152S(4 PSD Occasional Paper No. 18 September 1995 8 > Economic Strategy and Structural Adjustment in Tanzania Brian Van Arkadie I R I The World Bank Private Sector Development Department Private Sector Development Department Occasional Paper No. 18 Economic Strategy and Structural Adjustment in Tanzania Brian Van Arkadie October 1995 This paper has been cleared for inclusion in the occasional paper series by R. Shyam Khemani, Manager, Competition & Strategy. The views expressed are those of the author(s) and should not be attributed to the World Bank. This paper forms part of a forthcoming volume, Leila Frischtak and Izak Atiyas (eds.), "Govemance, Leadership and Communication: Building Constituencies for Economic Reform." The World Bank Private Sector Development Department ii Contents 1. Origin of the Economic Crisis ..................................................... 1 Radical Change Following Independence .............................................................. I The Agricultural Institutional Environment ............................................................ 5 1973: The First Economic Crisis and the Response ............................................... 7 1979: The Second Economic Crisis ..................................................... ......... 8 Governance and the Managed Economy .............................................................. 10 2. The Politics of Reform: Economic Disequilibrium as a Source of Change ..................................................... 15 3. The Reform Process .................. ................................... 23 Resistance to Change and First Steps to Reform: 1979-84 .......................... 23 Resistance to Exchange Rate Adjustment ..................................................... 25 The Reform Process: 1984-93 ........................................................ 28 The Slow Pace of Parastatal Reform ........................................................ 31 4. The Distributional Impact of Reform .............................................. 35 5. Sustainability of the Reform Process ............................................ 39 The Momentum of Economic Growth ........................................................ 39 The Managerial Capacity of the State ........................................................ 41 6. The Role of Donors ..................................................... 45 7. A National Vision of Development? ............................................... 51 Bibliography ..55 iii Al Origins of the Economic Crisis Radical Change Following Independence 1.1 In Tanzania, the need for structural adjustment in the 1980s had its origins in both the negative events in the international economy and the failure to adjust policies swiftly enough in response to the initial shocks. In many respects, the evolution of the Tanzanian economy since independence and the events that led to structural adjustment are quite similar to those experienced by many other African countries. Another, and perhaps more interesting, part of the story is more particular to Tanzania: the crisis in the national model of development that had incorporated the vision of the charismatic independence leader, Julius Nyerere. That crisis was brought to a head by exogenous shocks. Whether the system would have sustained itself in more propitious external circumstances is a matter for historical speculation, but faced with deep crisis, reform was unavoidable.' 1.2 The economic reform process, as it came to be understood in the Tanzanian context in the 1980s, had four main elements: (1) stabilization, through the use of orthodox macroeconomic policy instruments to restore macroeconomic balance, both domestic and external; (2) liberalization, reducing bureaucratic controls and government allocation of resources, and opening areas of the economy to private business that had been public sector monopolies; (3) reform of stale enterprise and privatization; and (4) reform of the government itself to enhance government capacity. 1.3 Some of the policy issues on the reform agenda, particularly in relation to stabilization, could be seen as more generally technical in character (although even issues such as the exchange rate, which are sometimes considered entirely technical matters, had strong political implications). Many of the policy issues, however, had obvious ideological content, particularly those related to liberalization and privatization, because they implied dismantling the system created following the Arusha Declaration. I In an early conmnent (Van Arkadie 1983) on structural adjustment in Tanzania, I argued that the evident need for stabilization and some institutional reform should not be used as leverage to enforce dismantling of the existing economic system. In retrospect, although I remain critical of some aspects of the structural adjustment orthodoxy of that period, I recognize that I underestimated the depth of the malaise in the Tanzanian system at that time. 2 Economic Strategy and Structural Adjustment in Tanzania 1.4 The Arusha Declaration, issued by the ruling party, the African National Union (TANU, later renamed CCM), in February 1967, engendered international approval because of its emphasis on the priority attached to rural development, its rhetoric of self-reliant development, and its commitment to egalitarian economic and social policies. The Arusha Declaration seemed to incorporate a clear vision of Tanzanian development, which was implemented with great vigor by a charismatic and honest leader. The argument in the declaration was that self-reliance and egalitarianism could be best achieved by combining a leading role for the state in all large-scale activities with the promotion of rural development through cooperative agriculture (rural development based on the principles of "Ujamaa"). 1.5 With its emphasis on rural development and concern for the broad spread of the benefits of economic growth (in essence an early espousal of what later became known as a "basic needs" approach), important parts of the Arusha vision seem as valid today at they did in 1967. It is of some interest, therefore, to consider whether the economic difficulties that engulfed the Tanzanian economy demonstrated the fundamental unsoundness of the Arusha vision-its impracticality despite its Utopian attractions-or whether the problem lay in the faulty implementation of basically sound ideas. In the contemporary Tanzanian political debate, both explanations receive support. 1.6 One important element in the evaluation of the Arusha vision depends on interpretation of the degree of consistency between policies implemented and intentions, as stated in the declaration. Particularly in the period beginning in 1972, policies were implemented that arguably were inconsistent with the positions developed in the Arusha Declaration. The formulation of the Basic Industrial Strategy involved a definite shift from the priority given to rural development in the declaration. There was increasing centralization of economic authority and urgency in the implementation of programs. The attempt to enforce the villagization ("Ujamaa Villages") program over two years, the abolition of cooperatives and local government, and the commitment to achieve universal primary education in two years all reflected an impatience to achieve political goals swiftly, even if it meant the loss of the principle of decentralized participatory rural development associated with the Arusha Declaration rhetoric and downplaying realism in the implementation of policies. At the same time, the emphasis on the ascendancy of politics-and politicians-over the bureaucracy and workers' participation in the management of state enterprise seemed to incorporate a heightened political radicalism, although at the expense of weakening the self-confidence and role of technocrats. 1.7 The reasons for the heightened radicalism in the 1970s are not entirely clear, and cannot be established in this chapter. One possible explanation is that most of the policy criticism came from the intellectual left.2 Nevertheless, it is unclear how much support the left-which 2 This was the view developed by N. Lipumba in commenting on an earlier draft of this chapter. He was in Tanzania at the time, and places emphasis on the influence of the 'left critique in raising the ideological content of Tanzanian politics and contributing to the gap that he sees as emerging between policy intentions, as set out in the Arusha Declaration, and practice, as implemented in the period from 1972 to the mid-1970s. While there was Origins of the Economic Crisis 3 argues, for example, in favor of the basic industrial policy-could summon. President Nyerere seemed able to ignore or contain criticism from intellectuals with little difficulty. The explanation for the period of heightened ideology probably lies in the evolution of the thinking of Nyerere himself, who exhibited an increasing impatience with technocratic conservatism and a belief in the need to accelerate the development process. Aspects of the extension of state control also had a more pragmatic element, in that it reflected an effort to defend established priorities in the face of the economic crisis of 1973. By the mnid-1970s the degree of state intervention and central economic control had gone well beyond that envisaged at the time of the Arusha Declaration. 1.8 Insofar as a quasi-planned socialist economy was created, economic reform involved not merely a technical change in economic policies, but also a more thorough change in the economic and social system. The transformation was not equal to those of Eastern Europe and the Commonwealth of Independent States (CIS), because Tanzania had remained a mixed economy, but it did imply fundamental institutional change. Economic reform in Tanzania was therefore more intensely political than the customary reform process-it involved comprehensive change in economic institutions, and eventually in political institutions as well. 1.9 The need for reform was evident at the beginning of the 1980s, when the economy was faced with a deep crisis. The immediate symptoms were an extreme shortage of commodities, the result of a crippling shortage of foreign exchange, which in turn reflected such long-term phenomena as stagnation in export crop production, excessive dependence on donor support, and low productivity of the public sector investment programs in industry and infrastructure. 1.10 The economic system facing the crisis had been created following independence, which had brought economic policy change in four major areas. The growth in public spending accelerated, not merely in response to the ambitions of the new government, but also because two other developments made it possible to translate ambition into practice. Development aid broke the link between domestic savings and investment and funded development spending with too little regard for the recurrent resource base to sustain the expanded activities. In addition, the monetary restraints of the colonial financial system were ended as the colonial Currency Board gave way to the new Central Bank, allowing for monetary and fiscal flexibility and a break in the link to sterling. * There was a decisive shift in control over the economy. Change was initially concentrated in the staffing of the public service. The rapid localization of the such a debate-involving, for example, intellectuals at the university-it is difficult to judge how far polemics affected government decisionmaking. By and large, Nyerere seems to have been fairly impervious to such influences. The alternative explanation may be sought in the character of Nyerere himself, who had become increasingly confident of his capacity to override technical arguments suggesting caution, as he had done successfully in the initial implementation of the Arusha Declaration, and at the same time placed increasing faith in the potential for using the party for mass political mobilization to accelerate development. 4 Economic Strategy and Structural Adjustnent in Tanzania bureaucracy was one of the first priorities of the new government. But with independence, a lack of congruence emerged between political and economic power. It was no longer acceptable for Africans to be limited to the roles of peasant and worker. Either expanded public ownership or public support for the transfer of wealth to an emerging African property-owning class (as in Kenya) inevitably appeared on the political agenda; in Tanzania, this issue was confronted in 1967 in the Arusha Declaration. * There was an impressive expansion in social service provision in the early 1960s, and the Second Five-Year Plan (1969-74) set out a strategy similar to what some years later became known in the donor world as "basic needs." * The fourth, and least desirable, change was the expansion of military expenditure. Tanzania was neither militaristic nor aggressive, but sovereignty had implications for necessary military spending. The imperial regime enjoyed economies of scale because of its ability to move forces from colony to colony, an efficiency lost as each newly independent country faced the need to form a national army. 3 1.11 Despite the greater short-term fiscal and monetary flexibility gained by the replacement of the Currency Board by the Bank of Tanzania, over the long term the real resources available to government from domestic sources were constrained by the growth of export earnings, which in turn were affected by the vagaries of international markets, technology in the primary production sector, and the incentives provided to farmers. In the 1970s, the combination of ambitious fiscal expansion and the costs of the expansion of the public sector placed an increasingly heavy burden on the export sector. As the budget expanded and the single-channel (monopoly) marketing system became increasingly expensive, the export sector was squeezed, and export production faltered. 1.12 Although the First (1964-69) and Second (1969-74) Five-Year Plans were intended to diversify the economy, and the Basic Industrial Strategy, devised during the Second Plan period, aimed to accelerate the pace of industrialization, industrial investment failed to generate either much export diversification or real import substitution. Import dependence was not reduced, and industrialization through import substitution effectively increased structural dependence on imports by encouraging the creation of industries that could only operate with imported inputs. At the same time, the donor-financed projects crucial to development had extremely low average productivity. As declining export performance and negative shifts in international terms of trade reduced foreign exchange earnings, the initial response was to attempt to defend the resources at the disposal of the state and its functionaries. 3 In Tanzania, the problem of managing the army was thrown into sharp focus when the army mutinied in January 1964. Later expansion in the army is in part to be explained by instability in neighboring countries, which eventually led to the Kagera (Idi Amin) War with Uganda, 1978-79. Origins of the Economic Crisis 5 The Agricultural Institutional Environment 1.13 Crucial to a consideration of the origins of the Tanzanian crisis and the need for reform is the institutional structure conditioning agricultural perforrnance. Despite the extension of public ownership after the Arusha Declaration to include most large-scale economic activity, the bulk of the population remained outside the state sector, living and working on smallholdings; the typical unit of production remained the household.' 1.14 Given the dominance of smallholder farming, policies toward agricultural marketing were the most important government influence on agricultural performance. The postindependence policy had been to promote single-channel marketing through a chain running from the primary cooperative, to the cooperative union, to a marketing board, with monopoly enforced at each stage. The effort to extend the cooperatives beyond the areas where they had thrived as voluntary institutions to become a universal, but imposed, marketing mechanism had created an inefficient system (in 1966, it had already proved necessary to convene a presidential commission of inquiry into the cooperative system). Moreover, where the cooperatives had flourished before independence-such as the coffee- and cotton-growing regions-they posed a political dilemma as an alternative to the party as a focus of local political power.' The solution adopted in 1973 was to abolish the cooperatives and to amalgamate the local marketing functions, the previous activities of the marketing boards, and some of the extension responsibilities of the Ministry of Agriculture under newly created crop authorities. These large monopolies were unable to deliver what was expected of them, and their inefficiency became a prime cause of the deterioration in agricultural incentives,6 despite subsidization of loss-making crop authorities, which became a heavy budgetary burden. 1.15 Another aspect of institutional policy that commanded international attention was the Ujamaa Village program. When the Second Five-Year Plan was published in 1969, a commitment was made to extend the principles of Ujamaa (Nyerere's vision of socialism) throughout the countryside by promoting Ujamaa Villages.7 The villagization program, implemented in 1973-76, 4 In the early 1960s, about half of export production (most notably cotton, cashews, robusta coffee, and a significant proportion of arabica) and the lion's share of food was produced on smallholdings (see Van Arkadie and Ghai 1969). 5 It could be argued that there was a basic political contradiction in the rural development strategy of government. The rhetoric tended to emphasize a decentralized and voluntary ideology of rural development, an aspect of Nyerere's thought that added to its international appeal. If genuinely spontaneous local institutions developed, however, they were likely to be identified as a threat to the authority of the party, as was the case with the Ruvuma Development Association, which seemed at the time to be implementing the Nyerere philosophy, but was nevertheless abolished. 6 The burden on farmers resulting from the inefficiency of the crop authorities was initially documented in a series of studies by Frank Ellis, Economic Research Bureau, University of Dar es Salaam. 7 The Ujamaa Village was probably Nyerere's most famous innovation, and also his greatest failure. There is not the space here to commnent on that experience, except to note the great gap between the initial 6 Economic Strategy and Structural Adjustment in Tanzania sought to transform the pattern of rural settlement by congregating the rural population-which previously had been resident predominantly on dispersed family smallholdings-in nucleated villages of sufficient size to be efficient (in bureaucratic terms) units for the delivery of services. Involved in this plan was the idea that the new villages could also become the basis for a socialist system of production. 1.16 The program was massive, although not universal, in extent.' The effect of the program on production remains unclear. There must have been considerable short-term disruption, because production areas were lost and energies were diverted to building the new villages. In some areas, villagization created an increased risk of environmental degradation. Nevertheless, the only case widely 1presented to demonstrate large, long-term output losses from villagization is that of cashew nuts. Even if it is not possible to establish that villagization had a strong negative effect of production, however, it certainly diverted energies from activities that could have enhanced productivity. Little was achieved in the nature of socialist production, and although there was a substantial effort to extend social services in the rural areas, much of the progress at the village level was eroded during the subsequent economic difficulties.'