) to one that does.

Notwithstanding attempted institutional restructuring and ambitious reform policies, the structure of the power sector continues to most resemble that of the traditional industry model, which is a state-owned, vertically integrated (generation, transmission, and distribution) monopoly, regulated by the government (i.e. not an independent regulator).¹³ This model has largely been dysfunctional in Tanzania and across the region for decades. Yet TANESCO the state-owned electricity company, continues to operate as a vertically integrated utility with considerable monopoly power. The (difficult) introduction of private sector participation in generation and the establishment of an independent regulator, the Energy and Water Utilities Regulatory Authority (EWURA), has not changed the situation substantially. Electricity supply in Tanzania is persistently unstable, often inadequate, and insecure. Owing to the high coping costs associated with drought periods and political pressure around tariff increases, the sector’s financial standing has remained precarious. Tariffs remain below cost recovery, planning has not translated into timely initiation of procurement of new power generation capacity, procurement has mostly not been transparent or competitive, and TANESCO’s technical and financial performance is poor.

Political interference in the sector and its institutions has been a constant feature of Tanzania’s power sector. For example, the president’s obstruction of regulator-approved tariff increases in 2016, followed by the dismissal of several executives at TANESCO, and finally the removal of EWURA’s Director General all constitute rather problematic interventions. These and similar decisions on the part of the executive seem to bely a strong short-run bias considering that, in the long run, suppressed tariffs and unstable governance reduce the capacity of TANESCO to invest in sector maintenance and expansion, discourage private investment in the sector, and increase the cost of doing business. While these decisions are often justified by protecting the (urban) poor, low electricity access rates and risk-inflated costs of supply do not benefit those who need modern energy services the most; nor do they benefit businesses in general, who frequently cite poor electricity services as a primary development constraint; nor do they contribute to the reduction of poverty through economic growth.

There have also been a number of high-level corruption scandals. These were related to deals made with independent power producers to enlarge the power generation capacity of the country or to respond to emergency needs. The Independent Power Tanzania Ltd and Richmond deals resulted in much delayed and absurdly overpriced power supply, which led to huge financial losses at TANESCO. Those deals were authorised by the government, which suggests that high-rank executives were involved. A prime minister had to resign in 2008 after the Richmond scandal broke out.

The problematic experience with corruption in both public and private deals underscores the vulnerability of sector governance to political interference in a country where the political economy system is prone to using publicly owned companies and their links to the private sector for rent-seeking and patronage, even where corporatisation efforts attempt to ring-fence such interests. However, much of the blame has been laid at the feet of the private sector, strengthening the ideological bias against private participation in the sector – especially against private energy providers that could function as least-cost suppliers to a publicly owned system operator. These outcomes reflect a lack of capacity in the government and/or political will in the executive to ensure transparent and accountable governance in the sector. This is confirmed by recent moves against the independent regulator, which had emerged as an island of excellence in terms of technical expertise in the region.

There are too few accountability mechanisms built into existing legislation, specifically the 2008 Electricity Act, that could really bring key directives into force. This leaves the political executive with considerable power over policy implementation. Reforms have been resisted or at best implemented only partially, a situation that has contributed to checking (keeping low) levels of transparency and accountability. While potentially an improvement, the ESIRSR 2014–25 roadmap to reform has not been suitably translated into robust legislation. In effect, the status quo is maintained through institutional rules that include vertical appointment structures, archaic guidelines on sector information and lines of communication, and the frequent application of de facto power by political actors. In his discussion, Antonio Estache summarises well the state of play by stressing that the partial or biased implementation of past reforms, the repeated scandals about the attribution of private power deals, and their limited efficacy and ill-managed tariff policy have together made future reforms essentially ‘non-credible’. Under these conditions, it will be difficult to attract private investors in order to expand the capacity of the sector to the extent needed by industrialisation and development.

Interestingly, Tanzania has emerged as a leader in the region when it comes to decentralised options – primarily because of EWURA’s adroit regulation and lower levels of political interest or rent-seeking opportunities in smaller power projects. Urban and rural access rates have significantly increased in recent years. After a slow start, the Rural Energy Agency (REA) – established in 2005 – has begun functioning satisfactorily and rural electrification progress has considerably accelerated over the past five years, even though Estache suggests that it could move faster with support from the government and development partners. This is an interesting case, which highlights the benefits of effective regulation. However, the fact that it is sometimes perceived as a threat to the monopoly power of TANESCO and the special relationship between the executive and the utility could become a challenge. Even as it stands, REA hands over much of the distribution infrastructure to TANESCO after construction – which has placed further financial strain on the utility, but more power in the hands of those who control it. There have also been some allegations of corruption, as government funding has flowed to REA at unprecedented levels in recent years.

From a political economy perspective, Godinho argues that the situation of the power sector may be interpreted as an institutional equilibrium that might be conceptualised as the relatively stable and mutually reinforcing structural relationships that exist between public organisations, private interests, economic beliefs, and political power. The energy requirements of development will continue to rise rapidly as industrialisation takes root. Increased investment in the energy sector will prove determinative, yet institutional constraints remain immense, and the public sector is not in a position to finance the state-led model that is championed politically. Hence, there is a need to rely on the private sector without the guarantee that this will be politically acceptable, unless the executive moves toward greater transparency, accountability, and independence of sectoral institutions. The Suluhu administration has recently come with signs of more private sector tolerance, as indicated by the composition of the new board of TANESCO, which is dominated by members with a private sector background. This suggests that the type of interference witnessed in the Magufuli administration might be reversed by the new administration.

E Conclusion of Part 1: Five Basic Institutional Weaknesses

In trying to classify the most often cited institutional weaknesses throughout the various chapters of this volume and the preceding summary synthesis, the following general headings become clear:

• ill-defined structure of public decision-making;

• selective distrust of market mechanisms and private sector;

• under-performing civil service;

• rent-seeking and corruption; and

• patronage and weak business regulation.

These themes were not only dominant in the in-depth studies of particular aspects of the Tanzanian economy but, rather remarkably, in the more general approach to institutions and development adopted in the first chapters of this volume. They indeed fit well the economic diagnostic that pointed, inter alia, to the lack of dynamism of the tradable sectors, excluding natural resources: agriculture being handicapped by an inefficient and partly ineffective land rights allocation system, and industrialisation being slowed down by a badly managed power sector and a weak regulation of the business sector. The slow development of these sectors and the poor delivery of public services such as education and skills development also explain the lack of inclusiveness of recent economic growth. The five institutional weaknesses noted here also fit the conclusions obtained from the analysis of the international databases on governance indicators as well as the results from our own survey of opinions among Tanzanian decision makers. There, the dominant themes were corruption, government ineffectiveness, and regulatory quality (or the investment climate). The general picture obtained from these diverse approaches to the relationship between institutions and development in Tanzania is thus consistent.

It is by choice that the preceding taxonomy has been limited to five items. Some of them could easily be broken down into autonomous sub-themes. For instance, the ill-defined structure of public decision-making comprises an often-mentioned overlap of responsibilities across different units of the public administration, but it also includes a centralisation bias of decision-making by which the central power may discretionarily intervene at all levels in the public sector, imposing decisions or changing the decisions taken at a lower administrative level. This may be considered as a special case of overlap of responsibilities with absolute priority given to central power’s objectives and rationale. In situations of asymmetric information and objectives between the executive and the decentralised administrative unit, this is bound to lead on occasion to economically inefficient decisions.

Another sub-theme under the ‘ill-defined structure of public decision-making’ heading is the gap between the passing of laws and reforms, or the recommendations emanating from publicly appointed expert commissions and their entry into force. A related issue is the succession of reforms in the same area. Examples of such situations were seen in relation to the management of the power sector, the reform of the civil service and the decentralisation law, multiple reports on the dysfunction of the land laws and the decentralisation law, with a particular situation in the latter case being a law that had not been fully brought to force, and parliamentary reports. These occasional gaps in the implementation of the law are also related to the centralisation bias, as they clearly emanate from the will of the executive, delaying the application of democratically grounded decisions.

It should be clear from this example that the identified institutional weaknesses in the list are generic in the sense that they may include deficiencies of a different nature. They will be analysed in more detail later. Before doing so, however, it must be emphasised that, although ubiquitous in the functioning of the Tanzanian economy, the five weaknesses are only the symptoms or the consequences of institutions that are working in a way that is not as favourable to long-run inclusive development as it could be. As such, they do not say much about the underlying causes of this unsatisfactory functioning, even though some of these causes have been mentioned on several occasions throughout the preceding chapters. It is now time to regroup them into a consistent framework to help us to understand the origins of institutional challenges for the development of Tanzania, and how to address them.

II The Institutional Diagnostic

The preceding list of institutional weaknesses corresponds to the symptoms of institutional dysfunctions. A diagnostic must go beyond the symptoms to uncover the causes behind them and envisage possible remedies. This is what is done in the second part of this chapter.

A Institutional Weaknesses, Proximate Causes, and Deep Factors: A Conceptual Framework

Table 9.1, presumptuously termed ‘The diagnostic table’, tries to relate the basic institutional weaknesses identified in this study with their general consequences on economic development on the one hand, and what may be their proximate causes on the other. These causes, which are amenable to changes through policies and reforms are then related to ‘deep factors’, which may be responsible for whether those policies and reforms could be undertaken or not.

