CREATING AN AFFORDABLE PUBLIC SERVICE: TANZANIA, 1995 - 1998 SYNOPSIS In the early 1990s, Tanzania launched one of the most wide-ranging civil service reform programs ever undertaken in a low-income country. Over a period of 15 years, reform leaders worked to create a government the country could afford and that would deliver services more effectively. They reduced the size of the civil service, reorganized some functions into separate agencies, changed recruitment practices, adjusted pay scales, and launched initiatives to improve performance. Reform leaders scored some notable successes, reducing the size of the civil service by more than 25% and dramatically improving some core economic services such as business licensing. They also encountered obstacles and made slow progress in some aspects of their program, particularly performance management. This case focuses mainly on the period 1995-1998. Andrew Schalkwyk and Jennifer Widner wrote this case study on the basis of interviews conducted in Tanzania in 2008 and 2009. This case is part one of a series. Case published August 2012. INTRODUCTION In 2008, former Tanzanian President Benjamin Mkapa recalled the problems his country faced when he first came to office in 1995. The government "had almost ground to a halt" in promoting economic development. The public sector wage bill was 40% over budget.1 People complained they weren't receiving services. They worried about corruption. University of Dar es Salaam scholar Benson Bana put it more strongly: "Moonlighting was the rule of the day. Misuse-and abuse-of positions and public property was rampant." Between 1967 and 1983, Tanzania had vested many economic functions in a highly centralized, single-party government. The East African country had also limited trade ties with other countries, pursuing a distinctive form of agrarian socialism that had not yielded the gains anticipated by the first post-independence president, Julius Nyerere. In the 1970s, economic growth had stalled and turned negative in some years. Institutional capacity had decayed. The late 1980s began to witness a deliberate but gradual transition toward a more market-based system. Nyerere helped move some of these changes forward, both as president and then party chairman in 1985, when Ali Hassan Mwinyi succeeded him. In the early 1990s, as Tanzania shifted to a more market-based system, leaders began to re-imagine the role of the civil service. They aimed to transform government from the country's largest employer into a leaner organization that could provide services more efficiently and promote economic growth. Mwinyi's administration began the process, but progress was slow. In the early 1990s, Tanzania moved toward its first multi-party elections. Benjamin Mkapa campaigned for the presidency by underscoring the need for a more efficient and effective public sector. When he won the 1995 election, Mkapa signaled his intention to move forward. Recalled one adviser, "At his inauguration, some of the senior public servants got outfitted with three-piece suits," an elevation of the public sector that was "unprecedented" in Tanzania. The inclusion of civil servants on the guest list foreshadowed changes to come. Mkapa rapidly assembled a team of senior public servants to plan and manage the reform effort. Among them was Joseph Rugumyamheto, whom he appointed permanent secretary. Rugumyamheto, who had worked in the Ministry of Manpower Training and Administration, had graduated from Stanford University with a master's degree in education in 1985, and had risen through the ranks in the country's civil service. With Mkapa's backing, Rugumyamheto worked with a number of other administrative leaders to plan, develop and implement the first stages of a reform program. Tanzania chose to enact broad changes across many ministries, implemented simultaneously or in rapid succession. Not all stakeholders felt that this approach was wise, preferring pilot projects or a selective roll-out of new policies. However, the decision makers noted that one initiative usually required other changes, making a "big bang" approach nearly inevitable. The country's experience holds important lessons about sequencing, coordination and leadership. THE CHALLENGE The 1980s were difficult for Tanzania. The economy shrank every year from 1980 to 1985, a period during which the country lacked the foreign exchange reserves necessary to import equipment and other goods.2 The population topped 25 million by the end of the decade and would rise to 43 million over the next 20 years, making it the most populous country in east Africa. Gross domestic product per capita was less than US$170. By the mid-1980s, what had once been a fairly functional public service had become "almost a totally dysfunctional outfit," secretary Rugumyamheto remarked. It had grown from 148,223 employees in 1975 to 355,000 by 1992.3 The size of the civilian central government as a percentage of the overall population exceeded the average level for sub-Saharan Africa.4 Government had become a source of guaranteed jobs, and it had become too big to manage. The real value of civil service salaries had dropped since the 1960s. Productivity and efficiency declined as political influence increased