° 1.17 Institutional weaknesses contributed to the long-term malaise in export crop performance. In the 1970s agricultural export production declined, in striking contrast to the buoyant performance of the 1960s. In summary, costly marketing arrangements severely reduced conceptualization, which emphasized the spontaneous and voluntary nature of the experiment, and the eventual mass implementation, which was managed and bureaucratic. -3 Important areas seem to have been largely untouched, such as the main coffee-growing (and therefore more developed) areas of Kilimanjaro and Kagera. The central, southem, and coastal areas of the country were extensively affected. Claims were made that after the completion of the program more than 90 percent of the rural population lived in villages; it is not possible to evaluate that claim because of lack of data and definitional ambiguities in the concept of the 'village.' For a critical (from the left) account of the Ujamaa Village experiment, see von Freyhold 1979. 9 The argument runs that with dispersed settlement, it was easy to tend a stand of cashew trees around the house with very little labor timne. Once settlement was concentrated, it was not possible to plant around the house, while the existing stands would generally be some way from the new settlement. Certainly, cashew-producing areas were subjected to intensive villagization, and there was a severe loss of output. There were other plausible causes for the decline, however, such as unattractive prices and disease problems. 10 Probably the most influential critique of rural policies in this period is by Goran Hyden (1980). He develops the concept of the *uncaptured peasantry' to describe the failure of the state and party to mold the rural population to the chosen development design. This concept needs to be handled with some care, as sometimes it can be taken to mean that the peasantry was resistant to penetration by the market. There was little evidence of such resistance, as in the 1960s the peasantry had responded with great vigor to market opportunities, and the decline in marketed output in the 1970s was not so much a resistance to the market as a response to the undermining of market incentives and institutions. It is clear, however, that the peasantry was not to be readily incorporated into some sort of bureaucratic-cum-socialist mode of production organization, if that had been seriously intended, which is a moot point. Origins of the Economic Crisis 7 returns received by farmers. Other negative factors included sweeping institutional changes, such as the abolition of cooperatives, villagization, and decentralization; poor design, by donors as well the Tanzanian authorities, of programs to promote agricultural innovation, long-term shifts in market prospects (for example, for sisal); and unfavorable weather. Moreover, the burden of a monopolistic marketing system meant that macroeconomic adjustments to improve farm incentives were inadequately passed on to the farmer (at the extreme, increasing official producer prices made little sense if farmers were not paid, or paid after long delays, as often happened). 1973: The First Economic Crisis and the Response 1.18 The Tanzanian economy proved vulnerable in the face of the new international economic disorder of the 1970s. In 1972-73 the system entered its first period of great economic difficulty. How the initial causes can be divided between exogenous shocks and policy failure remains a matter of debate."' The weakness and inflexibility of the export base demonstrated the validity of the shared premise of many policymakers throughout Africa in the 1960s that fundamental structural change was needed, but Tanzania found neither the means nor the time to effect the transformation before the onslaught of hard times. Even if economic difficulties had exogenous origins, policies were of varying effectiveness in responding to them. Domestic policy weaknesses included the excessive buildup of public expenditure, which placed pressure on available resources. The main proximate causes of the economic crisis, however, were two sets of exogenous shocks: drought resulting in harvest failure and the sharp deterioration in the terms of trade following the first oil shock. It was not perceived as a crisis in the system as such. 12 1.19 With access to increased aid, and the eventual recovery of the terms of trade (the result of the boom in coffee and tea prices in 1977-78), the system weathered the crisis without the sort of adjustment and liberalization measures that later emerged as the structural adjustment package. Response to crisis was actually in the reverse direction-government responded to an increasingly tight economic situation by increasing government control over the economy. Why was the system able to avoid adjustment in the 1970s, while not being able to do so a decade later? Apart from the fortuitous improvement in the terms of trade in 1977-78, another strikinig difference was the response of the donor community. Finance provided by the IMF (under the Witteveen and other facilities) during the first crisis had much softer conditionality than would be A study partly written by the author of this essay, R. H. Green, D. G. Rwegasira, and B. Van Arkadie 1980, describes the evolution of policy in that period. Critics of that study have argued that it was too much of an apologia for Tanzanian policy, ascribing too much of the explanation of poor performance to exogenous shocks rather than failures in policy. 12 Systematic criticism of the Tanzanian policy regime tended to come from critics on the left of the regime before the regime fell out of favor with the conventional aid establishment. Probably the most comprehensive and thoughtful critique of post-Arusha economic policies is to be found in Coulson 1982. The book drew on work produced and published in the early 1970s by Coulson and other commentators in Dar es Salaam who were critical of economic performance from a radical perspective. 8 Economic Strategy and Structural Adjustment in Tanzania the case ten years later. The period after 1973 also witnessed a rapid buildup of aid; the Nyerere vision proved attractive not only to a number of bilaterals, but also to the World Bank under President Robert MacNamara. This provided a much more sympathetic donor environment than was the case in the early 1980s-the World Bank had turned sour, the IMF was pursuing harsher conditionalities, and even the more sympathetic bilaterals were not willing to expand their support. 1.20 In the first half of the 1970s, weakness in the managerial capacity of the state was not yet evident, and the authority of the political leadership was at its highest, so that an aggressive, interventionist response to economic difficulties carried some credibility. In response to the crisis of the early 1970s, a number of decisions about the tactics of economic management moved the system in the direction of a controlled, command-style economy. Controls were increased to defend the state's access to increasingly scarce resources. Efforts to shield the domestic economy from the full force of the changing conditions included defending the existing exchange rate, when in retrospect it would have been better to have made a quicker response to new, harsher realities. Instead, pressure on the balance of payments was met by measures to tighten government control over the allocation of increasingly scarce resources, such as the extension of controls over foreign exchange, imports, and credit allocation. 1.21 Over the same period, inflationary pressures threatened the carefully designed prices and incomes policy, which had been a centerpiece of economic management in the post-Arusha economy. Stens to equitably share the burden of economic difficulty included further compression of official wages and salary differentials and the extension of the system of price controls. The policy response to the first round of economic crisis also included popular mobilization efforts, such as a mass drive to expand food production. 1.22 The response to the crisis of the early 1970s was thus to increase the role of administrative allocation of scarce resources (particularly foreign exchange) and government control over incomes and price movements, rather than to seek a new equilibrium through price adjustment. Nevertheless, at no time was a clear strategic decision made to opt for a "planned" or command economy in preference to the use of market tools (see Green, Rwegasira, and Van Arkadie 1980 for a sympathetic account of efforts to manage the economy). 1.23 It is difficult to judge how effective the Tanzanian state was in enforcing its priorities. Although it is difficult to pinpoint the timing of the decline in state capacity, however, it is evident that the overextension of the role of the state placed great pressure on its capacity, while the heightened ideological content and politicization of the government decisionraking process eroded the authority and self-confidence of the bureaucracy. By the end of the decade there had been an erosion in the capacity to carry out the economic management tasks of government. 1979: The Second Economic Crisis 1.24 By 1979 a new economic crisis had taken hold that was much deeper than GDP data, which could be interpreted to imply a modest dip in the level of economic activity, would suggest. Origins of the Economic Crisis 9 Constraints on the availability of imports led to what was described as a goods famine-even the most basic goods disappeared from markets. The cumulative effects of weakness in public sector administration were exacerbated by the foreign exchange shortages, so that deterioration in the maintenance of infrastructure became more acute. There was considerable inertia, shared by donors and government alike, in persisting with investment programs that were no longer viab!e. It made increasingly less sense to use available finance to fund new capital projects, which only added to already idle capacity, particularly when the existing public capital stock was rapidly deteriorating. Tanzanians became increasingly aware that the fairness and honesty of the public service was declining. 1.25 In weighing the relative importance of the internal and exogenous factors that brought the Tanzanian system into acute crisis in 1979 (see Stewart and Sharpley 1985), a judicious balance can be struck by noting that decline in export production and evidence of increasingly ineffective performance in the parastatals suggest that the system was stagnating, but that exogenous shocks were so severe that there would have been a serious economic crisis even if the underlying economic performance had been much better. 13 Negative exogenous influences included the end of the brief tea and coffee price boom; the breakup of the East African Community, with the consequent disruption of services and the need to replace them at the national level; the second oil shock, and the heavy burden of the war with Uganda (the Kagera, or Amin War, which began following border incursions and territorial claims by Amin, and was completed with the ouster of the Amin regime from Uganda). 14 1.26 There may have been an interplay between the dramatic political events of the period and the approach of the Tanzanian leadership to economic policy dialogue. Success in the Kagera War may have emboldened the leadership in other spheres, so that when Tanzania found itself in conflict with the 1MIF, it was not only possible for Nyerere to interpret the differences as essentially a political dispute, but also to see it as a dispute that could be won (I owe this insight to N. Lipumba). 1.27 The Kagera War had another effect: there was a general deterioration in the probity of the Tanzanian public service, which was faced with a sharp deterioration in the real value of official salaries associated with the difficult wartime conditions. This was compounded for public servants transferred from East African Community service to employment on much less advantageous terms with the Tanzanian government afler the Community's breakup. This 13 And, indeed, by the 1980s virtually all the economies of Africa were in crisis, even those touted at one time or another as models of successful, market-orientated development (such as Kenya and CMte d'lvoire). 1 4 There is an interesting parallel between the donor reaction to the Amin ouster and the roughly contemporaneous ejection of the Pol Pot Cambodian regime by the Vietnamese. In both cases, an innocent might have supposed that the invading forces were providing an unequivocally positive service to the world community by ejecting particularly obnoxious regimes, but in both cases the international reaction was negative- nonsupportive in the case of Tanzania and outright hostile in the case of Vietnam. Possibly such adventures are viewed as the exclusive preserve of the great power(s)? 10 Econornic Strategy and Structural Adjustment in Tanzania transfer included inland revenue and customs staff, and any deterioration in the morale and performance of these cadres was of real significance to the economic management capacity of the government. 1.28 One reason that the crisis at the end of the 1970s was readily interpreted as external in origin was that the dorninant feature of the crisis was the shortage of foreign exchange. The severe import constraint was associated with the sadly inadequate benefits generated by the substantial investment program implemented in the 1970s (see Tanzanian Economic Trends, 1988). Imbalances built up between new investment and the inadequate provisions to maintain the existing capital stock, between capacity expansion and the availability of foreign exchange to fund recurrent inputs, and between plant expansion and the technical and managerial know-how to operate the new plants. Most strikingly, the large program of industrial investment initiated in the mid-1970s generated no increase in industrial output because of import dependence, infrastructural bottlenecks, and design and managerial weakness (see Skarstein and Wangwe, 1986). Thus, underlying the short-term economic management problems that confronted the government was an increasing structural weakness. The economy was becoming more import dependent, while domestically generated funding for imports was stagnating. More subtly, the combination of increasing scarcities (and accompanying rent opportunities) with compression of salaries had created the conditions for the erosion of the probity and commitment of the bureaucracy. Governance and the Managed Economy 1.29 To use the current vocabulary, how should the "governance" of Tanzania, before the crisis of the late 1970s and the reforms of the 1980s, be characterized? There was a paradox in the condition of the Tanzanian state, demonstrated in the contrast between decisive leadership at the top-consistent with a quite "hard" state, capable of making and implementing difficult decisions and mobilizing mass support behind development projects-and a system of government with attributes of the "soft' state-with lax administration and a great gap between intention and achievement in the implementation and management of even routine government activities, such as road repair. Political scientists have explored that paradox (van Cranenberg 1990 deals with this paradox; see also Hyden 1980, 1983). 1.30 Governance can be seen has having two different aspects. The first is the effectiveness of the political system in maintaining stability, coherence, and legitimacy in response to the political interests in play. In this area the Tanzanian system scores well in some respects, particularly when viewed in the context of African political instability. The second area includes the managerial capacities and probity of state organizations, and here the Tanzanian record is less satisfactory. One source of weakness was the limited initial human capital stock at the time of independence, when there was inadequate national manpower, in striking contrast to the situation in most Asian and Latin American countries. The imperial bureaucracies of Africa had a history that was too brief, and locals were introduced to the system too late to have created anything like Origins of the Economic Crisis 11 the Indian Administrative Service, or to have laid down the deep historical roots of the Mandarin tradition that has influenced many parts of East Asia. The Tanzanian human capital stock was fragile, even when compared with other British colonies in Africa, such as Ghana or Uganda. 1.31 A bureaucratic elite swiftly emerged following independence. It was often professionally competent, but its consolidation was impeded by disruption, following the Arusha Declaration, when many of the most able civil servants transferred to manage the new parastatals, spreading available talent very thinly. The administrative system was next subject to a frenetic sequence of reorganizations. In addition, the emphasis placed on political leadership and the role of the party, particularly in the early 1970s, stunted the growth of a self-confident bureaucratic elite-in part, intentionally.'5 The weakness of the civil service made the system highly vulnerable to poorly thought out external proposals for administrative reform (in the early 1970s, the consulting firm of McKinsey was a particularly influential adviser on reform). 1.32 As the supply of qualified nationals increased, there was a systemic failure to utilize trained staff effectively. Part of the explanation is that in the postindependence decade too many tasks were loaded onto a bureaucracy still handling transition from the colonial authorities, overburdening administrative capacity when it might have been wiser to consolidate. Part of the "hard/soft" paradox arose from the contrast between a strong political leadership, driving to extend state management of the economy, and a weak bureaucratic apparatus, allocated the task of managing the new system. Even if there was no attempt to introduce a full-fledged command economy, the expansion of state ownership after 1967 and the extension of the planned allocation of scarce resources in the early 1970s stretched an already overburdened administration. 1.33 The overextended civil service found it increasingly difficult to induct new staff effectively, leading to a decline in morale, which accelerated with the decline in real incomes that followed the squeeze on the real resources at the disposal of the state in the early 1970s. The civil service sense of responsibility and accountability for the effectiveness of policy was also eroded by the primacy of the party and politics over administrators and technical considerations. The result was a decay of administrative systems and a consequent disinvestment in human capital: new entrants into a decaying bureaucratic milieu were exposed to a process of unlearning by not doing. 1.34 If inadequate numbers of trained staff had provided a plausible explanation of weaknesses in state capacity in the 1960s, it was much less convincing by 1980. While gaps remained in particular technical and scientific skills, the stock of human capital, at least as measured by formal qualifications, had been transformed. While in the 1960s it was necessary to draft scarce graduates into the public service through bonding schemes, by 1980 disguised 15 By the end of the 1970s, a slightly ironic usage had developed among senior civil servants, in which responsibility was implicitly disclaimed by reference to the 'policymakers,' that is, the politicians, who were seen as having essentially taken all control and responsibility out of civil service hands. 12 Economic Strategy and Structural Adjustment in Tanzania graduate unemployment was widespread in the public service, with an excess of graduates employed. 