Table 9.1 The diagnostic table

Deep factors		Proximate causes		Basic institutional weaknesses		Economic consequences
Political game (joint structure of political and economic power)		Under-capacity of State apparatus (skills, resources, technology)		Ill-defined structure of public decision-making (Responsibility overlap, centralisation bias, law implementation gap)		Ambiguous growth engine and long-run growth perspectives (slow productivity gain, slow industrialisation)
Social structure (ethnicity, rural-urban, education…) Donors’ views and leverage	⇒	Misaligned incentives (high and low levels of civil service) Imprecision and complexity of the law	⇒	Under-performing civil service	⇒	Poor investment climate
Historical legacy (Ideology, government credibility, trust, …)		Type of leadership (transparency and accountability) (state–business relationship)		Rent-seeking and corruption Patronage and weak business regulation Selective distrust of market mechanisms (Power sector, Land)		Low quality public services Regressive distribution and redistribution Occasionally inefficient public decision-making (e.g. power sector, local communities, land investments)

In examining Table 9.1, it is quite important to realise that there is no one-to-one relationship between the elements of the various columns. One institutional weakness does not have a unique general consequence and has more than a unique proximate cause. The relationship between the four columns is essentially multivariate. The important point is essentially the chain of causality. The whole set of institutional weaknesses is the consequence of the whole set of proximate causes, which depend themselves on the whole set of deep factors, even though, in both cases, the relationships may be stronger between particular items. On the other side, the set of institutional weaknesses affects how the economy works.

The relationships described here are complex and need some explanation. The right-hand side is fairly clear. Indeed, it has been seen throughout this volume that identified institutional weaknesses played a determinant role in affecting economic policies and economic outcomes in Tanzania and ultimately its long-run development path. Only some general implications of these weaknesses are mentioned in Table 9.1, but more detailed consequences could have been picked from previous chapters.

The items that appear under the ‘proximate causes’ column of Table 9.1 need more explanation. Considering institutions as a set of rules imposing tasks or behavioural constraints on the various economic actors, why would the latter not follow the rules? A first possibility is that they do not have the skills or knowledge to perform these tasks or to manage the constraints they are subject to. In the same vein, they may not have the equipment or resources to perform their tasks or deal with the constraints. In short, laws, administrative rules, and tasks may not be fully adapted to the level of development of a country. This lack of administrative capacity has been pointed to repeatedly in the preceding chapters, in particular in connection with the management of land rights and the civil service in general. It has also been emphasised both in the opinion survey and the analysis of governance indicators completed in the first part of this volume.

A second cause of the identified institutional weaknesses is the misalignment of incentives for economic and civil actors to fulfil the task asked of them and/or comply with the various institutional rules they are subject to. Incentives may be of the carrot or the stick type, rewarding those who behave well in the first case and punishing those who do not in the second. The main point here is that rewarding and punishing require that the evaluation of behaviour be properly conducted, with the appropriate skill and economic resources. It also requires that evaluators be themselves monitored to prevent transgressors and evaluators colluding.

Transgressing rules may also be unavoidable when rules are ill-adapted, inconsistent, or too complex. Several identified institutional weaknesses in Tanzania were thus due to ambiguities, imprecisions, or too much complexity of the law and administrative rules. We shall return to this point in more detail, but we may note here that this issue was found to be highly relevant in the case of the Land Act and recognised as such in the multiple official reviews of that piece of legislation.

The final set of proximate causes of institutional weaknesses may also come from the type of leadership, or more precisely the nature of the government, the strength of the supervision it exerts on the public administration, its transparency, and its accountability. The underperformance of the civil service, its rent-seeking behaviour, or major corruption scandals involving high-ranking political personnel may be the result of lax supervision and a weak obligation to accountability by the executive. Even when evaluation entities do exist, an example being the Comptroller Auditor General in Tanzania, they may be under the thumb of the executive or may deliberately not be given the resources they need to perform their task. An authoritarian leadership is not necessarily efficient if authority is used inappropriately, as may be the case with the centralisation bias mentioned earlier.

Of special importance is the nature of the relationship between the state and the business sector, as repeatedly emphasised in previous chapters. Does the government have the power to decide and implement policies that may constrain the private sector? Or does the latter enjoy an effective leverage on the government, through the strategic importance of some firms for the whole economy, so that their own interest prevails over that of the community? The absence of a well-articulated industrial policy suggests that the second situation is more likely in Tanzania, but this may not be true of all activities, and bargaining may take place leading to intermediate solutions. Of course, the possibility also exists that big business bribes government members or influential people within the dominant CCM party for personal enrichment or political survival.

It remains to be seen more precisely how these proximate causes apply and how they combine in the case of Tanzania to produce the identified institutional weaknesses, and which policy or reform could remedy them. Before doing so, however, a word must be said about the ‘deep factors’ column of Table 9.1. Deep factors have to do with the feasibility of policies that could remedy the proximate causes of institutional weaknesses.

The political game, that is to say the process by which a policy is actually put to work or stopped, is ultimately the main factor that makes a reform aimed at correcting institutional weaknesses possible or not. In this respect, the de facto structure of political power is essential. But, because some reforms also need the agreement or cooperation of non-political actors as well, for example big business, it is the joint distribution of political and economic power within society that matters. In the case of Tanzania, the presence of a dominant party, the CCM, tends to enhance the decision power of the government. Yet factions within the party and their links with economic interests may well hinder particular action and reforms intended by the executive.¹⁴ For instance, the long and continuing indeterminacy about the private or public status of power generation and distribution resulted from the inability to align interests of political and economic powers along one option. In several instances, the executive’s will to sue high-rank politicians for corruption has been stopped in ways that have never been elucidated, or at least made public. Likewise, industrial development policies have remained undecisive and opaque because of power conflicts between business-linked political powers within the dominant party.

To be sure, the joint structure of political and economic power may change over time. This certainly happened at the time of the transition back from socialism to a market economy and more recently under the Magufuli administration, whose anti-corruption goal may have led to an unexpectedly strong authoritarian shift in the exercise of power. Further changes may now be expected with the Suluhu administration.

The social structure of society may also be an obstacle to institutional reforms. For instance, the demographic importance of peasants in the population might block reforms that would affect their autonomy because this would substantially reduce the electoral support of the party in power. Such factors combine with the political game to make some reforms possible or impossible. In the case of Tanzania, this remark certainly applies to the difficulty of reforming land laws as well as the implicit struggle between local and central governments in tax collection.

The historical legacy may also create a context that affects the capacity and the effectiveness of institutional reforms. The role of ideology and the attachment to the socialist past that leads to some distrust of market mechanisms and the private sector in some areas is a clear example of such a legacy. The trust that citizens may have that political personnel and the public administration can effectively implement a reform may also be an important factor for facilitating or, in its absence, for blocking it or making it inoperative. There is indeed an issue of government credibility. Reforms that have repeatedly been promised but unsuccessfully implemented become non-credible and become unable to change the behaviour of people and firms. Something of this type may have happened in Tanzania with the fight against corruption. Things may have changed under President Magufuli, but it is still too early to know whether this has really been the case.

Donors are also mentioned among the deep factors, essentially because their contribution to the public budget and therefore their leverage on Tanzania’s government is sizeable. They reacted to several major corruption scandals by suspending aid disbursements, and they also directly imposed their own views on institutional reforms, as was seen in the case of decentralisation, and of course at the time of the shift back to the market economy in the 1980s.

Deep factors logically appear as the ultimate causes of basic institutional weaknesses. But, as mentioned earlier, they may change over time, so that reforms that were not possible may become possible. Because of this, an institutional diagnostic needs to be explicit about how the proximate causes for institutional weaknesses may be reformed, and on the gains and losses of all actors who would be involved in this process, so that all are aware of the consequences of exerting the power at their disposal.

This is what the next sections of this chapter intend to do.

B The Reform Potential for Correcting Tanzania’s Institutional Weaknesses

The preceding conceptual framework for drawing an institutional diagnostic of Tanzania is now applied to the generic institutional weaknesses identified in the first part of this chapter. For each generic weakness in turn, available evidence is briefly summarised, before reflecting on their proximate causes, possible directions of reform, and potential political economy obstacles. These various considerations aim to show the potential for mitigating or correcting identified institutional weaknesses, or, in other words, what would need to be done to reduce them and what would prevent this from happening.

1 More Efficiency in Public Decision-Making

In any economy, the structure of public decision-making is complex. Efficiency requires that, for most operations, information gathering, analysis, decision-making, and implementation occur in a precise place in the functional structure of the public sector. The thematic studies have shown that Tanzania’s decision-making system is set in such a way that it is not always clear who makes decisions and that the number of procedures needed to reach a decision may be burdensome. Law simplification and skill enhancement could be one way to straighten public decision-making.

Land transactions are a good example of this. In practice, investors must deal with many different actors before a final decision is made, from village councillors to district commissioners, to bureaucrats in the Land Commissioner’s office in the Ministry of Agriculture, not to mention several technical services. This complex procedure is likely to disincentivise investors, as already observed on some occasions – see Chapter 7.