and patronage systems took hold. Furthermore, the World Bank identified the resultant wage bill, which exceeded the budgetary allocation by 40% in 1994, as a key problem in closing the budgetary gaps that Tanzania perennially exhibited. In the late 1980s and early 1990s, several of the country's top political leaders realized that bold changes were necessary. To make politicians and public servants more responsive to citizen needs, the leaders started to open up the political arena, legalizing opposition parties and revising the constitution to allow for multi-party elections. They also began to reduce the size of government and encourage private sector growth. These big shifts in the character of politics and the economy required the country to alter its public administration as well. Rugumyamheto observed, "We undertook serious drastic changes in the economic arena as well as the political arena. And that created a big demand for drastic changes in the administrative sector." The country's governing circles widely acknowledged the need to re-think and reconfigure the public service. Kithingi Kiragu, a former Coopers-Lybrand director who lived in neighboring Kenya and consulted to Tanzania as a technical adviser, noted that the government "had done their own studies way back in 1975 ... and even in 1985 President Nyerere himself had tried to reduce the size of the public service. So they knew they had a public service which was bloated, which was not performing, which did not have the necessary skills ... whatever can go wrong with a public service was evident in Tanzania at that particular time." The country's supporters, particularly the Danish, Norwegian, and Swedish aid agencies that had long contributed to the country, underscored the need for change. The International Monetary Fund and the World Bank's International Development Association made assistance in grants and loans conditional in part on reducing the public sector wage bill. Mkapa reflected, "I was being driven not only by this urge to transform, but also to demonstrate to the donors and particularly to the international financial institutions that we were undertaking changes in better management of public finances, [and] also of the public service. Without a clear demonstration like that, we would not have qualified for debt relief, the Heavily Indebted Poor Countries Initiative." This debt reduction program, launched by the World Bank and the International Monetary Fund in 1996, adjusted the liabilities of poor countries so that the countries could better manage their debt burdens; debt relief was typically contingent on government policy changes. National pride may also have motivated reform. Despite the country's poverty, Tanzania was hailed throughout Africa as "the voice of reason; it was the voice of wisdom," Kiragu noted. He continued, "As Mkapa came into the presidency, [Tanzanians] wanted to regain part of their national pride." In spearheading reform, Mkapa and his team faced an array of political and practical challenges. Cutting jobs, never a popular initiative, was among the reformers' first tasks. Tanzania's economy offered limited private sector options for employees who were let go. In 1991, Mwinyi, Mkapa's predecessor, had introduced his own Civil Service Reform Programme to identify ghost workers and reduce the size of the public service. By the mid-1990s, there was little to show from the effort except roughly 80 preparatory studies. Despite consensus at the senior level on the need for change, "there were people who were benefiting from this mess," Rugumyamheto observed. Consequently, it was very hard to move forward without eliciting protest, even though there were no public service unions at the time. Mkapa knew he would have to draw on his knowledge and political skill to overcome opposition, then move quickly to produce results before the next election. His background served him well. He had begun his career as a civil servant in the Ministry of Foreign Affairs, worked as a journalist and then moved into a series of positions that gave him broad exposure to conditions and opinions, both in his own country and around the globe. He had served as editor-in-chief of the ruling party newspaper, press secretary to then-president Nyerere, ambassador to the United States, and minister of foreign affairs. He knew people in the party, the legislature and the civil service, and he put that familiarity to work. On taking office in 1995, President Mkapa moved quickly to build a core political coalition supportive of his proposed changes. Even detractors of the reform movement did not block Mkapa's proposals: "In terms of the politicians, they were prepared to give me as much rope as possible so to make it easy for them to hang me when I failed," he wryly observed. "So I didn't really have very much problem there." The early period of reform proved the most challenging. Mkapa reminisced, "The situation was pretty difficult. People would admire my courage, some would cite my foolishness in undertaking these reforms, but when the economy started turning around and revenues began to increase and then after we had reached out to potential investors both domestic and outside and investment starting picking up, they began