1.35 To emphasize the weakness of the bureaucratic structure is to emphasize the technocratic aspect of the content of governance. Turning to the political element, although the system can claim to have delivered national stability, in other respects the achievement fell short of expectations. Given the rhetoric of President Nyerere and the party, it might have been hoped that what was lost in bureaucratic competence might have been gained through grassroots mobilization. Numerous studies, however, have indicated that although the party was remarkably effective in organizing a political machine throughout the country, it was subject to the basic contradiction that development of the party structure was at the expense of the creation of autonomous, grassroots institutions (see von Freyhold 1979; van Cranenberg 1990; Finucane 1974; Shivji 1976). The abolition of local government, cooperatives, and the Ruvuma Development Association (RDA) and the contradictory reality of government "decentralization" could be seen as part of a pattern of concentration of power in the party. Nevertheless, having laid claim to a monopoly of leadership in development, the party was ill-equipped to carry out its self- imposed task. Phil Raikes (University of Dar es Salaam, Economics Research Bureau), in a pungent comment, suggests the destructive impact this had on agriculture: leaders whose knowledge about agriculture, economics, or the constraints of peasant farming were minimal, felt free, indeed obliged, to issue directives, to show their commitment. A significant proportion of ... extension advice was already of dubious value ... further simplification.... increased the proportion which was irrelevant or technically incorrect. . . . Apart from the considerable waste of time and/or energy involved in either complying with or evading directives, they appear to have contributed significantly to a decline in morale: a feeling that whatever one does to try to organize one's life sensibly, some clown in office will come along and mess it up. 1.36 Another aspect of the Tanzanian paradox resulted from the weakness of national interest groups pressing the state to provide effective economic services. Although the absorption of virtually all graduates into public service-in government and the parastatal sector-eventually undermined both the performance and the real incomes of graduates, it also meant that the educated elite were almost entirely co-opted into the official system. When it is also remembered that there was a very limited indigenous business class, the striking lack of opposition through extended periods of economic difficulty is not difficult to understand. The weak development of an autonomous middle class permitted a high degree of apparent relative autonomy of state action in the years following independence, as demonstrated by the surprisingly uncontroversial implementation of socialist measures following the Arusha Declaration. At the same time, there was an absence of pressure groups agitating for effective performance from the state, or of entrepreneurial groups capable of manipulating the state to support their own schemes of accumulation. The colonial authority itself had been relatively free of pressures from local interests. With independence, colonial power was replaced by a nationalist movement supported by the broad mass of workers and peasants, but unattached to the local concentrations of Origins of the Economic Crisis 13 economic power, which were largely non-African. The mass of the population could not be expected to play an active role in influencing policy-the peasantry expressed its discontent mainly by the passive resistance of withdrawal from the market, and workers could mobilize to defend a particular interest, but neither group could demand effective economic performance by the state. Nor did the parastatal managerial elite have any particular stake in efficiency as such, because for many years they could fiund their empires with borrowed capital. 14 Economic Strategy and Structural Adjustment in Tanzania 2 The Politics of Reform: Economic Disequilibrium as a Source of Change 2.1 Reform did not come in one discrete jump, but was the outcome of a somewhat tortuous process over the decade of the 1980s (see Mans 1994 for a systematic review of the steps in structural adjustment). The changing stance of the government can be traced from the denial of the need for reform in 1979-80, to the acceptance of some modest reforms in the Structural Adjustment Program of 1982 and the agricultural policy adopted at the same time, to the agreement of the Economic Recovery Program with the IMF and World Bank in 1986. During those years, there were contradictory movements in policy, and the process was described as "ambiguous" in its intent as late as 1986. The distance of the movement in policy attitudes can be seen by comparing the reports of two presidential commissions, one reporting at the beginning of the 1980s, the other at the end of the decade. The Mbylini Report on Agriculture (Professor Simon Mbylini was assistant to the president, and later principle secretary, Ministry of Agriculture), prepared in 1982 (Tanzania 1983), made only very subtle arguments in support of liberalized marketing, arguing the half-truth that local trade in food should not be restrained because it was a long-standing, "traditional" activity. In contrast, the Nyaribu Commission (Charles Nyaribu, retired governor of the Bank of Tanzania, and later ambassador to Washington) on the financial system, reporting in 1990, felt free to support thorough financial liberalization. In both cases, the commissions were drawn from the government and party establishment, and the members had similar backgrounds, but over the intervening years there had been far-reaching changes in what was understood to be politically acceptable. 2.2 Reform was supported from the early 1980s by many Tanzanians, including numerous public officials, in light of the obvious debacle of the established policy regime. Reform, however, represented a defeat for the Tanzanian political system. The party leadership, finding it no longer possible to command the resources required to sustain the system, had to accept a retreat. Nevertheless, the liberalization of imports and other initial reforms implemented in the last year of Nyerere's presidency (1984-85) could be also interpreted as temporary, tactical maneuvers, which relieved the leadership of the need to either confront the systemic failures of the previous decade or to fashion a comprehensive reform program and spell out its political implications. The attempt to mold the pattern of economic growth and to shield the system from the impact of economic decline had failed, but the extent of that failure was not fully recognized by the party leadership (Nyerere remained chairman of the party until 1987, two years after retiring as president). In that 15 16 Economic Strategy and Structural Adjustment in Tanzania period, a gap opened between the rhetoric and program of the party, which maintained the commitment to interventionism, and the policies pursued by government as it moved further down the reform path. 2.3 Acceptance of economic reform took many years. It started with the small, pragmatic steps required to manage an impossible economic situation, and moved on to more ambitious systemic reform, but only slowly recognized the deeper implications of piecemeal reform.16 Certainly the change of president in 1985 was significant. The Arusha Declaration was President Nyerere's creation, and he has never backed away from his socialist commitment, while President Mwinyi had no strong identification with any particular position. Nevertheless, there were striking continuities in political leadership. Ministers and senior officials from the 1 960s generally continued in office as long as they were physically able to do so. 2.4 Given one-party politics (until very recently), there was little open debate to serve as a measure of public sentiment that might require response from political authorities. During the period in which there was stiff official resistance to exchange rate adjustment and price decontrol, there was a concern among the political leadership about negative popular reaction in light of reactions to price decontrol elsewhere in Africa. But given the lack of hostile reaction to liberalization when it was eventually implemented, it seems likely that such a political assessment was wrong. It was evident in the early 1980s, however, that there was a significant current of opinion that was at least deeply cynical regarding the established economic policies.1" 2.5 Tiue attitude toward a change in policy regime must be influenced by interpretations of the nature of the crisis of the old regime. Even after more than a decade of reform, however, no consensus has emerged regarding the origins of the crisis. There are at least three positions implicit in discussions of public policy in Tanzania: 16 Nyerere retired from the presidency in 1985, when he did not seek reelection, and from the leadership of the party two years later. In his last year in office, the government acknowledged that it would have to accept substantial reforms in order to negotiate renewed support from the Bretton Woods institutions and the bilaterals aligned with them. That was a pragmatic response to the realpolitik of the aid process by a government in desperate need of additional foreign exchange support rather than the adoption of a new strategy. During Mwinyi's first term, policy positions adopted by government increasingly implied a new development philosophy, but this was not spelled out in a systematic fashion. 17 This conclusion is based on the personal observation of the author at that time-obviously not very scientific evidence. One indicator was the prevalence of a sort of ironic humor: black market prices were known as party prices; when questioned during the NESP (National Economic Survival Program), the popular response was to claim to be implementing PESP (Personal Economic Survival Program); the importance of 'technical know-who in acquiring scarce commodities was emphasized. That all having been said, it is striking that even with the advent of multipartyism in 1993, and the growth of an increasingly critical and investigative press, it seems likely that the CCM could well continue to dominate national politics, at least througb the 1995 national elections. The Politics of Reform 17 * The stated view of many party and government leaders continues to be that the goals articulated in the Arusha Declaration were fundamentally sound, and the crisis is to be explained by a combination of implementation errors, perhaps resulting from hastiness, and exogenous shocks. That is, there is a formal commitment to the old "vision," even while in the process of dismantling the system that embodied it. * The view of those who were fundamentally skeptical about the old system and were committed to a market-oriented, mixed economy-by the early 1980s, for example, most academic economists subscribed to that position, along with supporters of some of the new parties (perhaps Edwin Mtei, leader of Chadema, offers the most explicit support for privately based, market-oriented development). They would view the failures in the old system as inherent. Support for the market and private activity, however, does not necessarily imply unqualified support for a liberal economic system; industrialists support protection and indigenous businessman seek support in strengthening their position with foreign and minority community business interests. - There is also a radical perspective, which is deeply suspicious of the distributional and other social implications of liberalization, but at the same time is critical of the failures of the old system, particularly its failure to live up to its own rhetoric in relation to mass participation in the development process (it is sometimes forgotten that the most active criticisms of economic policy in Tanzania during the 1970s came from the left, from such writers as Coulson, Shivji, and von Freyhold). 2.6 One scenario for reform is for drastic political change to precede economic reform. This might appear the most propitious for success. The new political broom might be expected to sweep clean, and political change could generate a new "vision" of society to guide and justify changes and generate political momentum in support of changes that threaten vested interests. This is not what happened in Tanzania, nor is that necessarily a bad thing. Although a priori reasoning supports the virtue of combining perestroika with glasnost, experience in the CIS suggests that it may be quite difficult to fashion a new political system, while at the same time managing economic transition. An existing political leadership may be effective in carrying through a profound economic reform while maintaining its hegemony, although convoluted efforts may be required to imply ideological continuity." This might be described as the Asian (Chinese, or more recently, Vietnamese) model of transition, a model that allows the state apparatus to intervene to limit economic and social instability in the transitional process. The Tanzanian experience had some points of similarity with the Asian model: reform was carried through with substantial continuity of the political regime. Reform of the economic and social systems, however, was not primarily led by government initiative (see Van Arkadie 1989). The political authorities essentially responded to the initiatives of other actors in the economy who, in the face is In Vietnam, in the formulation of doi moi, the party went to great efforts to avoid a commitment to the market or a mixed economy, using instead the term *multisector- or a *multicomponent commodity economy.' See Ronnas and Sjoberg 1991. 18 Economic Strategy and Structural Adjustment in Tanzania of a decade of deepening economnic difficulty, adjusted their behavior and changed their roles. Economic change preceded policy reform-successful policy often involved the accommodation and orchestration of trends and initiatives from outside government. 2.7 Thus the political economy of reform was conditioned by the process of economic and social change provoked by the conditions of economic crisis. The first steps in the adjustment process in Tanzania were found in the survival strategies of the various actors in the economy- peasants, urban dwellers, businessmen, and the servants of the state itself-in the face of the breakdown of the existing order. State employees increasingly had to maintain their living standards through moonlighting or the pursuit of "informal" incomes, sometimes derived from the use of their public office. In a poorly functioning political and economic system, those requiring services from state institutions often found it easier to buy favors from functionaries than to press for a more general improvement in performance. 2.8 With mounting economic disequilibrium, by the beginning of the 1980s the economy had taken on many of the characteristics of a supply-constrained economy. Under conditions of extreme scarcity of foreign exchange, bureaucratic allocation determined the availability of inputs, and in principle most commodities were distributed to the end consumer under a rigid system of price controls rather than through a market-clearing price system. An irony of the period was that the materials allocation system of the early 1980s was developed not through the agency of a Tanzanian planning body, but through the allocations of import support by bilateral aid donors, who found themselves deciding whether toilet paper, textbooks, drugs, or soap would be available. 2.9 The government struggled with fiscal imbalances fueled by a cumulative spiral of decline. Efforts to control prices, including the exchange rate, led to diversion of income flows into the parallel economy, weakening the tax base. The erosion of the real salaries of public servants caused by tightening resource availability undermined the probity and effectiveness of tax administration, and further shrank the revenue base. The persistence of fiscal and monetary instability was fed by the lack of effective instruments to tax the growing parallel economy and to exercise financial control over public enterprises. 2.10 With the decay of the system of government control, the informal rules of the game changed, and activities that had been minor became more central to the economy. Just prior to the implementation of the Economic Recovery Program, parallel activity competed with the official economy in importance in many sectors (see Maliyamkono and Bagachwa 1990). The parallel economy was a critical phenomenon, feeding on the failure of the old system and contributing to the genesis of the new. The market economy asserted itself through the agency of the parallel economy as the planning system decayed. The black marketeer, the criminal of the old system, can be seen as the hero of the new. The black market not only acted to break the old system, but also began to provide a working basis for a new system of distribution. 2.11 In analyzing this process, two kinds of parallel economic activity can be distinguished. Some activity was genuinely "parallel" to the official economy, made up of a circuit of activity The Politics of Reformn 19 totally separate from the official economy, such as small-scale agriculture and petty craft production.'9 The failure to "capture" this activity in the official circuit reflected the weakness of the state apparatus and the unrealistic nature of ambitions to extend state controls (for example, over local trade in food), but had no particularly negative consequences for the internal working of state institutions. The thriving market in smuggled exports, which arose in response to an overvalued exchange rate, was more ambiguous in its effects. It subverted the operation of major policy instruments (such as foreign exchange controls) and eroded revenue. Other parallel activities were an outgrowth of the official system, ranging from black marketeering, where scarce products were sold under the counter above the official prices, to the diversion of state property to private uses. These activities might better be described as "orthogonal" to the official economy, in that they depended on using access to the state to generate private income and set in motion changes in the state system. 2.12 Breaking rules and the allocation of favors are usually necessary for any rigid bureaucratic system to operate. Survival in a supply-constrained economy depends as much on technical "know-who" as technical know-how. Nevertheless, black marketeering is often not only against the formal rules, but may also be socially undesirable when it subverts a system of distribution that, even if ineffective, provides the public with certain rights of access to scarce commodities. For the consumer in the queue, black marketeering, which diverts already limited official stocks to benefit privileged groups in the system, was unacceptable. The resulting cynicism in Tanzania led to the popular description of black market prices as "party prices" in the early 1980s. This provided a popular basis for official campaigns against black marketeers,20 but was also a step in the erosion of the system's authority. 2.13 By the beginning of the 1980s, the way the Tanzanian system operated was thus at increasing variance with the vision set out in the Arusha Declaration. The growing gap between the ideal and reality led to a loss of public support for that vision.21 Distress and cynicism about the operation of the system could either engender support for change (substitution of one vision for another) or support a reassertion of the values of the system to bring ideals and reality into 19 Thus, before the reforms that ended the official NMC monopoly on grain trade, much of the trade, both internal and to neighboring economies, was in the hands of private grain traders. 20 In 1983, when the Tanzanian economy was in the depths of crisis, the campaign against the walanguzi (black marketeers) virtually closed down the economy in some parts of the country, because by then the parallel economy was the only effective marketing system in many parts of the country, but the campaign was initially quite popular. Similarly, in the early stages of Soviet reform, there was a strong popular distrust of the "mafia,* which somewhat surprisingly seemed to be shared by Western donors, who claimed to be taking steps to ensure that aid went through 'proper' channels to the 'right' destinations. 21 And, of course, there is no way of determining the depth of support and understanding of the Nyerere philosophy among the mass of the population. Most observers agreed that there was widespread support both for him as a leader and for the ruling party; indeed, in the context of multipartyism, CCM still appears to enjoy the hegemony of the political system. 20 Economic Strategy and Structural Adjustment in Tanzania closer proximity. It was not unknown to find both these positions held by the same person, as a contradictory response to difficult times. 