There are many examples of lack of coordination and overlapping responsibilities among administrative units in charge of closely related operations. This is the case, for instance, for the National Land Use Planning Commission and the Director of Urban and Rural Planning in the Ministry of Agriculture, and in the absence of a clear line of communication between the Tanzania Revenue Authority, a public agency responsible for tax collection, and the Ministry of Finance, responsible for tax policy. Decentralisation raises the same kind of coordination challenge, with some confusion of mandates between local government executives and central government appointed officers advising on or controlling local decisions, and with an intricate set of relationships between LGAs, the prime minister’s office, the Minister of Finance and various sectoral ministers. That a decentralisation reform approved by parliament in 1996, and supposed to be entering into full force by 2008, is not fully implemented more than ten years later is symptomatic of the shortcomings in the structure of public decision-making. Together with the sudden decision by the executive to recentralise the management of local taxes in 2016, this is an example of both the inherent difficulty of coordinating various levels of government and a centralisation bias.

This gap between the passing of laws and effectively implementing them, or the ineffectiveness of government-commissioned expert reports on the land laws, also reveals not only a centralisation bias, but also governing dysfunctions. No fewer than six reviews and reports on the land laws have been produced since 2004.¹⁵ Yet little progress was achieved, and there are still serious difficulties in implementing the law.

Another telling example of the lack of coordination and centralisation bias can be found in the power sector. After decades of delay, a clear decision-making structure has been put into place with well-defined functions for the state-owned company TANESCO and the regulatory authority EWURA, whose technical competence is widely praised. Yet the central authority can veto decisions taken within that structure, as was the case during the Magufuli administration when a tariff rise was cancelled by central government and the managers of the two entities were simply demoted, their appointment, and demotion, being by law under the sole competence of the president. The argument put forward was that such a rise would have increased poverty, whereas not raising tariffs would have meant loss to TANESCO and a need for subsidy in order to operate without accumulating debts. Whatever the argument for or against the rise, such an event suggests either that the regulatory authority ignored some key social welfare aspects of its decision, which means something is wrong in the composition of the group of experts making tariff recommendations, or that it is indeed the expression of a very centralised power possibly acting under political motives. The same can be said about procurement decisions in the power sector. The multiple past corruption affairs linked to the choice of private power providers, which ended up both ineffective and exorbitantly expensive, and involving top government members are clear evidence that, even though well-organised on paper, the decision structure can be short-cut at a high level of the state.

The proximate causes for these institutional failures differ according to the issue being considered. In the cases of both land rights management and decentralisation, there is little doubt that the complexity of land laws and the ambiguity of the laws that rule the relationship between LGAs and the central state play an important role. An obvious direction of reform is therefore the simplification and the rationalisation of these laws and administrative rules. The lack of adequate skills, particularly at the local level, is another handicap to the efficient application of the law, which calls for serious efforts in recruitment and training.

Of course, some people gain from the complexity of some laws and the administrative organisation as they generate multiple rent-seeking opportunities. This is the reason reforms as simple as the previous ones – that is, legal simplification and skill enhancement – are not necessarily consensual. It has been seen that repeated changes in task allocation between local and central governments reflected a struggle between local elites and the dominant central party in capturing local rents. The involvement of the state in major decisions concerning the power sector may also be analysed as too much de facto power to the executive, and also in terms of hidden personal interest by powerful politicians who short-circuit decision making by the power company and the regulatory authority, for instance in procurement decisions. Finally, the delay observed in implementing laws such as the decentralisation laws or the limited responsiveness of various administrations to reform proposals of the land law is likely to reflect more a lack of consensus and possibly incompatible views because of conflicting interests by powerful political actors, rather than a practical difficulty in implementing reforms.

Reforms that would enhance state capacity may not be consensual in the short and medium term. Yet it bears emphasis that, in the long run, they will be beneficial to whoever is in power.

2 Overcoming Distrust of Market Mechanisms and the Private Sector

Distrust of market mechanisms and the private sector was observed in two areas in the course of this study: the allocation of land rights and the attempts at restructuring the power sector. In both cases, it was shown to have a heavy economic cost.

As in many other sub-Saharan African countries, the Tanzanian government faces a dilemma in the agricultural sector. On the one hand, it wants to foster the development of the sector with modern technology applied on large-scale farms and productivity gains among smallholders. On the other hand, it wants to protect traditional farmers from being displaced by private investors. This objective is all the more important since traditional farming areas are where most poverty is found. Unsettling that sector could thus have adverse consequences on poverty, as it did in the 1970s during the attempt at collectivising land, leading virtually to famine.

Based on the principle of public ownership of land, the Tanzanian land laws intend to combine these two objectives by distinguishing between village land for traditional farming and general land for modern productive uses of land, the former being governed by customary laws and the latter by statutory law. The difficulty is in the transformation of village land into general land for investment purposes.

Even though they were designed long after the end of the socialist planning period, the land laws very much retain the feel of a planning system. It seems as if it was attempted to move land management away from a market system as much as possible. However, there is no contradiction between the public ownership of land and the functioning of a land market if what is priced and exchanged is not the ownership but the right to use land, that is through long-term leases. To be sure, market operations for land rights take place today both in urban and rural areas, but the key operation that enables investors, or even the state, to access village land remains cumbersome, long-winded, sometimes uncertain, and possibly costly because of the multiple opportunities for rent-seeking by land officers at various levels of the hierarchy. The community’s wish to protect traditional farmers from being expelled from their land by commercial investors or speculators has led to a labyrinthine administrative system that is considered by many observers to be a major obstacle to development.

The current system has also led to some a priori antagonising of village communities with respect to agricultural investors because of the complexity and the frequent ineffectiveness of many of the deals between them, most often engineered by public intermediaries involving bribes, or sometimes signed by corrupt village leaders who are acting without consulting their villagers. Compensation packages that involve promises of employing villagers or building village infrastructure, such as schools or roads, are indeed more difficult to enforce than a simple transfer of resources to village communities.

The proximate causes for the inefficiency of the land allocation process lie in the complexity of the land laws, ineffective public operators, particularly at the local level, and weak control of corruption. To this must be added the incomplete mapping of the land, which makes the task of processing and recording land operations particularly difficult. The important point is that both the lack of capacity and the rent-seeking behaviour are themselves to a large extent brought about by the complexity of the law and the extremely slow mapping process. Improving things thus involves correcting those two weaknesses.

Simplifying the complex administrative procedure to transfer land from villages to investors might work through more reliance on market mechanisms, such that village communities and investors, rather than numerous public intermediaries, would be the main actors of the transactions. They would have to agree on the size of the land to be transferred, its location, and the price and nature of the compensation – including through shares in the investment venture – to the community, with the village being fully responsible for managing the reallocation of land that might be needed among villagers. Full publicity of transactions conducted elsewhere in the region or in the country, as well as of the consequences of the corresponding deals among smallholders, would be absolutely necessary to guarantee full information and market transparency. National interest would be taken into account only in the very final steps of the procedure, with some veto power in the hands of the government in case the deal related to very large swaths of land and was thought to be harmful for the national or regional economy.

Information on transactions and their consequences for village communities is essential. Knowing the price at which transactions took place in the region is a guarantee that no side will feel cheated if an agreement is reached. More importantly for the villagers, knowing whether smallholders elsewhere were ultimately harmed or benefited from the deal, possibly through the public goods the village could acquire thanks to the deal, is key to avoid social disasters or to obtain the right compensation.

Arrangements other than cash transaction for the unconstrained use of the land are sometimes made with investors: contract farming, smallholders acquiring equity in the investor’s venture and continuing to work their land, compensation in kind, and so on. Except maybe for contract farming, suitable only in specific cases, all these arrangements are bound to result in conflict at some stage, either at the time of sharing profits or when deciding on payment to labourers. Previous experience is not encouraging. Such contracts are simply too complex and cannot take into account all contingencies. However, considering that forms of equity arrangements have been working elsewhere (e.g. China), there is a case for exploring further possibilities. In the meantime, cash deals between an investor and a village, with redistribution or the acquisition of farm-productivity enhancing public goods being left to the local authority, could be promoted, as they are more transparent and make village authorities fully accountable.

Of course, for such a market-like solution to be possible, it is necessary that the whole land has been mapped so that precise limits to the land being transferred can be established. This requires resources to be made available for that purpose, although there seems to be some disagreement on the actual cost of this operation. Within villages, land management could remain as it is, with some possible land reallocation under the control of the village council in case the land transferred to investors impinges on the farms of smallholders. The titling process (CCROs) would keep developing, accelerating the extension of within-village and across-village transactions.

The preceding discussion refers to land transfers from small scale, near subsistence farmers in traditional rural communities and large-scale domestic or foreign commercial investors. But land consolidation is also needed within communities to increase yields. Progress has recently been made through the extension of medium-scale farms – 5 to 10 hectares – in some regions where land is still abundant. This has resulted from informal transactions that have led to the regrouping of small plots, with these transactions generating economies of scale and creating positive spillovers.¹⁶ Such a spontaneous restructuring or reform of village land and accompanying productivity gains would be quicker and stronger if market mechanisms were encouraged and made easier at the local level.

A larger role for market mechanisms in land right transactions should strengthen the potential and incentives for investors, medium-scale farmers, smallholders, and village communities to exploit the clear comparative advantage of Tanzania in agriculture and agro-industry, and reforms in that direction would probably meet some resistance. Civil servants and politicians, who are currently able to extract economic or political rents from the present unclear decision-making structure, would oppose them, as would possibly peasants and village communities, given their attachment to tradition. The structure of political power matters here. To the extent that village communities would retain some control of land transactions on their territory, they could probably be persuaded that these reforms were in their favour.

The proximate causes for a distrust towards the market and private actors in the power sector are of a different nature. Historical legacies matter, but so do the rent-seeking opportunities for politicians who are close to the central power to control the operations of a big public company, most importantly the procurement process.

History matters here because of the socialist past of Tanzania and the primacy of public companies in the socialist era, especially in the delivery of such a critical good as electricity. But history matters also because of several huge corruption scandals linked to subcontracting power generation to private firms. Keeping the electricity company as a vertical public monopoly is seen as a kind of guarantee against such corruption cases. But, of course, doing so also provides rent-seeking opportunities to influential politicians in all procurement operations made by the public company, including with private providers when the capacity of the company is overly constrained – as has been the case in the past.

Correcting the situation requires actions on two fronts. The first of these is re-establishing credibility in the procurement process. This means better supervision of procurement operations, but more importantly a composition of the contract-awarding commissions guaranteeing its objectivity. Allowing one or two reputable foreign experts to sit on this kind of commission may be a way of getting closer to that objective. Second, unbundling the public electricity company into power generation and distribution and then enhancing the current limited private providers to sell power to the public company would mean competition, and therefore more transparency and fewer rent-seeking opportunities in power provision. After all, this kind of reform was repeatedly recommended by expert reports and explicitly considered in the Electricity Act passed in 2008 that is presumably still in force today.

3 Enhancing the Civil Service’s Performance

In one of the first evaluations of governance in Tanzania in the mid-1990s, the civil service was found to be one of the key weaknesses of the public sector.¹⁷ Things have changed enormously since then. The skill of civil servants has substantially increased, and several reforms of the civil service have been undertaken, introducing a more rigorous recruitment process and a performance-based career management system. Yet the draft of a new wave of reforms explicitly acknowledges various problems faced by the public service, including staff demotivation and indiscipline, inadequate accountability, widespread corruption, underskilled staff, lack of meritocratic incentives, and weak and passive leadership.

Without anticipating the actual measures to be included in this new reform, improvements are certainly still needed in the civil service. Judging from the absenteeism among teachers and, to a lesser extent, health personnel, there is no reason for productivity to be much higher in other occupations in the government sector. Presumably, supervision is not stronger than in education, so that shirking is probably as frequent as teachers’ absenteeism. The estimate of the aggregate loss owing to absenteeism and shirking is high – roughly 2.5 per cent of GDP, not including the weak performance of schoolchildren or the low quality of other public services.

Besides the organisational inefficiencies analysed earlier in the public sector, an important proximate cause for this state of affairs has to do with incentives. In this respect, a first possibility is that salaries are insufficient to compensate for the efforts and time that teachers would spend in school if they were indeed there full time. Teachers and, by extension, other civil servants tend to shirk, moonlight, or rent-seek so that their standard of living is equivalent to what they could get with another (private) employer. On average, however, it has been seen that, other things being equal, salaries in the public sector are substantially higher than in the private sector, even though this may be more the case for low- rather than high-skilled workers.¹⁸ In short, some salary adjustment may be needed in some parts of the civil service – for instance, for graduate secondary school teachers or to attract candidates in maths and science – but, overall, the compensation of civil servants does not seem to be a major disincentive to effort. The same seems to be true of career prospects, which have supposedly been improved in the last programme of public service reform.

The alternative cause for weak incentives could be the ineffectiveness of the performance management system (PMS) and monitoring and evaluation (M&E) function. Shirking may be easier in the public than in the private sector because not enough resources are allocated to PMS and M&E activity in the former and the supervision structure of the civil service may be poorly organised. It may also be the case that PMS and M&E agents and supervisors are shirking themselves, or tend to side with the public-sector employees they are supposed to evaluate because of the public sector status they share with them. In both cases, they are not performing the task entrusted to them.

There is a huge body of literature on how to incentivise teachers.¹⁹ Experiments in this area have been conducted in many countries. In Tanzania, the need to improve the performance of the education system has been duly recognised, and much reflection and experimentation has been undertaken.²⁰ Reforms in that sector could take advantage of that knowledge in the educational sector. See the recent survey of this area by Dal Bo and Finan (2020).

Less is known about incentivising other civil service jobs. To the extent that supervision is closer than in the educational area and that both work intensity and outcomes can be more easily observed, a first step may be in the disciplining of public sector employees. Understanding why supervisors do not enforce better discipline on their subordinates and why PMS and M&E do not function better is thus essential for designing the measures most able to remedy that situation. As with other developing countries, Tanzanians have a long tradition of seeing a civil service job as a perk with essentially low expectations of the productivity of civil servants at all levels of the hierarchy. Formally, introducing performance-based remuneration as done in the recent reform of public service is undoubtedly going in the right direction, but is clearly bound not to work if performances are not properly evaluated because of evaluators’ lack of capacity or motivation.

Other proximate causes for the underperformance of the public sector are the lack of resources and the insufficient skill of civil servants in some crucial areas. Here, again, the education sector is a good example. The performance of the primary education system is low not only because of the scarcity of teachers and increased school enrolment – of course this is huge progress, in part brought about by the abolition of tuition fees – but also because of teacher absenteeism and because their training is recognised as inadequate. Things are possibly even worse now that basic teacher training has been reduced to a single year in teacher training colleges with a second ‘hands on’ year in school, with limited supervision. The situation may be particularly serious in teaching because of the need to increase the number of teachers to comply with a targeted student–teacher ratio at a time of fast-growing school enrolment. Indeed, a large part of the very fast increase in the size of the civil service over recent years has taken place because of a massive recruitment of teachers, and one may thus understand that the quality of recruitment, including teacher training, has gone down.²¹ Recruitment in other public activities has also been fast, with possibly the same effect on quality.

Remedies for these causes of the lack of effectiveness of civil service are obvious: the compensation system has to be re-examined to make sure that it is adapted to the reality of the labour market, especially in areas where the supply side seems short; PMS and M&E must be scaled up and made more effective at all levels, in particular by making sure there is no collusion between those who evaluate and those who are evaluated; and recruitment and training must be improved.

It is unlikely that such measures would meet any real opposition within the Tanzanian political elite, except maybe from workers’ unions complaining of excessive supervision and therefore workload. The main constraint here is resources. The Tanzanian budget is already extremely tight and relies extensively on foreign aid and foreign loans. Improving the quality of the civil service may thus require difficult trade-offs within the public budget. Of course, this does not lessen the need for more discipline.

4 Addressing Rent-Seeking and Corruption

Rent-seeking activity and corruption may be the clearest symptoms of weak-performing institutions. They arise from situations where a public sector agent with some autonomous decision-making power grants to a private sector agent an advantage in a way that is openly illegal or bypasses official operational rules. The cost of such behaviour is twofold. On the one hand, it biases the economic allocation process in comparison with what the institutional rules are expected to produce, which may not necessarily be economically efficient. On the other hand, it redistributes resources in a way that may be regressive, as when poor people find themselves forced to pay bribes to access a supposedly free public service.

The proximate cause for rent-seeking and corruption lies first in the incentives for agents to adopt such a behaviour. As with absenteeism or shirking, it may be due to an inadequate compensation system and agents compensating for it through extracting rents from others, and to weak PMS and M&E. However, as was clear from the discussion on land, rent-seeking may also be the consequence of ill-conceived institutions that grant too much de facto unchecked decision power to potentially untrustworthy agents, lack transparency, or impose rules that are unduly restrictive for economic activity. In the latter case, it is even possible that corruption is economically efficient, in the sense that it improves the lot of the briber, of the one who pays the bribe, and of the whole community if it helps to cut red tape.²² Another cause of widespread corruption is the lack of transparency of public decision-making in connection with the private sector.

Yet there is also an externality dimension in rent-seeking and corruption that must be stressed because it conditions the way to fight it. If this behaviour were limited to a few public officers, it is likely that they would soon be confounded with enough discipline on the part of the executive, a policy that would probably be accepted by most political actors. The difficulty appears when this behaviour becomes so widespread that everybody may hold everybody else under the threat of being denounced, thus neutralising this basic anti-corruption tool. Those uncorrupted people paying a rent will not denounce their rent-seeker either, as long as the institutional constraint that makes a bribe mutually advantageous has not been corrected. A true culture of corruption may then develop, where everybody acts under the expectation that none of the people aware of the corrupt action will denounce participants. Informal bilateral rules substitute for formal collective rules. This practice is most widespread in political elections, where most voters and seekers of political positions participate in corruption activities. From that point of view, and despite Magufuli’s general anti-corruption action, it is believed that the level of corruption in political elections in 2020 was higher than that in previous elections. The main challenge lies in political commitment to fight corruption, considering that the route to political positions is not free from corruption.

Economic theory provides an interesting representation of this phenomenon under the form of a multiple equilibrium. There is an equilibrium where most agents are corrupt, asking for or paying a bribe.²³ For this to be the case, the probability of denunciation must be low, meaning there are few ‘clean’ people. In the other equilibrium, a large proportion of people are clean, which makes corruption dangerous because of the high probability of being caught.²⁴ Most observers tend to believe that Tanzania is in the former equilibrium, with corruption so widespread that it is indeed ‘a way of life’, as a journalist wrote in view of the repeated corruption scandals at the end of the Kikwete mandate.²⁵

Although very crude, this model shows the enormous difficulty in fighting corruption when it is widespread. This is because the preceding equilibria are ‘stable’, in the sense that small departures from the original equilibrium with a slight change in the probability of corrupt people to be sanctioned cannot prevent the system from going back to the original equilibrium. This is what has been observed in Tanzania with various past attempts at curbing corruption. Moving to the clean equilibrium requires a ‘big push’, with a sufficiently large group of people above any suspicion of corruption leading the fight against corruption.