2.14 The genesis of market activity in the parallel economy leaves a negative legacy. The weakening of the morale and probity of public servants may erode the capacity of the state to undertake the tasks allocated to it in the reformed economy, while the experience of black markets generates unfortunate business ethics for the emerging market system. While parallel economic activity may make a positive contribution to the transition, it also involves a disrespect for the law, giving rise to a pattern of individual behavior that carries over into the reformed system a disregard for the legal requirements of business and a disrespect for normal fiscal obligations (for example, in Tanzania, lessons learned in avoiding the exchange control system have been readily applied to customs avoidance and manipulation of reimbursement of import support allocations). Businessmen became wise in the avoidance of all forms of government controls and impositions, which were implemented with increasing feebleness by a decaying state apparatus. The resurgence of private economic activity thus did not resolve the fiscal cfisis. 2.15 The weakness of the negative political reaction that followed liberalization is consistent with the view that liberalization essentially accommodated changes already under way in the economy. Even before the implementation of substantial measures, support for reforms was evident among the bureaucracy and technocrats. University economists, for example, became increasingly open in their support of the need for liberalization, and there was widespread recognition that something would have to be done to improve parastatal performance. Why, then, was the government's opposition to liberalization so persistent? Some of the opposition to selected reforms seemed to involve a degree of misunderstanding. In a number of public statements, President Nyerere argued that devaluation would make imports dearer-that the exchange rate would have a significant impact on the external terms of trade. Resistance to change also had an idealistic element. 2.16 The Arusha model was intended to create a more equitable development pattern through politically determined distribution of income and property. Critics of liberalization have compared conditions following liberalization either with those prior to the crisis or to an idealized view of what the system was intended to deliver, rather than the reality of 1983 under the full impact of economic crisis and the failure of the system to respond when practice fell far short of the ideal. But before the full impact of crisis, the system was progressing toward some of its goals, and it is perfectly understandable that the architect of that vision should remain committed to it. Liberalization involved much more than the adjustment of economic policy instruments, because it set in motion a process that would lead to the dismantling of the system, so that opposition to liberalization was a rational response for those who remained committed to the socialist ideal, just as support for liberalization was a rational choice for those who wished to have done with that particular vision of Tanzania's future. 2.17 There were interests as well as ideals at stake. As the economic crisis deepened, the parastatal sector became increasingly dependent on privileged access to resources. These The Politics of Reform 21 privileges involved monopolies in internal markets and external trade; subsidies; and privileged access to scarce foreign exchange, donor import support, and credit from the state-owned banking system. Liberalization inevitably eroded these privileges and placed in question the sustainability of the parastatals and the positions of parastatal managers, whose perquisites often liberally supplemented austere formal salary scales. 2.18 The most obvious beneficiaries of liberalization have been those able to take advantage of trading opportunities in expanding free markets, and the most obvious losers were those who benefited from privileged access to scarce, state-controlled commodities and undervalued foreign exchange. In some cases, however, these were the same people. Given that those most subject to a relative loss in income as a result of reform were in the state apparatus, and traders and rural producers who were potential gainers were not effectively organized politically, the pressure for structural adjustment generally came from donors, althouigh large segments of the Tanzanian population gained from reform. Only a minority of the workforce was employed by the government or in public enterprises. Much of the economy retained nonproletarianized, rural, labor surplus characteristics. Insofar as the unreformed system benefited those in official employment, it was supporting a small minority. This contrasted with Eastern Europe, where there was fUll employment, almost entirely in the state system. 22 Economic Strategy and Structural Adjustment in Tanzania 3 The Reform Process Resistance to Change and First Steps to Reform: 1979-84 3.1 Faced with a renewed crisis, over the period 1979-84 the government of Tanzania attempted a number of solutions before moving decisively in the direction of liberalization in the budget of 1984. In 1979 the government had agreed to a standby arrangement with the IMF to alleviate internal and external imbalances. A second standby agreement in 1980 fell apart when targets were not met. The government went public in rejecting proposals set forth by the IMF, setting in motion an extended period of acrimonious relations with the Fund.22 Despite frequent and protracted negotiations, no further agreement was reached with the IMF until 1986. 3.2 When it failed to agree to a Fund program, the government attempted to reassert control over the economy and promote exports through bureaucratic means and exhortation, without adjusting the exchange rate, particularly under the National Economic Survival Program (NESP), 1980-82. Targets were not met. Exports stagnated and external assistance declined, so that the import squeeze tightened. Deterioration in the fiscal balance became increasingly alarming as the government defended the existing exchange rate and controlled price structure through subsidies, while losing revenues through increased income generation in the parallel economy. 3.3 In 1981 the World Bank funded the Tanzania Advisory Group (TAG) to provide assistance to the government in framing a structural adjustment program. The Structural Adjustment Program, which was adopted in late 1982, included steps to restore macroeconomic balance, improve incentives, and move toward liberalization. Commitments were made to improve incentives to producers of export crops, reduce the government deficit and the growth in the money supply, reduce imbalances by shifting resources from public investment to recurrent uses, and improve the efficiency of parastatals. Some steps were taken that involved difficult choices: 22 Some aspects of the deterioration in relations between the IMF and Tanzania are in the public record. For example, the planning minister set out the IMF proposals and the government's response in the Bunge (see Van Arkadie 1983). Other aspects were not made public. There was, for example, an unfortunately abrasive meeting between President Nyerere and an IMF delegation, which led to a change of leadership of the IMF team, and a change of Finance Minister, Mr. Mtei. Edwin Mtei, who was later nominated as executive director of the IMF and is now leader of an opposition party, Chadema, has since made it clear that he supported the IMF proposals. 23 24 Economic Strategy and Structural Adjustment in Tanzania for example, stopping implementation of a number of important investment projects. Some modest steps were taken in the direction of liberalization, such as internal food trade and private transport. Nevertheless, the government was unwilling to adjust the exchange rate significantly. TAG had proposed a devaluation that was considerably less than the amount the IMF was seeking, but still more than the government was willing to accept. Moreover, in important areas such as parastatal reform, the program was little more than a statement of intent. The macroeconomic targets in the program were predicated on its acceptance by the donor commnunity, resulting in increased external finance and a shift in donor funding from project to import support. 3.4 Although there was a continuing shift in the composition of aid, with the balance shifting toward import support, and some improvement in the fiscal balance, the program did not prove acceptable to the IMF and the World Bank. Unlike the mid-1970s, this time the required increase in external funding did not materialize. Nor was the program sufficient to turn the economy around without such funding. Although the steep decline in agricultural output bottomed out, the collapse in industrial output continued, and there was a further deterioration in export performance. In these areas, the TAG exercise failed, but it did place a number of crucial issues on the internal agenda of government policy debate, including exchange rate adjustment and parastatal reform, and it sustained a positive dialogue between the Tanzanian government and part of the donor community,23 which became an important part of the process leading to reform. 3.5 Along with the Structural Adjustment Program, the government introduced a New National Agricultural Policy in 1982 (see Tanzania 1983), which affirmed a commitment to improved agricultural incentives, restored agricultural cooperatives, reduced the role of the crop authorities, made a commitment to increase investment in agriculture, encouraged an expanded role for large-scale private agriculture, and accepted the role of the traditional local food trade. 3.6 At the top of the political system, however, there was strong resistance to devaluation and liberalization of the trading system. The ambivalence concerning the modest steps toward liberalization was demonstrated when an intensive-if brief-national campaign was launched against the walunguzi (black-marketeers) in early 1983. This campaign was supervised by the late Prime Minister Sokoine. It would appear that it was not so much that he was committed to the regulatory structure (soon after he was associated with pragmatic steps in the direction of liberalization) as he was alarmed by the lack if discipline and the cynicism associated with the 23 Ambassador Michanek made great efforts to keep the dialogue alive, particularly with the Nordic countries. Of course, it can (and has) been argued that the sympathetic stance adopted by the Nordics may have postponed refomis that might have been imposed by a more united donor stand in support of tough conditionality. Against that view, the Nordic govemments took great care to make clear to the Tanzanian government that they were unprepared to provide an altemative to the multilaterals and that it was essential that Tanzania should come to an accommodation with the Fund, leading to access to Fund finance and World Bank structural adjustment lending. It could also be argued that in 1982-83, economic conditions in Tanzania became so desperate that any further downward push would have led to social and political collapse rather than reform. The Reform Process 25 widespread subversion of the law. In retrospect, the 1983 campaign can be seen as the last major effort to reassert control over the growing parallel economy. The futility of the effort was demonstrated in parts of the country such as Kagera, where the campaign was implemented with particular vigor, and it became evident that to close down the parallel economy was to close down the economy itself.24 It was in 1983 that economic difficulties in Tanzania became most extreii-e. Goods disappeared from the markets and there was a lack of commodities in agricultural areas, rendering financial incentives almost irrelevant. 3.7 Even if efforts to rein in the parallel economy had some temporary impact, vigorous growth in the parallel economy continued, and, as became evident later, this included the growth in exports bypassing the official system. This grew from the government's unwillingness to make sufficient price adjustments. Although government would boost farm prices, it maintained price controls, was unwilling to adjust other prices, and made only minor adjustments to the exchange rate. The gap between official prices and parallel prices (including foreign exchange) widened, increasing the incentive to trade in parallel markets. 3.8 Under conditions of foreign exchange scarcity, and given the priorities of the official foreign exchange allocation system, there was a virtual famine of consumer goods in Tanzania by the early 1980s. Faced with extreme shortages of commodities, which placed tight limits on consumption and made it increasingly difficult to operate any productive enterprise that lacked access to donor commodity support, the government had to seek ad hoc solutions to its problems. One such step was to allow those with access to foreign exchange to import pick-up vehicles without having to channel their funds through the rigid exchange control system. It was hoped that this measure would alleviate the severe transport constraint the economy faced. This was successful, and it became a harbinger of the future pattern of liberalization. 3.9 In the past, the development strategy had been inward-looking and exports were discouraged by weak incentives, although autarky was neither attempted nor achieved. There was a decline in international trade because of the balance of payments crisis, and this had a severe effect on an economy still structurally dependant on imports. It was recognized that there was an increasingly difficult structural problem in the country's balance of payments (developed in Green, Rwegasira, and Van Arkadie 1980). Exports financed too few imports to fuel the ambitious investment program and provide the imported inputs to utilize capacity. Nevertheless, exchange rate adjustment was not seen as an appropriate response at the time. Resistance to Exchange Rate Adjustment 3.10 Resistance to reform in the early 1980s was particularly notable in exchange rate policy, with strong resistance to IMF and World Bank advice regarding the need for substantial devaluation. This was such an important issue that it needs to be explored in some detail. While 24 It was at this time I heard that in my grandmother-in-law's village, even salt disappeared, reflecting an extraordinary collapse in the trading system. 26 Economnic Strategy and Structural Adjustment in Tanzania there would now probably be a broad consensus that the failure to adjust exchange rates in the mid-1970s was a bad mistake, it might be useful to understand some of the reasoning and historical background behind the resistance.25 3.11 During the days of the East African Currency Board, currency in circulation had been adjusted in response to fluctuations in external reserves, with only a very conservative provision for a fiduciary issue. In an era of fixed exchange rates (the 1967 sterling devaluation was only the second sterling-dollar adjustment since World War II, the first being the Cripps devaluation of 1949), the Currency Board system underwrote the stability of the extemal exchange rate by its conservative, quasi-gold-standard monetary policy. International orthodoxy leaned in the direction of fixed exchange rates until the dollar crisis of 1971. Even in the early years of central banking, monetary policy was cautious, inflation modest, and the exchange rate not obviously out of line with purchasing power parity.26 Indeed, except for the discussion regarding the appropriate response to the 1967 sterling devaluation, exchange rate policy had not been a feature of the East African policy debate. The kinds of issues that traditionally concern the IMF were not a source of policy controversy for some time, as evidenced by the benign relationship between the IMF and Tanzania, even during the period when Tanzania had access to IMF facilities following the first oil shock. 3.12 The case for not pursuing a flexible exchange rate policy in the Tanzanian context went as follows. * Tanzania was a price-taker in most export and import markets, so that extemal terms of trade were not influenced by the shilling exchange rate; exchange rate adjustment would only affect intemal relative prices. * The intemal prices of most export crops were set administratively, so that in principle incentives could be adjusted selectively, increasing prices for commodities with a good expected supply response without generating windfall gains to exports with a more inelastic supply. * Govemment was implementing a prices and incomes policy to restrain inflation, which would be difficult to maintain in the face of large exchange rate adjustments (and if 25 There is obviously an element of self-justification at this point, as in the role of economic adviser to the Planning Ministry in 1967, the author of this essay was party to the analysis that concluded that the East African currencies should not follow the sterling in devaluing against the dollar, and in so doing helped develop a line of argument that supported a fixed exchange policy throughout the 1970s. The case that resulting weaknesses in the incentive system were an important contributory cause of poor export performance is presented by Jennifer Sharpley (1985). 26 However, the potential for this to change with the creation of national Central Banks was obvious (see Van Arkadie 1965). The Reform Process 27 inflation were not restrained, the impact of exchange rate adjustment on relative prices would quickly erode)." 3.13 The two crucial assumptions implicit in this line of argument were that the government had the means available and the willingness to use them to fine-tune the internal incentive structure without using exchange rate adjustments, and that excess demand in the foreign exchange market could be handled by rationing (that is, exchange controls), without significant loss of efficiency. It became increasingly evident, however, that the apparatus of public agricultural marketing was not a policy instrument that could fine-tune anything, and the weight of its inefficiency was a prime cause of the erosion of incentives. The monopoly position of the crop authorities (and of single-channel marketing in its various manifestations) meant that both peasant and government lacked alternatives; the peasant had to sell through the single official channel (or smuggle), while the government had to fund its continuing operation, however inefficient, or countenance a disruption in crop marketing. As a result, the vast sums transferred from the government budget and the banking system to the crop marketing system-justified as a means of sustaining incentives-did little to improve farm prices, but instead disappeared into a bureaucratic black hole. 3.14 Although in retrospect it is clear that the failure to adjust the exchange rate and other market variables had severe political as well as economic costs, the use of administrative means to control the economy was sustainable as long as the degree of disequilibrium was not too large. This meant that foreign exchange available for allocation was sufficient to sustain a reasonable level of capacity utilization; goods were available in sufficient quantities to distribute under price control without provoking a generalized black market; and subsidization of parastatals could be handled without fueling excessive inflationary pressures. By 1981 those conditions no longer held. As disequilibrium increased, parallel markets in foreign exchange and imported commodities developed, and individuals and institutions with access to foreign exchange at the official rate were able to reap increasingly generous rents. This, of course, led to the development of strong interests within the government and parastatals in the maintenance of an overvalued exchange rate. 27 The case against the use of exchange rate adjustment was later carried one step further by Ajit Singh, during the debate over the Structural Adjustment Program in 1982. In his usual persuasive style, he argued that export supply was likely to be largely price inelastic and it would probably be very difficult to adjust the real wage rate through exchange rate adjustment. Apart from any detailed weaknesses in Ajit Singh's position, the overall flaw is that it implied that the existing exchange rate would always be the best of all possible exchange rates, and there was a strong a priori case for not making any use of a key macroeconomic policy instrument. 