There are few examples of countries that have successfully combated corruption, or at least reduced it to sporadic cases, starting from a culture of corruption. Singapore and Hong Kong may be the most famous cases. In both countries, a big corruption scandal moved public opinion. A politically powerful leader then put together an anti-corruption team made up of police personnel and judges who were above suspicion, whose mission was to sue corrupt civil servants and businesspeople with maximum publicity. It took time, but it worked. In the case of Singapore, interestingly enough, the strategy also included redesigning the law so as to eliminate situations that would provide rent-seeking opportunities, as well as making sure that civil servant salaries were well aligned with remunerations in the private sector.²⁶

There are many more examples of countries where anti-corruption policies and agencies have been unsuccessful or short-lived, including Tanzania. The main difficulty is forming the ‘above-suspicion’ anti-corruption core team, giving it all the resources it needs, politically supporting its decisions, and maintaining the fight for long enough.

Corruption seems to be sufficiently severe in Tanzania that such policies should be seriously considered. This was the position of the Magufuli administration, which launched an aggressive anti-corruption campaign based on a solid team around the president and on re-establishing discipline in the upper tiers of civil service. After some apparent early progress, doubts have been expressed about whether fundamental change had been obtained at the time he passed away, six years later. His noticeably rigged re-election in 2020 would seem to confirm those doubts.

It should not be forgotten that, as mentioned in the case of Singapore, simplifying the law, making the administrative apparatus more transparent, reducing the monopoly power of civil servants in particular decisions, and aligning their compensation with the private sector are an equally important side of the fight against corruption. Most importantly, such measures are much less likely to be resisted by those political forces that would be under threat in a direct attack against corruption.

5 Strengthening Business Regulation and Competition

The relationship between the government or the political elite and big companies hides another aspect of corruption and rent-seeking through economic favouritism. Practically, politicians in power, or with influence on the executive, offer advantages to a company, in exchange for their economic support. This is legally reprehensible in some cases, as with vote-buying, but economic support may take other forms, for instance through contributions to electoral campaigns, grants to regional development, or employment creation in particular localities.²⁷ Even though such practices are close to lobbying, the difference from straight corruption is small, especially when the political support granted by companies is hidden, as it is most often the case.

Because electoral support mostly comes from business, this patronage system is equivalent to the latter extracting a rent, under the form of a competitive advantage or otherwise, from the government or, more exactly, from citizens. They are able to do so for three major reasons. First, they have the resources to support the electoral campaigns of their political allies. Second, as economic growth relies to a large extent on big business, no government would take the risk of antagonising them. Third, major firms are often in monopolistic positions, which limit the leverage the government may have on them. Such a state of affairs is certainly not a guarantee of economic efficiency since the interest of big groups is not necessarily that of the nation. Pursuing an economic development strategy that would be collectively beneficial but would hurt the interest of some major firm would be a risky bet for governments, as the firm might retaliate by favouring rival political groups or, when in partial control of the media, campaigning against the government’s decisions. The institutional weakness here is the lack of independence of the government or public decision-making process with respect to the interests of major private companies and the opacity of that relationship. The joint structure of political and economic power is thus what determines policies that affect big business.

The Magufuli administration faced this problem, as its predecessors did. Shortly after taking power, President Magufuli stated that he would not let businesspeople ‘play with his government’. He then took vigorous action in that direction by fighting tax evasion and intervening in several corruption cases that involved big companies. Without really releasing the pressure, he nevertheless made clear a year later that he was open to signing agreements with them provided they contributed to the industrialisation agenda.²⁸

This was probably the right strategy by then. However, even though the relationship with the executive improved, big firms retain considerable power, including of course that of influencing the political game. As the dominant political party, CCM, is somewhat fragmented in various political factions; there has always been room for large companies to play one faction against another and potentially unsettle the executive. As seen earlier, this might change if the leadership were able to exert more control on the party.

The other effective way of reducing the power of big businesses is to weaken their monopoly position both as leading economic actors and in their areas of specialisation. This requires the fostering of competition on two levels. First, the emergence of new economic leaders should be favoured, which includes making sure that promising business managers will not be ostracised because of their ethnic origin or other personal characteristics. In other words, there should be no glass ceiling, most notably when raising funds, for entering the business elite other than talent. Second, it should be checked that incumbent firms are not restricting competition in their area of activity by raising barriers of entry, including through artificial business regulation obtained from the government. In short, this would mean empowering the Fair Competition Commission and sectoral regulatory authorities, for example the Sugar Board,²⁹ so that they are more proactive and more rigorous, this being particularly important in the financial sector. Foreign investment and trade policy, in particular the dismantling of various types of non-tariff barriers since base tariffs are set at the level of the East African Community, should also be used to check domestic monopoly power and foster the development of new activities. At the same time, trade and tax authorities should be encouraged to tighten the control of counterfeit imports and the levy of legal tariffs. Transparency of the long-run industrial policy pursued by the government is also a necessary condition for resisting the political influence of local big business.

In relation to big domestic companies and the design of industrialisation strategies, the development of South Korea bears some lessons. Although industrial development was based on big companies, the chaebols, the Korean government managed not to become their hostage, at least during the years during which development took off, and to develop its own strategy. This was done by creating competition between chaebols, especially in export markets where they would not be constrained by domestic demand, never granting support to one group that would not be granted to at least one other and, most importantly, making those grants conditional on successfully meeting precontracted objectives.³⁰

Promoting competition among leading companies in Tanzania would not bear fruit in the short run, and it may well be the case that it would be partly neutralised through their political leverage. Coupled with an authoritarian stance on corruption and an adequate communication policy on the government’s industrialisation strategy, it could nevertheless progressively improve the competitive landscape, restrict the de facto political power of the business elite, and allow for more effective and more open collaboration between the government and business.

C The Need for More Transparency and Accountability of Central Government

Summarising the preceding sections, the proximate causes for the institutional weaknesses identified earlier in this volume are clear. They can be divided into two groups. The first comprises three basic causes for economic actors not behaving according to institutional rules: lack of capacity or skills to apply the rules, weak incentives not to deviate, and rules that are too complex together with an ill-adapted administrative organisation. Several types of remedies have been mentioned, depending on the institutional weakness being considered. Some of them essentially require more resources and capacity building among civil servants in various positions. Implementing them should not raise any opposition. The same is true of the cases where the compensation of public officers is thought to be inadequate, this being an incentive for absenteeism and rent-seeking. Another type of remedy is to enhance the PMS and M&E function in public service in terms of frequency and effectiveness. This is most likely to raise some resistance from unions and to reinforce political opposition to the government. Finally, top civil servants would oppose administrative reorganisation that would terminate their source of rent. In all these cases, however, central power should be able to impose those remedial policies. Political will is what is needed. The Magufuli administration took measures that went in that direction, even though these were not ambitious enough.

The second group of causes is much more difficult to resolve. Hidden political opposition is likely to be strong when attacking grand corruption, a key institutional weakness now deeply rooted in Tanzanian society, or trying to regulate big business to enhance its contribution to development. Institutional improvement in these areas touches upon deep political factors. The success of such reforms depends on the political game and to a large extent on another deep factor that has been little alluded to before, the trust that citizens may have in the government and the president. In this respect, the transparency and accountability of the government play a key role because, through the clarity of public decision-making and of the evaluation of the executive’s actions, they deeply affect the structure of political power. A government that is trusted by the public may take key decisions in some areas that would otherwise be blocked by particular interests. Conversely, lack of transparency and accountability feed suspicion by the public and make governments weaker in the face of vested interests.

Asked about transparency in Tanzania, a former Minister of Justice and Constitutional Affairs, said a few years ago: ‘It’s something that is not in our culture!’, adding: ‘Our government has always been run on confidentiality.’ It is difficult to say whether Tanzania does better or worse than other countries on that account because of the ambiguity of available cross-country indicators for governance. However, in view of the thematic studies conducted in this project and of the general challenges discussed in this chapter, it seems fair to say that Tanzanian governance is barely transparent and, when transparent, barely accountable.

Of course, governing a country cannot be done in full sight of everyone. That is not the point. Transparency and accountability essentially require government objectives and intended policy reforms to be publicly set out, and, ex post, properly and publicly evaluated and debated, so that lessons are drawn from experience.

In the case of Tanzania, it is sometimes argued that the representation of the dominant party, CCM, in parliament is so overwhelming that debates do not have much content. This was the case with the preceding legislature where CCM held 72 per cent of the seats. Things are much worse today, since the opposition has been muted through the 2020 election – which is widely believed to have been rigged – and the violent repression of political opponents by the Magufuli administration. The question of transparency thus became completely irrelevant, and the public debate was simply shut down. Magufuli’s successor, Samia Suluhu Hassan, may hopefully prove less authoritarian, but she will run the country practically unopposed until a new election in 2025, except maybe for diverging factions within what has become again the ‘single party’.