28 Credit for initially making the public case regarding the weaknesses of the agricultural marketing system must go to Frank Ellis, who produced a series of careful studies of the operation of the crop authorities at the Economic Research Bureau, University of Dar es Salaam (see Ellis 1982, 1983). 28 Economic Strategy and Structural Adjustment in Tanzania The Reform Process: 1984-93 3.15 Two years before agreement on the Economic Recovery Program with the IMF and the World Bank, the government began to explore its own route to reform. In the 1984 budget a series of measures were introduced that incorporated an important shift in policy. Steps included: * import liberalization through an "own account" import provision * substantial reduction of the number of items covered by price controls, especially imported items * removal of food subsidies, increased liberalization of food markets, and rationalization of the food marketing price structure * the removal of agricultural input subsidies and the reflection of the new level of input costs in producer prices * an enhanced commitment to defend the real level of agricultural prices * first steps in cost-sharing in the social services. 3.16 The most important of these measures in immediate economic impact was the "own account" import scheme (see Tanzanian Economic Trends 1988). The own account import provision allowed those with access to foreign exchange outside the official exchange control allocation system to use the funds to freely import a wide range of goods. The initial rationalization for this policy was that residents might have access to such funds from relatives abroad, from savings while studying overseas, and from other similarly innocuous sources. The provisions had an immediate and strong impact on the import of incentive goods and transport equipment, significantly changing their availability in local markets. The size and persistence of the import flow funded in this way, which far exceeded anyone's expectation, soon made it clear that the funds were being replenished from sources other than those initially envisaged-that is, from the flow of export earnings that were bypassing the official exchange control system. Own account imports were being funded by parallel market exports.29 3.17 This measure was important for a number of reasons. First, it was of macroeconomic significance as an initial, and substantial, step in relaxing the import constraint that had throttled the economy. Second, it implicitly accepted a dual exchange rate, because transactions supported by this trade were at a parallel exchange rate. Third, it involved an 29 Numerous efforts have been made to estimate the size of the parallel market foreign exchange earnings implied by the level of 'own account' imports by those monitoring the economy (including Benno Ndulu at the University of Dar Es Salaam and Enrique Rueda Sabatier in the World Bank office). Given the nature of the activity, it was not directly observable, while even data on the import side were not very reliable. Nevertheless, it seems that figures for parallel market foreign exchange earnings spent on own account imports could well have been in the range of 70-90 percent of official, recorded export earnings! The Reforn Process 29 implicit acceptance that there was a great deal of unofficial-and illegal-activity in the Tanzanian economy, which could make a positive contribution to economic performance if allowed to do so. 3.18 The change in the government's policy stance, as exemplified in the 1984 budget, took place while President Nyerere was still in office. The steps toward reform could be seen as a pragmatic response to the need for action to revive an economy in deep crisis, recognizing the need for immediate measures to loosen the constraints on economic activity and the realistic recognition of the need to restore relations with the donor community, a need that became urgent with pressure from donors who had maintained support, but were becoming increasingly alarmed by the deterioration in the economy. The policy change did not represent a fundamental shift in the ideological stance of the government, and the top leadership was not entirely happy with the measures adopted. Indeed, Nyere;e continued to express concern about the distributional implications of trade liberalization after he left government in 1985. 3.19 Whatever misgivings may have existed in government, once begun, liberalization gathered momentum swiftly and met with little resistance. This is not difficult to understand. The depth of the preceding crisis meant that by 1983 there was widespread hardship. The crisis was reaching such great proportions that even those protected by access to rents were being badly hit by the negative effects of the dislocation in the economy. Critical parts of the initial liberalization legalized activities that were already established, and these measures mainly had the effect of reducing petty harassments, such as roadblocks checking food movements, and increasing ease of access to commodities for the mass of consumers. There were, of course, objections voiced, comparing liberalized prices with the previous controlled prices, but the force of such criticism was diminished by the knowledge that in the preceding period many goods had either not been available at all, or were purchased by the end consumer at black market prices no lower than the new liberalized prices. 3.20 In 1985 Nyerere was succeeded by President Mwyni, although he retained his position as chairman of the party during most of Mwyni's first term.30 It was under Mwyni, therefore, that the Economic Recovery Program (July 1986 to June 1989) was adopted and agreement was reached for support with the IMF and World Bank. It is true that the movement toward the reform program was begun under Nyerere, and that even with the adoption of the reform program, the rhetoric of the party remained socialist. Nevertheless, Mwyni sought to define his own image as more pragmatic than Nyerere, and as his administration proceeded, the socialist content of the government program diminished. 3.21 Under the Economic Recovery Program, in addition to heightened efforts to tighten the budget, a number of innovations were made in the direction of financial orthodoxy. Substantial and continuing adjustments were made to the official exchange rate, which 30 A presidential termn runs for five years, and under current provisions a president is limited to two terms in office. President Mwyni therefore completes his period in office in 1995. 30 Economic Strategy and Structural Adjustment in Tanzania significantly narrowed the gap between the official and parallel rates. Interest rate policy was radically adjusted, and the objectives of moving to positive real interest rates and tight ceilings were adopted for credit expansion. The price control system was further dismantled, and additional steps were taken to rationalize and liberalize the grain marketing system. Incentives for exports were enhanced. In the 1988 budget, import tariffs were simplified and rationalized. 3.22 In response to the program, there was a large expansion in extemal financial support, including two Fund programs, several sectoral adjustment credits from the World Bank, and an expansion of bilateral commitments, particularly from donors who had held off during the period in which there was no Fund program. Some measure of debt relief was offered under Paris Club auspices, although Tanzania remained heavily indebted, particularly to the World Bank, reflecting the generous but not very productive boom in World Bank finance in the 1970s. An important aspect of donor support was the considerable shift of funding from projects to balance of payments support. This had already begun under bilateral import support programs, which continued on an expanded basis, and beginning in the second year of the program there was support for the introduction of an Open General Licence System (OGL), which further increased the flexibility of the foreign exchange allocation system at the margin. The Economic Recovery Program was followed by an Economic and Social Action Program (ESAP), initiated in 1989, which incorporated measures to revive delivery capacity and extend user charges (cost recovery) for social services. 3.23 The subsequent movement of policy toward continuing liberalization of the foreign exchange regime continued through expansion of the coverage of OGL, the disappearance of import controls, and the movement toward a unified exchange rate, which was achieved in 1993. There has been total liberalization of intemal trade and nontraditional export trade, but much slower liberalization of traditional agricultural exports. Generous incentives were introduced for foreign investors. Price control has ended. Financial sector legislation has been enacted, opening up the banking system to private banks, and by 1993 the first private banks had opened.3' Donors have provided program support and substantial finance to rehabilitate basic infrastructure, particularly the road system, which had deteriorated badly. 3.24 This chapter is not primarily concemed with the economics of adjustment policies, and therefore discussion of the technical merits of the timing and dosage of particular economic measures is slight. Nevertheless, there are two elements of the interplay between the adjustment in economic instruments and institutional factors that require comment here. 3.25 In the early stages of the dialogue over reform, external advocates of orthodox economic measures probably placed too much emphasis on the appropriate levels of the needed 1 Initially, Meridien and Standard were established. In April 1995, the Bank of Tanzania had to take over control of Meridien, which was faced with liquidity problems from its overseas dealings. At the time, it looked highly probable that the Bank of Tanzania would find it necessary to meet Meridien's obligations to depositors if therm were a shortfall-private busines is not always immune from some aspects of the soft budget constraint." The Reform Process 31 policy instruments, and too little emphasis on the institutional reforms necessary to make the instruments effective. Large adjustments in interest rates could be expected to have great effect in a financial system (both banks and parastatal borrowers) not subject to a hard budget constraint. The level of official prices for the purchase of farm produce was not very meaningful to farmers who were subject to long delays in payment-or even nonpayment-by ineffective marketing institutions. The notional price of foreign exchange allocated under import support programs was of limited interest to recipients of foreign exchange allocations who did not pay the local currency debt for the foreign exchange received. In all these cases, a change in the institutional arrangements to make the price system operational was a more important initial task than getting the notional prices 'right." 3.26 A second general weakness was the failure to take account of the effect of economic adjustment measures on government performance. Insofar as exchange rate adjustment and other liberalization measures were intended to shift resources in the domestic economy from nontradables to tradables, the shift in incentives implied a reduction in the real wage bill for state employees. To the extent that liberalization measures made the activities of selected state employees redundant, that would be of no particular consequence. And, of course, if effective civil service reform were to be implemented, the decline in the real wage bill need not necessarily result in a commensurate decline in real wage levels if numbers in state employment were to be reduced. In practice, adjustment measures led to sharp declines in the real wage levels of public servants undertaking tasks necessary to the economy, and the failure to consider the implications of that decline was a glaring weakness in standard adjustment packages. Civil servants were no less sensitive to incentives than peasants (see Van Arkadie 1986). The Slow Pace of Parastatal Reform 3.27 In contrast to the success in liberalizing the macroeconomic policy regime, reform of the parastatal sector moved slowly. Market reform requires not only that prices be right, but also that they be operative-that is, act as working constraints or incentives influencing actors in the economy. Realistic prices do not constrain enterprise behavior if the firm can avoid the consequences of losses (the soft budget constraint). The hard budget constraint is needed both to spur enterprise efficiency and to reinforce macroeconomic stabilization. The difficulty of imposing hard budget constraints at the early stages of the reform process was partly a question of political will. The soft budget reflected political unwillingness to enforce a harsh pace of adjustment, and the government was either a lazy shareholder or gave higher priority to objectives other than enterprise efficiency. In areas where there was a public sector monopoly, such as crop marketing, the government had to consider the impact of a parastatal collapse on the users of its services. 3.28 The situation was improved by two reinforcing movements. Over a decade, the monopoly position of the parastatals was dismantled. This began, as did most steps in the reform process, as a tentative and pragmatic response to changes already under way in the economy. The 32 Economic Strategy and Structural Adjustment in Tanzania legalization of private trade in food and private urban passenger transport accommodated changes already in progress. The "own accounts" import provision eroded the parastatal dominance of the import trade. The success of private trade and transport in easing the extreme scarcities from 1985 onward made these changes popular and quickly created new interest groups in the private trading sector. 3.29 The other source of change was the extemal pressure on the government to improve the fiscal balance and to impose monetary constraint. Under donor pressure, beginning in 1985 the government moved to reduce budgetary subventions to the parastatals, but financial discipline was cushioned by access to other sources of finance. For example, when budgetary support was withdrawn from agricultural marketing institutions, it was quickly replaced by credits from the state-owned banking system, which was not subject to effective financial discipline, but was open to political pressures. Public enterprises were financed through growing interenterprise debts. And in the absence of a strong assertion of ownership interest, wage and salary bills were in some cases met by cannibalizing the real assets of enterprises. 3.30 A slow pace of change was particularly evident in dismantling the single-channel monopolies in trading export crops. The slowness of reform in this area is somewhat surprising. The weakness of the marketing system was well documented and reform was presumably in the interest of the largest single economic group in Tanzania, the small farmers. Nevertheless, there were strong bureaucratic interests embedded in the marketing system, both at the national level, in the marketing boards, and at the regional and local level, in the reconstituted cooperatives. For example, in the cotton zone, the cooperative union had a clear interest in maintaining the monopoly position of the union-owned cotton ginneries. Moreover, the ending of the single- channel marketing monopoly was not placed very high on the agenda of donor conditionality in the first round of structural adjustment lending. Major steps in the introduction of competitive marketing in traditional agricultural exports were therefore delayed until late in the reform process. It was substantially implemented in 1994, but only after most other aspects of the international trade and payments regime were fully liberalized. 3.31 The difficulty of enforcing discipline on public enterprises provides the strongest argument for privatization. Privatization in Tanzania has been accepted as an objective, but implementation has been very slow,32 partly because of opposition from vested interests in the parastatals. An additional political difficulty resides in the limited supply of large-scale indigenous entrepreneurs. There is a weak national large-scale private sector, too poorly developed to take over large-scale public enterprises en masse. Although there are now many successful African businessmen, few have the experience and capital to take over a large parastatal. The wholesale transfer of public assets to foreign or minority community ownership could prove politically 32 Currently there has been some acceleration in the process through the mechanism of the liquidation of the assets of firms in debt to public financial institutions through the agency of LART, set up to carry out that task under the financial sector reforms. The Reform Process 33 controversial; the desire to reduce foreign and minority community dominance of the economy explained much of the popular support for public sector expansion (see Van Arkadie 1973), the expansion of the state enterprise sector reflecting an aspiration to increase national involvement in the large-scale economic activity rather than a doctrinaire commitment to public ownership as such. The fast growth of Asian and Arab mercantile activities under liberalization, and the implications of planned privatization, have led to a provocative debate about "indigenization," involving aspiring African capitalists and newly emerging political groupings in the mobilization of popular sentiment against restoration of economic power to minority communities. 3.32 Another difficulty is coming to terms with past investment mistakes. Many parastatals have little potential value as going concerns in a liberalized economy. They are often burdened with a heavy, foreign-exchange-denominated debt. While a reasonable economic case could be made that assets would be better transferred at a realistic market value-often liquidation value-so that they could be put to some alternative, productive use, public suspicion of sales greatly below historical costs is understandable. 3.33 Although the pace of privatization has been slow, the urgency of privatization in many sectors has diminished as the liberalization process has opened previous state monopolies to competition. Indeed, in areas such as crop marketing, there would have been no gain from privatizing a parastatal as a monopoly, and as crop marketing has been liberalized, there is no particular harm in parastatal marketing institutions competing against private traders, providing they do not require subsidies to survive. 34 Economic Strategy and Structural Adjustment in Tanzania 4 The Distributional Impact of Reform 4.1 Much of the international debate about structural adjustment has related to its distributional effects. The proponents of structural adjustment have had to respond to demands for a "human face" by exploring the "social dimensions of adjustment." Loosely structured criticisms of the distributional impact of policy over the past decade are often based on a comparison between the distribution of welfare following reform and the situation before the onset of crisis (for example, the mid-1970s) or the intended result of the Arusha strategy. Arguments that the poor have suffered are often based on the decline in real earnings of the employed labor force since the mid-1970s. 4.2 The short-term impact of reform on income distribution includes the following: * Compared with the situation before the introduction of the first reform measures (1983), the majority are probably better off, particularly in light of the great hardships experienced by almost all strata of society in the depths of the crisis. * Some income groups that suffered a severe decline in real income during the crisis have not done well during the reform period, particularly the formally employed. Their decline, however, was as much as result of crisis as of subsequent reform. Also, the formally employed were not among the lowest income groups. A decline in their incomes cannot be presented as an income loss for the particularly vulnerable. Groups that most obviously lost out from reform include those who had been able to defend their real income under conditions of crisis through privileged access to officially allocated resources. * There is some evidence in the GDP figures that there was a general shift in the internal terms of trade in favor of agriculture, which may be seen as an egalitarian shift. * The period of crisis and reform has provided the setting for fast accumulation on the part of an emerging property-owning class, which is creating a society with much more evident inequality, but not necessarily greater poverty. 35 36 Economic Strategy and Structural Adjustment in Tanania * The decline of the capacity of government to deliver social services, a result of both crisis and adjustment, has reduced coverage, which has had a differentially greater impact on the less privileged. 4.3 The crisis and the reform process resulted not only in changes in cash incomes, but also in the whole system of social service entitlements. An effort had been made to provide basic social services on a universal basis, but as an incomplete and underfinanced social service delivery system became strained, and even before reforms were introduced, access became differentiated on the basis of position in the system and income (for example, state employees began to demand payment for services that were, in principle, free). There has been a shift in public policy from free universal provision to cost recovery (user charges), which represents a significant shift from earlier ideology, but perhaps not so significant in its effects in light of the system's decline. 4.4 During the period of great scarcity in Tanzania in the early 1980s, access to commodities at official prices was an important source of benefit. In government service and parastatals, as wages were strictly controlled, fringe benefits (such as housing, transport, entertainment, and international travel) became increasingly important determinants of real income. It was through differential nonmarket entitlements that the greatest inequalities in the system emerged by the end of the 1970s. Losses in real income among groups in the public sector came as much from declines in privileged access to scarce commodities as from loss of salary (while those currently enjoying high incomes from public service have done so through access to perquisites of office, including allowances to senior officials, rather than from salaries). 4.5 One difficulty in assessing welfare changes during crisis and adjustment is that the strategies of actors in the economy change. It tells little about the real income of university teachers, for example, to note that real salaries have collapsed if the teachers have changed their own strategies by taking on all manner of new income-generating activities. Perhaps the most important impact of crisis and adjustment is the degree to which many Tanzanians have been forced to adopt new strategies of econornic survival. This has had both a negative (where people have been forced to great efforts to make a precarious and bare subsistence in the urban informal sector) and a positive side (where pressures for economic survival have uncovered entrepreneurial talents). 4.6 There have also been changes in the flow of private transfers through the extended family system. In the 1960s, transfers from public servants contributed through the extended family network to the welfare of those outside the official system. By the 1980s the bureaucracy had to seek sustenance in informal activities, and urban dwellers sought to supplement their incomes from the countryside. 4.7 Evaluation of the distributional impact of reform must also address the long-run consequences of the change in economic system. The Arusha strategy had included an explicit commitment to egalitarian income distribution, opposed the growth of an African capitalist class, and embraced a "basic needs' strategy for the provision of social services. For those who feel that The Distributional Impact of Reform 37 such goals were desirable and potentially achievable, the systemic change introduced through reform must be a source of disquiet. 38 Economic Strategy and Structural Adjustment in Tanzania 5 Sustainability of the Reform Process The Momentum of Economic Growth 5.1 The achievements of the economic reform program so far have included a revitalization of private economic activity as domestic actors in the economy have responded to new market opportunities. The positive response has been greater than suggested by conventional GDP data, which incorporate most of the official, public sector economy and underestimate the unofficial, private (often informal) economy. As it is the official economy that has borne the brunt of stabilization, and the private economy that has been the main beneficiary of liberalization, the GDP data have a downward bias in measuring recovery under structural adjustment.33 5.2 It seems unlikely that the reform program has had a negative impact on the poor. The real deterioration in the incomes of many of the poor was as much the result of the crisis that gave rise to the need for policy change as the structural adjustment policies as such. If anything, the policies have a pro-rural bias, and hit the formal incomes of the urban middle class particularly hard. Policy reforms have received more widespread popular acceptance than politicians expected, although there has been understandable public suspicion of some measures. 5.3 Despite these successes of reform, the hard economic benefits achieved so far have been less than projected (see Ratso 1991 for a negative assessment of Tanzanian economic performance under structural adjustment). A certain fatigue can be observed in Tanzania, both on the part of government and the donor community, because the economic results of policy reform have fallen short of expectations. If, on the positive side, evidence can be offered that the reform medicine is working, and a near moribund economy is showing signs of life, the recovery thus far is less than complete. The modest success achieved looks impressive against the sorry background of the depths to which the economy had sunk, but although markets are working and the life of 33 The official GDP data suggest that there was positive GDP growth from 1986, but still at less that I percent per capita annually. Because of the manner in which comparative GDP is recorded in the World Development Report, Tanzania's relative position in the GDP table has declined as the economy has progressed, so that it is now recorded as having the second lowest GDP of the nations included. This is entirely the effect of exchange rate adjustments, although it is reported in the Dar es Salaam press as evidence of the failure of the reform program! (See Van Arkadie and Komba 1992.) 39 40 Economic Strategy and Structural Adjustment in Tanzania many-but certainly not all-is improving, it is not yet evident that the material and institutional basis has been laid for sustained growth over the longer term. While there has been widespread revitalization of the private economy, partly as a consequence of deregulation and the liberalized trade and payments regimes, it is still unclear that the conditions exist to support a sustained high rate of growth, and the economy remains too dependent on aid. 5.4 One question mark about the impact of reform relates to export performance, at least as measured by official data. This is partly a result of the deplorable state of the world markets for leading traditional exports.34 Continuing vulnerability to fluctuations in markets for a few traditional crops, however, demonstrates continuing structural weakness, a problem that has yet to be solved. The assault on "inward-looking" trade policies is often supported by appeals to examples of successful, export-led growth in East Asia. The success of the Asian tigers was based on dynamic adjustment of their comparative advantage, however, whereas the export trade of Africa continues to concentrate on traditional commodity markets. 5.5 In the period preceding structural adjustment, government interventions to alter the location of Tanzania in the international division of labor were signally unsuccessful. Under structural adjustment, diversification into new export areas has made a small beginning, but performance of recorded nontraditional exports has done no more than restore volumes to 1980 levels.35 There is considerable unrecorded export trade-for example, the artisanal mining sector, which produces gold and precious stones, exports much of its output without it being recorded in official data. There is a diversified cross-border trade in consumer goods, including food, which is also only partially recorded. Efforts to estimate the magnitude of unrecorded exports-for example, based on estimates of the financing requirements for recorded imports-suggest that unrecorded exports could equal official recorded exports. If that is so, Tanzania's recovery may well be much more robust than impressions based on official data suggest. This is highly speculative, however. Even if it can be inferred that unrecorded exports are considerable, no one has yet been able to estimate a growth rate. 5.6 In the absence of hard data, it is reasonable to question whether a plausible vision exists of how Tanzania can earn its way in international markets. Even if market stimulus is a necessary ingredient for economic success, it is far from obvious that it will be sufficient to generate the export growth to fund Tanzania's increasing import and debt servicing requirements. The response to new economic opportunities by actors in the economy has been significant, and given time this may have a cumulative impact on economic performance and provide the 34 Although as this paper is being prepared, Tanzania seems to be facing the prospect of improved markets, because the price of coffee escalated in the early months of 1994. 35 Official export data show stagnation throughout the 1980s and up to 1992, with year-to-year fluctuations but no clear trend. As the discussion of parallel markets and the foreign exchange system noted, however, there has been a significant undercounting of exports. Whether the undercounted total has grown significantly is unknown. With the unification of exchange rates and a fully liberalized trade and payments regime now operative, a higher proportion of exports should flow through channels captured in the official statistics. Sustainability of the Reform Process 41 institutional basis for export diversification,36 but there is little hard evidence of the magnitude of the growth impetus that this will sustain over the medium term. Moreover, part of the improvement that can be observed has been at least partially the result of a self-fulfilling donor hypothesis: policy reform is necessary, therefore increased donor support is conditional on reform, but increased support following reform itself results in improved performance. Thus, ou:pu: revival may have been as much a result of aid increases as the positive impact of policy reform. 5.7 Since the adoption of the Economic Recovery Program, Tanzania has achieved positive growth in per capita income, probably at rates higher than suggested by official GDP data, which undercount informal and private economic activities. Nevertheless, both the Tanzanian population and the donor community tend to perceive growth as sluggish. It may be difficult to sustain the vision of market-orientated development in the absence of a clearer breakthrough into higher growth, particularly in the absence of the powerful inspiration of other African success stories.37 For the economist, a Panglossian view that the free market provides the best of all possible worlds may be sufficient. For others, that may be less persuasive, particularly if the memory of the economic difficulties faced under the old regime in the early 1980s fades, and the ideals espoused in the Arusha Declaration revive. In the longer term, it is likely to be necessary for government to again address issues of economic structural change, learning lessons from the failures of the earlier efforts of 1960s and 1970s, but seeking out a vision of change that is less passive than simple dependence on market forces. The Managerial Capacity of the State 5.8 A continuing cause and result of economic crisis that structural adjustment policies have been unable to address is the erosion of the capacity of the state. The development of a market economy needs to be backed by reasonably strong state institutions to provide a stable macroeconomic setting and a supportive institutional environment and to offer the economic and social services that are necessarily the responsibility of the state. The recent concern about governance-for example, in the World Bank literature-includes an effort to spell out a vision of the state and its role in the reformed market economy. But the reform process as it has been experienced so far has done little to restore the capacity of the state. 5.9 One of the main thrusts of structural adjustment has been to prune the economic tasks undertaken by state. The removal of exchange controls, price controls, and the reduction in all sorts of state intervention have dramatically reduced state tasks. These reforms attempted to 36 See the combined issue of Tanzanian Economic Trends (vol. 5, nos. 3 and 4, Sept. 1992 and Jan. 1993), which is devoted to nontraditional exports. See, in particular, the article by Dik Bol. 37 In 1991 I took on the intresting asignment of accompanying a tearn of senior Victnnmese on a study tour of Malaysia and the Republic of Korea. The powerful impact of the economic success of economics in the same region was obvious. It is less persuasive to hold up the image of Asian success as an inspiration to an African audience, but this is what proponents of market reform find themselves doing. 42 Economic Strategy and Structural Adjustment in Tanzania confront the reality of an enfeebled state apparatus that was increasingly unable to enforce its rules and regulations. Economic reform has narrowed the range of government intervention and should have left government with a more manageable set of tasks. Yet as private economic activity has revived, it has become increasingly clear that the state no longer has the capacity to undertake the basic tasks necessary to support the private economy. The expanding private economy faces constraints because of the inadequate provision of services that remain the responsibility of the public sector. Where economic infrastructure has been rehabilitated, it has been the result of donor support. 5.10 Adjustment policies could have contributed to the enfeeblement of state institutions. Exchange rate and budgetary adjustments were intended to shift income to the tradeable goods sectors and away from the state bureaucracy, both in nominal income and access to rents. The shift in the intemal terms of trade in favor of the private sector and the rural economy involved some loss of privilege for state functionaries, one reason for resistance to change. The availability of external support cushioned the impact, and increased imports allowed a general revival in living conditions, but the public service was left underpaid and poorly motivated. In recent years there has been an effort to restore the real incomes of senior civil servants, largely through enhancement of perquisites rather than by decompressing salary scales, but this has not affected the large mass of public servants. 5.11 The deterioration in the delivery of social services has persisted as a reflection of the stringent resource constraints and slow implementation of new strategies of delivery and cost recovery. Macroeconomic structural adjustment did nothing to improve the situation in the social sectors. While there is little evidence that structural adjustment had a regressive impact on the primary distribution of income, fiscal stringency has had a strong negative impact on the public provision of social services. Formulation of new sectoral strategies (for example, in the ESAP) has been partial, support for the social sectors has not been informed by a coherent strategy to promote institutional reform, and progress in implementing new strategies has been limited. In light of the fiscal debility of the central government, one way forward might be through local initiatives and control, either through local government or self-help, even though this feeds regional inequalities and might give local privileged groups greater control over access. Given the importance of investment in human capital, the continuing decay in education and health delivery systems may prove to be the most debilitating of the weaknesses currently observable over the long term. 5.12 The tasks left to government are not only critical, but also in some ways complex, including the manipulation of market-related policy instruments, the implementation of public investment programs, and the management of public infrastructure and the delivery of social services. The government needs to fill gaps in the institutional arrangements for the private economy, including modifying the legal framework in such areas as land law and corporate law; sponsoring the development of new markets, such as financial markets; and encouraging coordinating agencies, such as trade councils and chambers of commerce. There is little sign of restoration of governments' ability to handle such essential tasks effectively. Sustainability of the Reform Process 43 5.13 Real salaries in the public service collapsed during the crisis and adjustment, and the performance of many employees precisely matches their vestigial incomes ("the government pretends to pay us, and we pretend to work"). Senior civil servants survive on the perquisites of office.38 The public investment that does take place is donor-financed, so that public investment planning is almost entirely a matter of donor coordination. It could be argued that what is required is nothing less than the reconstruction of the state. Even if an essentially private economy is the vision for the post-structural adjustment era, a more effective set of state institutions is required. The state needs to be reformed to be able to meet the needs of market-orientated development, including the provision of essential infrastructure and social services and, as the private sector grows, the implementation of appropriate public actions to accommodate and orchestrate private economic activity. 5.14 What is required is a more careful definition of the responsibilities of government in the emerging economic situation and a systematic effort to put in place a reformed bureaucratic structure to tackle the identified tasks. The agenda for the reconstruction of state institutions includes clarification of the required institutional structure, identification of appropriate manpower requirements, and the definition of an effective incentive system for the public service. In principle it is not particularly difficult to formulate a blueprint for reformed systems of government administration; in practice, however, very little gets done as the existing systems, for all their weaknesses, operate as a subsistence sector for large numbers of employees. 5.15 Civil service reform might result in fewer public servants, although concentrating attention on reductions in numbers has so far not proved a very effective reform strategy. This is partly because redundancy is expensive, so that there is little initial benefit from staff reduction, and because appeals to reduce the "bloated" civil service tend to oversimplify a complex set of problems. It is certainly the case that for some cadres employment has expanded well beyond need, and a reduction in numbers combined with an improvement in terms should be one strategy to improve performance. This is probably the case in a number of major economic management tasks of government that do not require large numbers to implement, but do require a certain minimal level of professional competence to perform effectively. 5.16 Much of the expansion in the Tanzanian public service, however, occurred through the increase in the number of schoolteachers (the largest group of public servants) and primary health care workers, as a consequence of the decision to expand the coverage of the education and health systems in the 1970s. In these cases, the reduction in the numbers employed would imply a decision to reduce the coverage provided; conversely, mass provision of such services will only be possible if the primary contact staff (primary schoolteachers and primary health care workers) are paid at levels that can be sustained by the population serviced, either through taxes or cost 38 Although these may be sufficient to provide a reasonable income for top functionaries, a result of the process of using ad hoc mechanisms to shore up incentives is a lack of transparency regarding the system of remuneration operating in the civil service and the parastatals. 