In a perceptive analysis of governance in Tanzania, Lawson and Rakner (2005) show how the institutions in place and the presence of a dominant party were preventing the legislature from really holding the executive to account. Formerly, this has not always been the case, though, as could be seen at the time of the Escrow scandal – when several hundred million dollars were siphoned from the Central Bank, and the Public Accounts Committee of the parliament, chaired by a member of the opposition, exposed those members of the government whose carelessness had permitted it to happen and obtained their resignation. It turns out, however, that this was made possible because of one faction of CCM working against another, which was supportive of those party members involved in the scandal.³¹

Party fractionalisation is not a substitute for a lively opposition, as differing views tend to dissolve within party unity with respect to the electorate. It does not provide any incentive for transparency, except when it cannot be avoided as in extreme cases such as grand corruption scandals, as revealing within-party clashes would be electorally counterproductive. Note also that, even in front of publicised corruption scandals, central power is not necessarily held to account, as culprits within the party are not even sacked or properly sued.

Transparency and accountability are also missing in areas other than corruption and big business patronage. As stressed earlier in this chapter, not enough efforts are being devoted to the evaluation of the functioning of public services or to specific policies. The decentralisation reform has apparently not been fully implemented. Why is this the case? With respect to land laws, there have been multiple official assessment reports over the last ten years or so, yet little has been done. Is this because of inefficient governance or because of hidden vested interests? In the first case, there would be pure advantages in the government openly drawing the lessons of these reports and taking appropriate action.

It may also be stressed that none of these reports was a true evaluation, in the sense of attempting to quantify the economic and social losses due to the difficulty of implementing the laws or to the law itself in comparison with alternative rules. It is striking, for instance, that no fully reliable database showing the number and general features of land transactions is available to evaluate the present system of land transfers. This lack of evaluation and data gap is also observed in the educational sector, despite public knowledge that it does not function well, or in other activities of the public sector.

Only systematic evaluations of policies allow progress to be measured and adequate decisions to be taken. This is true at all levels of the governance of a country and irrespective of the structure of political and economic power. Generating the data for such evaluation, even internally within the executive, is a necessity and, of course, a step towards transparency, and possibly accountability. There might be capacity constraint in producing such evaluations, but this could easily be remedied if the political will is present. The path towards public management efficiency and, at a later stage, transparency, accountability, and citizens’ trust, starts with producing the adequate data, and allowing public access to them – unlike some laws passed during Magufuli’s rule.³²

D Three Key Principles for Action

The basic institutional challenges singled out in the thematic chapters of this study can be traced back to proximate causes that take different forms depending on the challenge being considered, but they result from a few basic mechanisms analysed in the preceding sections. Likewise, when examining the various remedies that have been suggested to overcome the main causes of basic institutional weaknesses, it is striking to realise that, assuming that the political will for reform does exist, they rely on essentially three principles for action.

The first principle is the necessary improvement of state capacity on the one hand through the rationalisation of laws deemed too complex for easy implementation and the structure of public decision-making, and, on the other hand, both equipment and skill enhancement of the civil service.

The second principle is to allow for more competition and competitive market mechanisms to play their role as much as possible but under adequate and strict supervision and regulation. This in no way should be taken as the blind promotion of ultra-liberal dogmatic economic principles in favour of fully competitive markets. Their limits are well known. This recommendation must be interpreted exactly in the other way as the need to avoid unchecked monopoly situations in all possible decision-making areas, whether in the administration, the delivery of public services, or indeed in the market. This ‘more competition’ – or ‘fewer monopoly situations’ – principle applies to land management with decentralised cash transactions between village communities and potential investors, to power generation with competition among providers, and to sectors of activity dominated by big business. But it also applies to the public sector with reforms that would reduce the monopoly power enjoyed by some operators – for instance, by allowing users to use alternative paths to obtain a licence, permit, or official document, and rewarding the most effective and expeditious ones. The extreme case is, of course, the use of information and communications technology facilities that simplify procedures and keep records of the performance of public officers. Making public service more competitive also brings about the right alignment of compensation with labour market conditions and individualises careers by relying more heavily on performance evaluation, assuming of course personal performance evaluation is properly, that is independently, conducted and based on objective criteria rather than on personal relations or cronyism.

The third principle is the systematic, regular, and rigorous evaluation of the functioning of the public sector, policy outcomes, and socioeconomic progress. Concerning the functioning of the public sector, this goes beyond applying results-based management principles based on a few indicators that can more or less easily be manipulated. The same applies to policy and social progress. In all these cases, evaluation is often performed ex ante on the basis of stated objectives and intended inputs rather than ex post on observed outcomes. Performance evaluation should be part of the evidence that measures the outcome of government actions, and its results, as quantitative as possible, should be made public.

Several countries have now made the proper independent evaluation of policies and reforms in some specific areas a constitutional obligation.³³ This could be an example to follow in Tanzania. Note, however, that this requires substantial progress to be made on the collection and analysis of statistical material and for the evaluation to be conducted by teams of independent analysts, which could possibly include foreign experts or observers who could guarantee the quality and impartiality of the evaluation.

At the political level, this principle of evaluation includes the checks and balances principle that defines, together with a few others, the functioning of democracy. Together with a strong and dominant party, the Tanzanian constitution gives extensive power to the president, to such an extent that the first president, Julius Nyerere, was known for joking about the constitution potentially allowing him to behave like a dictator. Strong central power, within reasonable limits, may be a good thing if wisely exerted. It is sometimes necessary to encourage reforms that are fundamentally good despite wide opposition by the elite, including within the dominant party. This is the case in Tanzania with corruption. Strong central power is not an advantage if it means that the executive is able to avoid its actions being properly evaluated by popular representation and public opinion, and to ignore the recommendations on key subjects that emanate from official expert commissions. The centralisation bias and the law implementation gap found to be key institutional challenges in Tanzania may be the result of such a concentration of power, a concentration that dangerously increased under Magufuli’s presidency. Strengthening checks and balances, in particular by introducing plurality in the filling of key positions presently under the sole control of the president and institutionalising the public evaluation of policies by independent experts, possibly including foreign observers, could contribute to improving the institutional context that conditions Tanzanian development. In this respect, it bears emphasis that properly publicised, evaluations emanating from donors and international organisations and conducted in collaboration with local experts may be extremely valuable.³⁴

Applied to specific sectors, such reinforcement of independent policy evaluation capacity and open reliance on expertise might in some cases greatly help the central power, even when it is close to being autocratic, by publicly exposing the forces that go against development and justifying policies. On the contrary, opacity or purely self-conducted evaluations lead to public distrust and, with time, risk adverse political turnarounds in case of a major exogenous shock.

E Conclusion

Following the medical metaphor, a diagnostic should be a statement about what disease is possibly causing observed symptoms of dysfunction. It is accompanied by a prescription describing the cure that is suggested to eradicate the disease. It is an illusion to believe that we can be as rigorous and effective in our institutional diagnostic. We have identified a list of symptoms caused by underlying institutional problems in Tanzania and the way they affect economic development, our institutional diagnostic exercise shedding light on the proximate causes of these dysfunctions, and general directions for reform being suggested. However, emphasis has also been placed on several ‘deep factors’, most often of the political economy type, which may prevent reforms from taking place or being effective. It is hoped that these reflections and conclusions will be of some help for thinking about long-run growth in Tanzania.

Given the use of the word ‘diagnostic’, some readers might have expected to find at the end of this volume a long prescription of things to do in most areas of public intervention. We have made such a list and mentioned in it several of the policy suggestions made throughout the volume – see the Appendix. However, the real contribution of this study is elsewhere: it is in the deep analysis provided in relation to the basic institutional weaknesses, including the political economy factors that may prevent effective reforms, and in the few fundamental principles of reform stated at the end of the preceding section. These principles should guide the reflection of policymakers in elaborating reforms and monitoring the progress made.

Economists are expected to quantify the effect of all possible reforms on the level of economic activity, the rate of growth, and the reduction of poverty. This is not an easy task when reaching beyond purely economic issues. In the field of institutions, this is simply impossible because of the extraordinary complexity of the way in which institutions affect development and the not-less extraordinary complexity of the mechanisms, including economic development, that could make institutions evolve. Economists regularly produce cross-country analyses that address these deep issues, but they are much too rough and aggregate to describe the reality of a specific country. Progress can be made only through detailed analysis of real cases, which has been attempted in this volume.

The reason why the relationship between development and institution is so complex is because it cannot be analysed without explicitly accounting for the structure of political power. It has been shown that this is of primary importance in Tanzania, as in any other country. Our discussion of corruption is a case in point. It would clearly make no sense for an institutional diagnostic of Tanzania to simply conclude that ‘corruption is damaging and should be eradicated’. Such a statement is certainly true, but useless for two obvious reasons. First, it is difficult to know how damaging corruption is to the economy, and what kind of corruption exists. Second, even if it can be shown that it is indeed very damaging, effectively eradicating it is not a simple policy decision. It requires the political determination for doing so, a huge political power, as well as the possibility of committing to such a policy for a sufficiently long time. Not all governments are in such a position.

An institutional diagnostic cannot get into the detail of the political economy of some of the pertinent reforms that could address institutional weaknesses. Yet it may be hoped that making the political actors aware of existing weaknesses and possible cures will also make them face up to their own responsibilities, whatever their political leanings. This is the reason why, irrespective of the political economy context, it is essential to show the nature of the institutional constraints that must be relaxed to enhance development, and the nature of the instruments that can be used to eliminate particular institutional weaknesses. This is what we hope this study has contributed in the case of Tanzania.

Afterword

François Bourguignon and Samuel Mwita Wangwe

The Tanzania Institutional Diagnostic was mostly completed in 2017 and 2018, at a time when it was too early to detect structural changes attributable to the Magufuli administration that had come to power in November 2015. Five years later, at the time of publishing this volume, many changes took place during the last two years of Magufuli’s first mandate and during the first year after his re-election, including his unexpected death and his vice president, Samia Suluhu Hassan, succeeding him in March 2021. One could have expected some continuity in this transition from a president to his deputy. The elements of discontinuity one can observe – treatment of COVID, freedom of speech, private sector development, foreign relations including with foreign companies – suggest that, indeed, either she did not agree with everything Magufuli did or she may have learned from experience that some reversals, adaptations, and improvements were necessary. As far as the present study is concerned, it also means that some of the disruptive measures taken by Magufuli might be about to be reversed, so that institutional changes since the core of the present study was completed might be less than could have been expected a few years ago, for instance at the time of the electoral campaign for Magufuli’s re-election, when a clear authoritarian drift in Tanzanian institutions was evident.

Under these conditions, the question arises of whether, with more visibility over the recent years, the main conclusions of the Institutional Diagnostic should be revised.

Some institutional aspects of the way both Tanzanian society and its economy function have been subject to change, and have been profusely commented on. Some changes were in the right direction, in the sense that they address or mitigate some institutional weaknesses emphasised in the Diagnostic. This is the case regarding the implementation of new anti-corruption strategies, even though it is still too early to evaluate what is, or what will be, their actual impact. Some other changes are more debatable but, for them too, it is difficult to identify the effect they are likely to have on development, especially in view of the fact that they may be in the process of being reversed. In what follows, therefore, we only identify the changes or decisions that are the most salient and we ask whether they aggravate or possibly attenuate the institutional weaknesses identified in the Diagnostic.

Referring to the five basic weaknesses that were singled out there, we ask the following question: how would major events, policy, strategies, and reform decisions observed over the last three or four years modify the Diagnostic?

Ill-defined structure of public decision-making was the first weakness. It is difficult to say whether progress has been made about the overlapping of responsibilities, an important point that was underscored in the Diagnostic, without surveying insiders. On the other hand, the ‘centralisation bias’ that was also emphasised has not weakened; indeed, quite the contrary given the numerous presidential interventions on detailed aspects of the functioning of the economy, from the export of cashew nuts to the management of the Dar es Salaam Port Authority, to regulating mining companies and to the formation of agencies taking over some functions of local government authorities such as the Rural Water Supply Agency and Tanzania Rural and Urban Roads Agency. It may also be recalled that the centralisation bias diagnostic was also partly based on the early overruling by Magufuli’s regime of decisions made by the agency supposedly responsible for the regulation of the electricity sector, the EWURA. Finally, the point made in the Diagnostic about the long implementation delays of laws and reforms still seem to apply. No progress has been made on adjusting the land laws and a recent report on decentralisation (Ewald and Mhamba, 2019), meaningfully entitled ‘Recentralization’, similarly suggests there is no noticeable progress in pushing forward the Local Government Reform Programme. However, there are indications that the so-called Samia administration is reviving the operationalisation of National Decentralization Policy and statements are made in support of local economic development such as building industrial parks. However, it is too early to discern the significance of this change.

Concerning selective distrust of market mechanisms and the private sector, the numerous statements of President Magufuli against the private sector – including both national and foreign firms operating in Tanzania – lead one to think that this institutional weakness has been aggravated for a while. Nonetheless, this opinion must be nuanced. On the one hand, these presidential critiques and actions were probably not directed to the operations of the private sector per se but rather to their consequences in the area of corruption. Reducing corruption, which was the absolute priority of President Magufuli, has entailed not only sacking ministers and top civil servants earning illegal income but also exposing the private firms providing it. Since the beginning of his mandate, he stated on many occasions that it was that aspect of the behaviour of private firms that he was criticising, certainly not their key role in economic growth, employment, and industrialisation. On the other hand, most indicators referring to the business climate have deteriorated in recent years. Tanzania has dropped 13 places in the World Bank Doing Business global ranking in a few years.¹ Furthermore, its ‘regulatory quality’ score, which includes business environment, in the Worldwide Governance Indicators significantly worsened between 2013 and 2018, after improving over the previous period.²

Of course, this worsening of the business climate may only reflect the fight against corruption and tax evasion resulting in many private firms being the victims of some harassment by tax people. Business managers at some stage expressed their frustration to the President, who then demoted the Commissioner General of the tax collection agency – the TRA. The President took this opportunity to castigate the approach TRA was adopting to harass businesspeople, which was leading to the closure of some businesses. He thus expressed a wish that the TRA would exert more effort to collect taxes without undue harassment. However, the formation of a task force parallel to TRA to collect tax, which threatened some firms of being charged of money laundering, with no bail, if they did not pay tax as estimated by the task force did not really go in that direction. In addition, attachment of tax obligations to bank accounts became common causing uncertainty to private sector operations.³

On the contrary, some other reforms have been undertaken to improve the business climate, like the easing of administrative procedures and approval of the blueprint for easing the business environment in May 2018 and its implementation from 2019 onwards. They may be seen as a move in the direction of reversing the backsliding trend on the business climate indices and global rankings. Overall, however, the approach Magufuli adopted towards the private sector was perceived as hostile while favouring the public sector institutions to perform functions that were usually performed by the private sector. This bias towards public sector operations was most explicit in sectors such as insurance and construction.

The Samia administration is reversing the negative perceptions on the private sector. She has held meetings with the private sector associations culminating in the meeting of the TNBC three months of her coming into office, has travelled abroad with private sector delegations, and has accelerated the implementation of the Blueprint. Improvements have been made in registration of businesses, permitting engagement of foreign experts and easing immigration procedures for investors. President Samia’s attitude towards the private sector is a reversal of Magufuli’s. It is too early to discern the results of the new direction but there are indications that the Samia administration is reversing the negative stance towards trust in markets and the private sector.

The underperformance of the civil service was found to be a third institutional weakness. Because this is to a large extent the consequence of a limited availability of skills and resources as well as putting in place performance management systems, changes in this area are necessarily slow. Incentives matter too, however. From that point of view, action was apparently taken to dynamise parts of the civil service and public agencies. At the same time, however, fiscal austerity led the executive to freeze promotion and salaries of civil servants, thus severely reducing their incentives. In addition, the fight against corruption in the public sector, as well as efforts to increase revenues, has led the executive to shake up the TRA several times. This crackdown has also affected other parts of the public sector. However, it is difficult to know whether it addressed the basic functioning of the sector or was limited to sacking managers. New managers were supposedly uncorrupted and expected to undertake efficiency-enhancing reforms. These efforts need to be reinforced by structural and institutional reforms to ensure sustainability, though, while re-establishing incentives throughout the civil service hierarchy, including performance management systems.

The Samia administration has now reversed the ban on promotions and the freeze on salary increments and has made promises about improving incentives for the civil service and about enhancing civil service reforms. In fact, salary increases have been announced to take effect from 1 July 2022. The style of micromanagement into the affairs of civil service has changed to one of greater autonomy. President Samia has instructed the Ministry responsible for civil service to review all regulations, update them, and then make sure they are followed. The tone is that of institutional reform and application of the rule of law. However, it is too early to make a judgement about what will have been the change in the efficiency of the public sector over the last few years. Moreover, the data that are needed to evaluate progress are most often missing.

The last two institutional weaknesses identified in the Diagnostic were rent-seeking and corruption and patronage and weak business regulation. These have clearly been the top priority of President Magufuli’s mandate, based on the personal commitments he made during the 2015 electoral campaign. The rigorous anti-corruption policy that was launched right from the start of the Fifth Phase Government focused on two sides of corruption: on the one hand, civil servants, especially at the top of the scale, and on the other hand big business. Action was taken against many people in top positions in the public sector, including ministers, top managers, and board members of public agencies who were either suspected or found guilty of corruption. Also, several firms were sued, and their operations suspended for evading taxes, most often with the complicity of top civil servants and influential politicians. Famous cases include a row with the Canadian Acacia mining company, which was accused of under-reporting the gold content of the ore it was exporting and then forbidden to ship it abroad. The company had to cease operations before an agreement, which is more favourable to Tanzania, was found. Another notable regulatory dispute took place with Aliko Dangote, the Nigerian businessman, about a cement factory for which the government wanted to renege on promises made by the previous administration. Here too, the plant was closed for some time until an agreement was reached.

These affairs had a demonstration effect to other investors that is likely to have contributed to worsening the business climate and, as mentioned above, reinforced the feeling that the government has an anti-private sector bias, particularly against foreign companies. In more recent times, President Magufuli had made pronouncements inviting investors including foreign investors to create or expand their businesses in Tanzania. It must also be stressed that the target of government’s attacks may not have been the companies themselves but rather the politicians and civil servants covering their illegal operations and benefiting from their largesse. If the objective was officially to make the operations of these companies more transparent, some observers also think that there may have been a political strategy behind it. Indeed, restricting companies’ ability to bribe powerful politicians, who are most likely to belong to some faction of the dominant party, could have helped the President to better control that party.⁴ This assertion is given credence by the observation that no action was taken against his close associates who were perceived to be corrupt.