44 Economic Strategy and Stuctural Adjustment in Tanzania recovery. This suggests that important groups of public servants cannot be paid at levels out of line with the household incomes of the communities they serve. 5.17 In practice, public service incentives in Tanzania have been handled increasingly through the manipulation of nonwage perquisites, so that the differential in actual real incomes is much broader than the high degree of compression in formal salary scales. Perhaps more important than reducing total numbers of public employees-which tends to be achieved by shedding the lower-level employees who are ready to continue at the pitiful existing salaries, and are no great burden on the budget-is the need to create a more complex career and incentive structure that recognizes the needs of different groups of public servants and ensures that an essential core of senior public servants are looked after sufficiently well to attract the high-level staff and to motivate them. At all levels of the public service, however, commitment and morale are not just a matter of payment levels, but also a matter of the status and responsibility accorded to the service. One need in the Tanzanian system is to restore the level of commitment at many different levels of the public service. 5.18 One aspect of deterioration of governance in the reform period, which is by its nature difficult to document but widely perceived, has been the growth in corruption. Corrupt practices were unusual in the 1960s. Under the conditions of economic crisis in the 1970s, widespread petty corruption was associated with the development of the informal economy. During the period of econornic reform, corruption is perceived to have spread to the higher levels of government.39 If economic policy emphasizes the market and accepts free-wheeling private accumulation, it is unclear how the political system can be insulated from the temptation to use political power as a base for private accumulation. 39 The accusations of conuption from the donor community have focsed on the administration of foreipn exchange program support (import support and the OGL scheme), and came to a head in 1994 with public accusations regarding the maladministration of revenue collection in the Treasury. This led to suspension of aid disbursements and changes in the leadership of the ministry. In 1994 and 199S, former President Nyerere launched devastating attacks on the probity of the government, including explicit accusations that the corruption spread to President Mwyni's own family. 6 The Role of Donors 6.1 One change in the political economy of Tanzania in the 1970s was the rise in the importance of donors. It was during the first period of economic difficulty that donors-including the IMF and the World Bank-built up their support; total aid grew from US$100 million in 1973 to US$302 million in 1975 (see Havenik and others 1988, table 6.6). In the decade following 1973, Tanzania received around US$5 billion in aid (Havenik and others 1988, table 6.6). This transformed the structure of the Tanzanian balance of payments, accommodating a large and persistent negative trade balance, in contrast to the 1960s when the trade balance was modestly positive. It was almost as though the adoption of self-reliance as a national goal was particularly attractive to donors, perversely resulting in a high degree of aid dependence! It also meant that with aid donors funding nearly half of the import bill, donors necessarily became important actors in the policymaking process.40 6.2 Any discussion of the evolution of policy in Tanzania must thus address the role of donors. They have become important political actors. Major macroeconomic policies, such as exchange rate adjustment and fiscal and monetary policies, have fallen increasingly under the influence if not direction of the Bretton Woods institutions, and even at the level of project choice, design, and staffing, donor agencies have often played the leadership role. 6.3 The assertion of donor authority has occurred despite the rather weak evidence of donor competence. The record of disaster in the performance of aid projects in the 1970s almost matched the earlier debacle of the British Groundnut Scheme.4" Frankness at the level of operational evaluation of projects has not, however, led to any particular modesty on the donor 40 Whether all that aid was of much use is a legitimate question, given that after a decade of heavy aid flows, per capita incomes were probably lower. In a contribution to the most recent World Bank Country Economic Memorandum, I offered the judgement that at the margin, aid had negative productivity-certainly OED assessments have demonstrated that nmny World Bank projects in that period produced nothing, and they all had domestic costs (see Therkildsen 1988). 41 The record is documented in many frank operational evaluations, such as the World Bank's evaluation of such agricultural failures as the National Maize Programme, Geita Cotton, the cashew nut processing projects, and the like. Therkildsen (1988) has documented the failure of a number of agencies in supplying rural water supplies. The Morogoro Shoe Factory must be one of the worst disasters in the World Bank industrial portfolio. 45 46 Economic Strategy and Structural Adjustment in Tanzania side in claiming wisdom in policy formulation. Nor has donor failure been associated with any noticeable accountability. 6.4 Donors were slow to recognize weaknesses in Tanzanian economic performance. By the second half of the 1970s, critical performance indicators began to suggest that the system was in trouble-notably the decline in production of export commodities and slow completion of investment projects as limited resources were spread over far too many projects. Nevertheless, the system was not yet in open crisis and donor alarm was slow to manifest itself The World Bank 1977 Basic Economic Report included a number of warnings. For example, it noted the operation of the parastatal sector and the possible highly negative macroeconomic scenarios, and it could be seen as a turning point in donor attitudes, but such warnings were still muted, and did not yet motivate harsh conditionalities.42 The Tanzanian government showed some flexibility in responding to advice from the World Bank to relax import controls. Unfortunately, this had a disastrous result. Imports surged in 1988, just as the coffee price boom came to an end, and imports again had be brought under tight control. 6.5 Donors not only participated in the dissipation of resources at the project level, but also contributed to systemic deterioration. Donor willingness to fund the development budget with little consideration for the sustainability of the recurrent spending requirements also added to the chronic deterioration in public expenditure management. Donors also contributed to the overelaboration of the bureaucracy. New departments of government and parastatals were funded with little concern for their eventual sustainability, either in manpower or funding. The flow of technical assistance itself contributed to a cancerous proliferation of bureaucracy as development projects spawned new sorts of bureaus and government agencies. For economic planners, nothing succeeded so well as failure: the failure of macro planning led to the creation of sectoral and project planning units, failure at the center motivated the development of regional and district planning, and failures in aid implementation led to the creation of project implementation and monitoring units. 6.6 The global shift in donor thinking was as important as other exogenous factors in influencing the reform process. The World Bank signaled a much clearer commitment to the 42 It is difficult to pinpoint when donors moved over to a generally critical stance on Tanzanian policy. In the mid-1970s, when in retrospect it seems that problems were already building for Tanzania, the World Bank was financing projects in a manner that bordered on the reckless (the Second Cashew Nut, National Maize, Geita Cotton, and Morogoro Shoe Factory were among the failures that originated in that period). While the World Bank Basic Economic Memorandum of 1977 began to raise general issues of policy and strategy, the subsequent comment by the Bank itself on that report illustrates the limited scope of the criticism: Though the BER threw light on linkages between 'real' aspects of the economy which with the Bank had been dealing, and the macro- and monetary aspects, the latter were still treated in a cursory fashion. Monetary and foreign exchange policy were virtually ignored. Replacement of quotas by tariffs was advocated to rationalize production, but it was not even mentioned that the proposed changes would have a positive effect on fiscal revenue' (World Bank 1990, p. 58). The Role of Donors 47 market with the publication of the Berg Report in 1980 (World Bank 1980).43 In influential Organization for Economic Cooperation and Development (OECD) capitals there was a more substantial ideological shift than at any time since World War II (notably in Washington, with the advent of Reagan, and in London, under Mrs. Thatcher). The IF was also less accommodating by the end of the 1970s than it had been when responding to the first oil shock under Witteven- although in the Tanzanian case, its unbending stance was in part the response to rnishandling of negotiations, for which the Tanzanian government must bear some responsibility. By 1980, the donor community had shifted its stance significantly compared with the mid-1970s, and the Tanzanian leadership was possibly rather slow in picking up the change. 6.7 The donor community did not, however, present a united front. There was a virtually universal recognition that major reform was required, but a number of bilaterals were unwilling to shift from a supportive stance, seeking reforms through friendly dialogue rather than a more belligerent stance of hard conditionality. This was particularly true of the Nordic group and the Netherlands. Even inside the World Bank, there was an effort to hold off outright confrontation through the agency of the Tanzanian Advisory Group (TAG;1981-82). In 1981, in an effort to encourage reforms and to avoid confrontation between the World Bank and the Tanzanian government, the Bank proposed and agreed to fund TAG. This was an independent team, funded by the Bank but acceptable to the Tanzanian government, that sought to mediate and find a reform package acceptable to the Tanzanian authorities that could form the basis for a new IMF/World Bank program. The three-member advisory group, agreed to jointly by the Bank and the government of Tanzania, was backed up by a small secretariat.44 6.8 TAG presented a report to the government of Tanzania that was partially incorporated into the Structural Adjustment Program. Whether the proposals of TAG-which included a significant devaluation, but still well short of the adjustments proposed by the IMF-would have provided the basis for an IMF program is in itself doubtful; the watered-down version accepted by government was not acceptable. The attempt to reach a compromnise with the donor community therefore proved unsuccessful, and the Structural Adjustment Program (1982) did not provide an acceptable basis for an IMF/World Bank program. 6.9 Bilateral donors who maintained their programs during the period in which Tanzania did not reach agreement with the Bretton Woods institutions (1979-85) had to recognize that the 43 This report is commonly know as the "Berg Report.' The team that produced the report was led by Elliot Berg, who could claim a reasonable degree of consistency, because he had already been a critic of etatist development strategies in Africa in the mid-1960s. 44 The Advisory Group was led Ambassador Ernst Michanek, who had been the first head of SID. The other members were Professor Gerry Helleiner of the University of Toronto, who had been the first director of the Economic Research Bureau, University of Dar Es Salaam, and Professor Cran Pratt, also of the University of Toronto, who had been first vice chancellor of the University of Dar es Salaam. The secretariat was led by the current author, and included Professor John Loxley, University of Manitoba; Rolf Hofmeier, University of Hamburg; and Peter Ngumbullu, now principal secretary to the Tanzanian Treasury. 48 Economic Strategy and Structural Adjustment in Tanzania main thrust of their support should no longer be directed to increasing the capital stock, but should instead provide the minimal resources required to sustain recurrent activity and to rehabilitate critical elements in the decaying infrastructure. The "like-minded" therefore found themselves in a holding operation, using their support to keep the economy going, while actively lobbying the Tanzanian authorities to reach agreement with the IMF and World Bank. Critics of the aid agencies within the like-minded countries have questioned whether the "soft" approach adopted did Tanzania a disservice, noting that reform might have happened sooner if the donors had presented a united front. Against this it can be argued that the Nordic group was able to act persuasively in encouraging the Tanzanian government to reform, and that a total donor boycott would have had severe economic effects and quite unpredictable political consequences. 6.10 In the 1980s the euphemistically titled "aid dialogue" was finally successful in promoting policy reform. This proved effective in promoting devaluation and the shift to a more liberal payments and trade regime and reductions in budgetary subsidies to parastatals. Donor conditionality has been less successful in promoting other institutional reforms. In relation to reforms in agricultural marketing, where the objective should have been the straightforward end of single-channel marketing monopolies, the World Bank was surprisingly tentative in pushing conditions. The liberalization of food marketing, which was a corollary to the withdrawal of subsidies to the National Milling Corporation, was carried through with remarkable success, partly because it involved the accommodation of an informal pattern of food trade that was already well established. The promotion of competition in the processing and trade of traditional export crops was much slower, and remains incomplete. Given that an early Bank structural loan was conditional on agricultural sector reform, it suggests that an opportunity was lost; compared with Fund conditionality, Bank program loans tend to carry more conditions of much less precision and strategic significance. 6.11 The Bank and Fund did not do very well in promoting financial sector reform following the completion of the Presidential Commission of Enquiry into Banking (the Nyaribu Report) in 1990. Swift progress should have been possible through decisive support for the implementation of the commission proposals. While no doubt imperfect, they were radical and comprehensive. 6.12 So far there has been little success in donor promotion of comprehensive public sector reform. The initial steps in structural adjustment involved changes in key macroeconomic policies and decisions to dismantle numerous government controls; as such, the programs were implemented by a small cadre of officials in the economic coordinating ministries and the Central Bank. Donor conditionality initially focused on a few macroeconomic policy instruments that could be readily monitored. Reform of the institutions of government is a more complex task, and donor promotion of such reform faces the contradiction that their own role has been one source of the erosion of governmental capacity. Even where piecemeal efforts to improve the performance of portions of the government have had some success, it has been at the expense of making the overall system more confused (for example, the provision of ad hoc donor incentives The Role of Donors 49 to public servants in priority aid areas, which render the overall incentive system even less coherent). 6.13 Administrative capacity was not enhanced by donors continuing with technical assistance approaches developed in the 1960s to handle a quite different set of problems, inserting highly paid foreign advisers or consultants into systems that suffered not from a lack of skilled nationals, but rather from an inability to utilize them effectively. This bred resentment and frustration on both sides. 6.14 Reform of government is most likely to be achieved by a process owned by and embedded in the national government structure, with the understanding and eventual support of the top political leadership and civil servants. Such United Nations programs as the Management Development Program (MDP) and National Capacity Assessment (Natcap) have addressed aspects of administrative reform, as did two World Bank public expenditure reviews and numerous bilateral technical assistance exercises. An effective donor strategy for promoting the required degree of reform in public administration has yet to be identified. Donors have promoted ceilings on public expenditures and employment, programs to ease the hardships of redeployed civil servants to soften opposition to reform, and piecemeal and uncoordinated interventions to bolster particular segments of the public service, but these initiatives have so far fallen short of an effective reform strategy. 6.15 Could donors do more to promote administrative reform? One problem may be that while the donor promotion of a market economy has been based on a plausible vision of the way the economy could operate more effectively, the accompanying vision of the way politics and the administration should perform is less clear, or where it can be identified, it may be too Utopian to be of much practical significance. The vision of a small, well-paid, elitist public service led by an uncorrupt and committed political leadership has a striking similarity to the ideal of colonial administration. Added to that, however, is the desire that the system function as a multiparty democracy. While aspects of the internal logic of that vision can be defended as consistent- multiparty democracy might be expected to check the more flagrant forms of high-level corruption-there are also inconsistencies. Democratic politics in poor countries is unlikely to be supportive of an elitist administration (one component of the "good governance" package) and is likely to encourage patronage and an overexpanded civil service. Likewise, in a poor country, emphasis on the virtue of private accumulation can too easily lead those in positions of political authority to use their political power to participate in the accumulation process. 6.16 One way for donors to pursue their vision of better governance would be to seek to buy a more effective administration. Setting aside a significant proportion of funds currently devoted to technical assistance, consulting, and the like over a period of years to fund a new incentive structure for national professionals that could be introduced with reforms to improve working conditions could raise standards of performance. This would contrast with donor- sponsored attempts at reform so far, which have emphasized cuts in the public service without offering noticeable benefits to the remaining staff Such a donor initiative would require modification of long-standing donor thinking limiting the role of aid in supporting local costs. 50 Economic Strategy and Structural Adjustment in Tanzania Given the de facto contributions donors have been making to recurrent budgets through import support programs, this would primarily involve an effort to think through the logic of existing practice. Perhaps the most alarming implication of such proposals is that donors would be taken further down the road toward becoming the meaningful government. 