In effect, there is a perception that double standards were applied in the fight against corruption during the Magufuli era, and that corruption may not have declined as much as claimed. On the other hand, corruption accusations may have been used as a way of getting rid of some opponents. Among the many people who were accused of corruption, some of them stayed in custody for one, two, or more years without being taken to court. When President Samia came to office many of these people were released and the Director of Public Prosecution was removed. This signalled the respect of the rule of law. However, the Prevention and Combating of Corruption Bureau, which had been very much strengthened during the Magufuli administration, continued to function as before in the Samia administration. Yet, one point that is relevant to this study is the emphasis she has placed on institutional capacity building in fighting corruption.

Effectively fighting corruption in a country where it has become a culture, of course, comes with some collateral economic costs, such as antagonising investors and the business community. To be sure, affairs like the Acacia–government row and the hassling of domestic and foreign firms by the tax authority have received widespread coverage by the national and international business media. Yet this reputational cost may be worth incurring if the expected impact of such a strategy on corruption and long-run economic growth is large enough. It is, however, too early to make any judgement on growth. As far as corruption is concerned, it is true that most indicators have improved. Tanzania’s rank in Transparency International’s ‘perception of corruption index’ went up from 130 in 2015 to 99 in 2018. Likewise, the Worldwide Governance Indicator for the control of corruption improved somewhat in 2018 as compared to previous years, without the change being enormous either. The same is true of the World Bank Country Policy and Institutional Assessment index for corruption.⁵ It must be realised, however, that those indicators reflect intended policies more than their outcomes, the problem in the case of corruption being that outcomes are extremely difficult to apprehend in the short and medium terms and that it may take a long time before changes in actual behaviour are observed.

Probably the greatest challenge of corruption is political corruption as demonstrated in the general elections of October 2020. That election was marred with a high level of corruption. Magufuli had threatened that corruption within the ruling party would not be tolerated and that any candidate who was discovered of engaging in corruption would be dropped from the race. In reality, it is believed that the level of corruption turned out to be higher than in previous instances. This reveals two further challenges. First, it is unlikely that such a magnitude of corruption could have been possible without candidates themselves making use of huge amounts of corrupt money. Second, as the pool of political leaders is selected from that group, the commitment of political leadership to the fight against corruption is seriously challenged. The ruling party won all parliamentary seats except a couple. Opposition parties cried foul, but their voices did not make a difference. Usually those who are dissatisfied with the results are allowed to petition This time there were no election petitions, a sign of loss of confidence in the objectivity of 2020 General Elections under President Magufuli’s watch as well as a sign of loss of people’s trust in the impartiality of the judiciary.

In summary, the fight against corruption by both subsequent administrations is without a doubt addressing one of the major institutional weaknesses of Tanzania. This represents a major change compared to preceding administrations. Nevertheless, some time will be needed to ascertain whether this really modifies the institutional diagnostic conducted in the present study. Experiences in other countries suggest that beating corruption takes a long time and undiminishing efforts accompanied by institutional reforms. In this regard, corruption is likely to still remain a pertinent problem in Tanzania for some time.

Putting things together, it seems that, although positive from several points of view, the action led by the two subsequent administrations (Magufuli and Samia) over the last three or four years has not modified the basic institutional weaknesses identified in the Diagnostic. There is an improvement in some areas, but there are signs of deterioration and possibly new weaknesses in others. This is the case of transparency and accountability, the two general principles of action recommended in the Diagnostic to help improve other institutions. The kind of departure from these principles seems to have changed, but the overall goal is still distant. While the Diagnostic found there was ‘some transparency with little accountability’ in previous administrations, the last three years during the Magufuli administration seem to point rather to greater accountability with less transparency. Formerly, many cases of misconduct were detected without much action being taken. In comparison, fewer cases have been exposed recently but action has been taken, that is, accountability is now upheld without delay. Overall, however, the transparency of the government in relation to its citizens seems to have been reduced, in some cases threatening the sense of democracy that has characterised Tanzanian society ever since independence, a quality that the Diagnostic considered, albeit perhaps not explicitly enough, as a major institutional strength.

Deliberate opacity of policies and outcomes, infringement on individual freedom, and rising authoritarianism during the Magufuli era were largely underscored and commented in the international press, whereas the national media found it increasingly difficult to raise these issues domestically. The list is long of events and decisions that contribute to this judgement. A few examples will suffice here.

The amendment to the Statistics Act tabled in Parliament in September 2018 that made it a criminal offence to publish statistics both ‘to disseminate or otherwise communicate to the public any statistical information which is intended to invalidate, distort or discredit official statistics released by the National Bureau of Statistics without authorisation’. This is one example of the opacity the government attempted to impose on the debate on policies and policy outcomes. To be sure, this decision was reversed in 2019 after bowing to pressure from the World Bank. Even so, however, the Statistics Act of 2018 restricted the debate on policy, in a direction opposite to the ‘evaluation culture’ recommended in the Diagnostic. It also cast doubt on the accuracy of published statistics, a doubt that was apparently shared by the World Bank, which estimates 2018 GDP growth has been noticeably lower than officially announced (World Bank Group, 2019, 11). In this vein, the vetoing by the government of the publication of the IMF Article IV statutory 2019 report, which suggested growth was slower than reported and was also critical of some of the government’s policies, inspired the same doubts. Because of such issues, it is hard to be convinced that efforts were truly being made under Magufuli towards more transparency and accountability in the economic policy sphere.

Beyond economics, several events and government decisions by the Magufuli administration, which also pertain to transparency, have led several international NGOs of repute to raise some concern about the treatment of human rights in Tanzania. It is not the place here to go into any detail on issues that are largely outside the scope of the Diagnostic. Yet the fact that an NGO like Human Rights Watch expressed concerns about the freedom of expression and the freedom of assembly in Tanzania, or that 30 civil society organisations called the attention of the United Nations Human Rights Councils to their perception that ‘the space for human rights defenders (HRDs), civil society, journalists, bloggers, the media, LGBTI persons, and opposition and dissenting voices’ was ‘shrinking’ are to be underscored.⁶ Also, a bilateral donor like the EU temporarily recalling its ambassador in November 2018 or suspending some aid programmes a little later, in both cases due to human right concerns, is not to be taken lightly, even though the crisis was quickly resolved.⁷

Since she assumed power in March 2021, the Samia administration seems to have reversed these trends of erosion of transparency and the treatment of human rights in Tanzania. Freedom of the press and other media has been restored and the threats to democracy have been mitigated.

In summary, the conclusion of this brief review of the way accounting for the recent past might lead to a revision of the Diagnostic is threefold. First, among the institutional weaknesses identified in the Diagnostic, progress may have been achieved on the corruption front, although it will probably take time before evidence of a real improvement is available. In this respect, it is fair to say that the resurgence of political corruption at the time of the 2020 presidential election is not a positive signal. There is no reason to modify the Diagnostic on other weaknesses, as President Samia seems to be decided to correct the most unfavourable institutional changes introduced by her predecessor.

The second conclusion is that a new weakness may have appeared. Unlike as was recommended in the conclusion of the Diagnostic, it seems the Magufuli administration was moving away from the pursuit of transparency in many ways. Such an attitude might have been part of a political strategy meant to ensure more authority rests with the President, possibly to win the anti-corruption battle. Most developing countries need ‘strong’ heads of state to push collective development rather than letting private interest rule. However, getting too strong without corresponding strong institutions may be dangerous if it means policymaking becomes less and less transparent. Besides mounting political discontent, the risk is that no countervailing opinion and correcting force would then prevent the country from heading in a wrong direction. This trend is being reversed in the Samia administration which is apparently restoring transparency and democracy. However, the ease with which this is being done is reinforcing the point made in the Diagnostic of weak institutions.

This is our third conclusion, namely the institutional weakness revealed by the flexibility of Tanzanian institutions depending on the president. There is a fundamental flaw in political institutions that permit the excesses observed during the Magufuli era to take place and eventually to be corrected by the next administration. What was stated as a joke by Nyerere that the Tanzanian constitution allowed him to rule as a dictator is in fact true. The constitution does not protect against a President who would strongly deviate from democratic principles.

A possible explanation of the Magufuli’s authoritarian shift was his need to take full control of the various factions within the CCM ruling party in order to launch his anti-corruption strategy. This initial fragmentation of the ruling party into distinct groups of interest was a key feature of the political economy of Tanzania before Magufuli, and a potential obstacle to the adoption of some development-friendly reforms. He apparently succeeded in controlling them, as shown by the results of the 2020 general election. There is full uncertainty about what would have happened next: would he have moved back to a less authoritarian leadership or not? Samia’s leadership of the CCM so far has shown two types of changes. First, she appointed new leaders in the Secretary General and Vice-chairman positions. Second, she has demoted or removed from their positions leaders who were identified as aspirant for presidency in 2025, which implicitly means that she does not have her hands free from CCM factions.

The Magufuli era may have been a kind of parenthesis in the political history of Tanzania, during which a disruptive president attempted to impose a tough anti-corruption strategy, seen as essentially advantageous for development, which required exerting authoritarian power over various political factions. However, the negative attitude towards the private sector was not favourable to long-term private investment that is crucial for development. Such a state of affairs had been unseen in Tanzania over the last 30 years, and maybe since independence, and certainly was a major change over preceding administrations. President Samia took over, reverted to a more democratic and decentralised rule, but is now probably facing re-built factions within the ruling party which aspire to power in the next general election, and may represent obstacles to the conduct of policy by the new administration.

Even though there is some continuity with the previous administration, it cannot be discarded that the situation is back to what it was under Magufuli’s predecessor. All in all, there is no evidence of need or reason to change our diagnostic.