6.17 The possible donor contribution, however, is likely to be limited. The effective organization of government is necessarily a national matter, and there is little evidence that donors have a comparative advantage in providing government services. Indeed, it is evident that pervasive donor intervention has become part of the problem, because it has undermined local initiative and accountability. The commitment of substantial resources by agencies such as the World Bank to the production of numerous highly elaborate mission reports seems to have only a sporadic impact on local policymaking, but it can serve to displace any national commnitment to undertake serious policy analysis. 6.18 It is also difficult for donors to establish and maintain a realistic and internally consistent set of expectations. Donors have sponsored structural adjustment and have lectured on the limited capacity for successful government intervention in the economy. Yet it is the donors who press to add items to the policy agenda that require sophisticated interventions, such as targeting social groups that have suffered from structural adjustment policies. Having pressed for market economics, donors tend to be shocked by the realities of market relations in very poor economies. And having extolled multiparty politics, there is surprise that the emergent lines of political division are based on populist demagoguery, rather than a measured discourse regarding policy. Therefore, in the face of the evidence that economic reform in Tanzania has been more of a success than most commentators and donors would have expected when the process started in the early 1980s, donors turn sour, infected by fatigue that reflects the failure to define realistic expectations and goals, and the short institutional memories chara,cteristic of donor agencies. 7 A National Vision of Development? 7.1 The reform period in Tanzania has been associated with a blurring of political vision; perhaps it is best described as a period in which politics have become less important and political attitudes more cynical. Even the initiation of multiparty politics in Tanzania did not respond to any upsurge of political enthusiasm or unrest; the testing of opinion that preceded its introduction suggested that there was not majority support for abandoning the familiar one-party system. 7.2 The tentative and sometimes ambiguous ways in which a number of reforms have been introduced could be seen as resulting from weakness of leadership and failure of vision. It could also be argued that the movement away from the clearly defined vision of the Arusha Declaration and the accommodation of a process of change from below was well served by a politics that was low-key and rather nonideological. In the politics of the market economy, profits rather than prophets become the order of the day. That Tanzania has been able to experience profound economic and social changes with remarkably little tension and confrontation may in part be because of a politics, which increasingly avoided issues of principle and ideological choice. 7.3 Economic reform came to Tanzania during a period in which a charismatic leader with a clearly defined vision of development retired, and his vision has not been replaced by a clearly defined alternative. Nyerere had eventually been willing to reform, and had often been willing to be self-critical about failings in the implementation of the Arusha vision, but he has never acknowledged that the economic crisis the system faced demonstrated inherent inadequacy of his socialist vision of the Tanzanian future. Since Nyerere, the tone of policy presentation has been pragmatic in character, and the ideological shift has been implicit, rather than being spelled out in a new political ideology. 7.4 Effective economic performance by the state is not only a matter of the quality of leadership and clarity of vision-the record shows that Tanzania got into economic difficulties with high-quality leadership endowed with a clear vision. Perhaps more important is the sort of pressure likely to come to bear on the state to influence its performance. Tanzanian politics is in a state of flux, and it is quite unclear how the economic groups strengthened under liberal economic policies will participate in the political process, how a lively ("vibrant" to use a favored World Bank term) small-scale sector and fledgling indigenous capitalists, operating through the new multiparty populist politics, will impinge on enfeebled state institutions. 51 52 Economic Strategy and Structural Adjustment in Tanzania 7.5 The rhetoric of multipartyism (including that of the ruling party) leans toward supporting the virtues of good government and rooting out corruption. At the same time, there seems to be an increasing public perception of venality in the upper reaches of government, although how correct this perception is may be impossible to judge. Part of the current disquiet no doubt reflects the impact of an increasingly open, muckraking press.4' This suggests that there is a growing constituency for improved governance, which could result in the definition and implementation of a new vision of the role and management of government, possibly in the context of the elections of 1995, the first ever multiparty presidential and parliamentary elections in Tanzania. 7.6 If the reform process has been guided by a vision, it has been that of the international donor community. This is not to say that there has not been national support for reforms, but that acceptance has so far fallen short of the definition of a new set of clear national goals to replace those articulated in the Arusha Declaration. The vision of the donor community can be pieced together from numerous donor documents, particularly from the World Bank, which address elements of economic policy. It is probably not too far in the direction of parody to suggest that the recipe on offer is "take one teaspoonful each of free trade and private investment, plus a dram of property rights and the rule of law; and wash down with a swig of multiparty democracy" (Mortimer 1994)." The donor vision is reflected not only in the reforms promoted by the World Bank and other agencies, but perhaps even more sharply in donor exasperation at the recalcitrance of the African pupils in not learning their lessons. 7.7 The vision of economies operating with undistorted commodity markets and freely flowing capital, with just social policies that deliver social services in an equitable manner, officiated by a multiparty political system, innocent of corruption and pure of purpose is an attractive Utopia, but is not particularly realistic. The lack of realism shows itself in the oscillations of donor concerns-the emphasis on the market followed by shock at the inequalities the market generates, the adulation of private business and the embarrassment at the processes 45 Such matters are difficult to assess. Certainly there were a large number of scandals and suggestions of scandal aired in the press during 1993, and the level of popular cynicism and anger in Dar es Salaam seems to have increased considerably. There is also a perception in the diplomatic community that the system has become more corrupt, which specifically focused on the operation of the allocation of foreign exchange under donor program support (the OGL) during 1992-93. It should be noted that the level of probity in the public service was at a high level in the 1960s and early 1970s. Behavior certainly deteriorated at the lower levels of the public service in the economic difficulties that emerged during the 1970s, when falling real incomes created the need, and a supply-constrained, malfunctioning system of control created the opportunities. There is now a widespread perception that dubious practices have penetrated the political hierarchy-a view articulated, for example, by Mr. Mrema, Deputy Prime Minister and Minister of Home Affairs, a populist spokesman in support of law, order, and honesty in public life. 46 Edward Mortimer explores the possibility of viable alternatives to this Western model in terms of a possible Asian counterpart (see Mortimer 1994). A National Vision of Development? 53 whereby business classes emerge at the early stages of capitalism, and endorsement of multiparty politics and subsequent surprise that party politics is about the sectional competition for spoils. 7.8 The vision of welfare distribution associated with the adjustment process is far from clear. Advocates of the market tend to view the inequalities that emerge from an efficient market process as functionally justified, and they tend to believe that in freely operating labor markets and undistorted commodity markets, poor workers and peasants are likely to benefit from the opportunity to sell their labor services and products. It is, however, difficult to find a specific view in the structural adjustment literature about the likely structure of ownership that will emerge under a liberal policy regime. There is an explicit belief in the virtues of private business, but little attempt to spell out the inequalities of wealth and power that are likely to result. The official vision of the CCM government seems to be to accept the need for a shift to private sector development, with the accompanying pattern of income distribution, while translating the continuing verbal commitment to socialism into a concern for the provision of services to the masses (Kolimbo 1993). 7.9 The intemal dynamics of policy reform can be seen as a process of state accommodation of new groups, first implicitly, by turning a blind eye to practices that were widespread, although in principle illegal, and then explicitly by adjusting the rules. From the changes begun under conditions of crisis and consolidated under structural adjustment, an African property-owning class is emerging, alongside the resurgence of minority (Asian and Arab) interests. With structural adjustment, there has been a flowering of private economic activity. Successful entrepreneurs have emerged from the minority trading communities; they had hung on through periods of state control, and became adept at pursuing rents during the period of crisis. African entrepreneurs have emerged as small businessmen and from the state system. 7.10 A politically viable vision of development through the market must address the issue of national participation in the ownership of private, large-scale economic activities. Within the Tanzanian polity, one view about the future under reform has emerged from groups critical of CCM policies that restricted the scope of African businessmen in the past, and see the possibilities in the new dispensation for the emergence and consolidation of an Afiican capitalist class. Criticism of aspects of economic liberalism are voiced by those who believe that the state should support African capitalists in competition with ethnic minorities and foreign business.4' The interests of segments of the African business community have been articulated through appeals for "indigenization" by such groups as the Chamber of Commerce and Agriculture and Industry, and one line of division in the new multiparty politics could emerge around that issue. There is no sharp line of division between those in politics and government and the emerging African capitalist class. Part of the process of accumulation in the development of that class involves primitive accumulation by those with access to state resources. Quite open private accumulation by leaders in the party, government, and parastatal sectors, disallowed in the self-denying 47 The case for this is set out quite explicitly by Edwin Mtei, leader of Chadema and one-time executive director of the IMF, in a recent interview in the EEC Courier, November 1994. 54 Economic Strategy and Structural Adjustment in Tanzania "leadership code" adopted after the Arusha Declaration, has been increasingly countenanced during the 1 980s. 7.11 The emergence of an African property-owning class has not yet been associated with an accentuation of ethnic tensions within African society. Given the lack of concentrated tribal makeup in Tanzania, it is less likely to be prone to the sort of ethnic divisions that have emerged in Kenya, although there have been some indications that religion could become a political factor.4' The new African business class, however, is emerging from among the small traders, craftsmen, and contractors, activities in which the Moslems are well represented, as well as from the bureaucratic and technocratic elite, which is predominantly Christian, and it might be that the balance between the religions will be still more equal within a growing private sector than it was within government and the parastatals. 7.12 While there is no shortage of African entrepreneurial talent in small-scale activities, one important contrast between Africa and East Asia is the lack of large-scale national capital in the former and its active presence in the latter. The prospects seem good for accelerated African small-scale development; one source of optimism is the widespread evidence of small-scale entrepreneurial initiatives in agriculture, trade, transport, construction. and craft activities. It is less clear whether a national capitalist class is emerging that is capable of working with the state to develop a program for a more advanced stage of capitalist development. 7.13 Successful visions of social change are not often handed down on tablets to great men in the mountains, but are more likely to be founded in a concrete reality of what is likely to receive support and what is likely to work. A vision of a successful economic future should be based in empirical reality, as well as incorporating acceptable ideals. There is an underlying dilemma in Tanzania, as in much of Africa, that in the absence of success, it is far from clear what that vision should be. 48 Religion is mentioned herm because one pouible division in Tanzanian society is along religious lines The debate over the Union has had religious undertones, nd there was a brief outbreak of religious conflict in Dar es Salaam in 1993. It is difficult to predict the likelihood of future conflict. In Dar es Salaam and on the coast, there is some coincidence of religion, tribe, and cooomnic datus, with the Muslims in an underprivileged position, which could provide the basis for political dicontent on rligious lines. But up country the situaion is more confused (for example, in Kagera and Par)-membera of the san com nunity and even the same family may be drawn from both religions, blurring the divisions. Also, it is posible that liberlization has eased some of the economic inequalities, because Muslims ae probably bettr represented in private commercial activities tha in the bureaucracy. Bibliography Bol, Dik. 1993. "Non-traditional Exports in Tanzania: Trends and Prospects." Tanzanian Economic Trends 5 (3 &4): 4-40. Coulson, Andrew. 1992. Tanzania: A Political Economy. Oxford, U.K.: Clarendon. Ellis, Frank. 1982. "Agricultural Pricing Policy in Tanzania." World Development 10 (4): 263-83. 1983. " Agricultural Marketing and Peasant-State Transfers in Tanzania." Journal of Peasant Studies 10(4):214-42. Finucane, James R. 1974. Rural Development and Bureaucracy in Tanzania: The Case of Mwanza Region. Uppsala: Scandinavian Institute of African Studies. Green, R. H., D. G. Rwegasira, and B. Van Arkadie. 1980. Economic Shocks and National Policymaking: Tanzania in the 1970s. The Hague: Institute of Social Studies. Havenik, Kjell J., Finn Kjaerby, Ruth Meena, Rune Skwrstein, and Ulla Vuorela. 1988. Tanzania: Country Study. Norwegian Aid Review. University of Bergen. Hyden, Goran. 1980. Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry. Berkeley: University of California Press. 1983. No Shortcuts to Progress: African Development Management in Perspective. Berkeley: University of California Press. Kolimbo, Horace. 1993. EEC Courier, November. Maliyamkono, T. L., and M. S. D. Bagachwa. 1990. The Second Economy in Tanzania. London: J. Currey. Mans, Darius. 1994. 'Tanzania: Resolute Action." I Ishrat Husain and Rashid Faruqee, eds., Adjustment in Africa: Lessonsfrom Country Case Studies. Washington, D.C.: World Bank. Mortimer, Edward. 1994. "Has the West Triumphed at Last?" flnancIal Times, January 1. Mtei, Edwin. 1994. EEC Courier, November. Ratso, J. 1991. Structural Adjustment in Tanzania. Norwegian Ministry of Foreign Affairs. Photocopy. Ronnas, Per, and Orjan Sjoberg. 1991. Doi Moi: Reforms and Development Policies in Vietnam. Stockholm: Swedish International Development Authority. Sharpley, Jennifer. 1985. "External versus Internal Factors in Tanzania's Macroeconomic Crisis, 1973- 1983." East African Economic Review n.s.l (1). Shivji, Issa G. 1976. Class Struggles in Tanzania. New York: Monthly Review Press. Skarstein, Rune, and S. M. Wangwe. 1986. Industrial Development in Tanzania: Some Critical Issues. Uppsala and Dar es Salaam: Scandinavian Institute of African Studies and Tanzania Publishing House. 55 56 Economic Strategy and Structural Adjustmnent in Tanzania Stewart, Frances, and Jennifer Sharpley. 1985. Economic Policies and Agricultural Performance: The Case of Tanzania. OECD Development Centre. Tanzania, Ministry of Agriculture. 1983. The Agricultural Policy of Tanzania. Dar es Salaan. Tanzanian Economic Trends. 1988. "Trends in Capital Formation." Tanzanian Economic Trends 1(2). .1988. "Import Liberalization." Tanzanian Economic Trends 1(1). Therkildsen, Ole. 1988. Watering White Elephants? Lessons from Donor-Funded Planning and Implementation of Water Supplies in Tanzania. Uppsala: Scandinavian Institute of African Studies. Van Arkadie, Brian. 1965. "Central Banking in an East African Federation." In C. Leys and R. Robson, eds., Federation in East Africa. Nairobi. 1973. "State Cooperation and Economic Independence in Africa." In Dharan P. Ghai, ed., Economic Independence in Africa. East African Literature Bureau. 1983. "The IMF Prescriptions for Structural Adjustments in Tanzania." In Karel Jansen, ed., Mfonetarism, Economic Crisis, and the Third World. London: F. Cass. 1986. "Some Realities of Adjustment: An Introduction." In External Finance and Policy Adjustment in Africa. Development and Change 17(3). . 1989. " The Role of Institutions in Development." In Proceedings of the World Bank Conference on Development Economics. Washington, D.C. Van Arkadie, Brian, and John Komba. 1992. "Long-term Trends in GDP." Tanzanian Economic Trends. Van Arkadie, Brian, and D. Ghai. 1969. "The Economies of East Africa." In P. Robson and D. Lury, The Economies of Africa. London: Allen & Unwin. van Cranenberg, Oda. 1990. Thje Widening Gyre: The Tanzanian One-Party State and Policy Towards Rural Cooperative. Delft: Eburon. von Freyhold, M. 1979. Ujamaa Villages in Tanzania: Analysis of a Social Experiment. New York: Monthly Review Press. World Bank. Various years. World Development Report. Washington, D.C. 1980. Accelerated Development in Sub-Saharan Africa: An Agenda for Action. Washington, D.C. 1990. 'World Bank/Tanzanian Relations, 1961-87." World Bank Operations Evaluation Departrnent Report 8329. Washington, D.C. PRIVATE SECTOR DEVELOPMENT DEPARTMENT OCCASIONAL PAPERS No. I Rhee, Katterbach, Belot, Bowring, Jun and Lee, Inducing Foreign Industrial Catalysts into Sub-Saharan Africa No. 2 Mody and Wang, Explaining Industrial Growth in Coastal China: Economic Reform ... and What Else? No. 3 Biddle and Milor, Institutional Influences on Economic Policy in Turkey: A Three-Industry Comparison No. 4 Lanjouw and Mody, Stimulating Innovation and the International Diffusion of Environmentally Responsive Technology No. 5 Tan and Batra, Technology and Industry Wage Differentials: Evidencefrom Three Developing Countries No. 6 Navarro, Reversal of Fortune: The Ephemeral Success of Adjustment in Venezuela, 1989-93 No. 7 Morales, Bolivia and the Slowdown of the Reform Process No. 8 Ibrahim and Lofgren, Governance and Structural Adjustment: The Egyptian Case No. 9 Tan and Batra, Enterprise Training in Developing Countries: Overview of Incidence, Determinants, and Productivity Outcomes No. 10 Sundaram, Teik and Tan, Vision, Policy and Governance in Malaysia No. 11 Tzannatos, Labor Policies and Regulatory Regimes No. 12 Kessides and Willig, Competition and Regulation in the Railroad Industry No. 13 Nagaoka, Antidumping Policy and Competition No. 14 Atiyas, Bankruptcy Policies No. 15 Brandao and Feder, The Case of Land Markets No. 16 Jebuni, Governance and Structural Adjustment in Ghana No. 17 Atiyas, Uneven Governance and Fiscal Failure: The Adjustment Experience in Turkey I - - -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~