to see that there really had been no way out except these reforms had to be undertaken." Given broad support for the ruling party, Mkapa leveraged his position as party head to win politicians to his side. The party's national executive committee and its central committee both met regularly and Mkapa used these occasions to explain his proposals. Still, building a coalition was difficult. People were reluctant to object too loudly as long as Mkapa had Nyerere's backing. Mkapa banked on his ability to build support as the economy improved, initiating some difficult reforms early in his presidency. Mkapa tried to breathe life into the reform initiatives by vesting responsibility in a reform secretariat located within the President's Office. The secretariat would operate independently from other Civil Service Department operations and separately from the ministries. Mkapa also picked a team. "I had been in government for quite some time, so I knew who was who," he observed. Matern Lumbanga, then head of the civil service, helped Mkapa identify Rugumyamheto and George Yambesi, who would ultimately be Rugumyamheto's successor. Mkapa and Lumbanga also selected permanent secretaries for key ministries. By 1996, the reform secretariat had a cohort of public servants drawn from various ministries or pulled out of retirement and sent for special training, locally or abroad. It also drew on the help of some technical advisers, Tanzanians with broad experience or non-nationals with special skills and information. Mkapa believed that competent advisers, including outsiders, could "tell you the realities which the civil servants may not be able to tell you and which your fellow politicians may be hiding from you." They could assume some of the analytical burdens others could not manage. The president also opened his door to the secretary of the cabinet, the head of the civil service department, and the permanent secretary in charge of reform. He encouraged the reform leaders to speak candidly and he kept them in their positions for years, establishing a strong rapport. He also encouraged persistence in the face of obstacles. "I think much depends on the political leadership of the country," he observed. "You must keep the flag of reform flying, which means you must breathe down the necks of those who are charged with implementation. If you relax a little bit it's not enough." FRAMING A RESPONSE The reform team's ambition was to create a public service that the country could afford, that delivered services more efficiently and effectively, and that planned ahead and adjusted to changing demands and conditions. The primary goals of the program's first phase were to reduce staff numbers, use some of the savings to pay better salaries to those who remained, create systematic controls over staff size, and introduce a more professional and business-like structure for the civil service.5 Given the limited number of managers in the country, Rugumyamheto noted that it was important not to overextend supervisory loads. "Have fewer ministries and make sure they are focused on easy to identify outputs and outcomes," he urged. However, "downsizing" and "rightsizing"-the policy terms for adjusting staff size-posed four major problems. The first was how to identify the people to remove. The second was how to mitigate the social cost of widespread retrenchment, including helping people find new jobs. The third was how to identify and eliminate ghost workers from the payroll. The fourth was how to ensure that the public service sector didn't simply grow again, a common problem among countries that make big staff cutbacks. In devising their approach, Mkapa's team looked at the experiences of other countries, including the highly experimental approaches of Singapore, Malaysia, and New Zealand. The team sought to expand job options for people whose employment was terminated as a result of the reforms. Yambesi, an expert on redeployment already on staff, was responsible for the human face of the program and later became the director of policy development and the coordinator of implementation efforts. At the outset of his presidency, Mkapa made important decisions that influenced the later course of reform. He imposed strict limits on the size of his new cabinet, partly at the instigation of the planning team. Previous regimes had cabinets of approximately 30 members, whereas Mkapa's was 17. Mkapa's technical adviser, Kithinji Kiragu, noted that a smaller cabinet was critical to reducing overall headcount: "[I]f you are going to get all these people appointed as ministers, they are not going to be the ones deciding to make a smaller cabinet. It means they would be sacking themselves."6 GETTING DOWN TO WORK The reformers proceeded on several tracks at once. Their efforts built on an earlier legislative authorization to overhaul the civil service, negotiated by the previous president Ali Hassan Mwinyi. Mkapa's team knew that they had little time left on the clock; the authorization would expire at the end of 1998, so they moved quickly, hopeful that the ultimate success of their initiatives would extend their mandate. Reducing the number of employees To reduce the size of the civil service, the reformers first identified the tasks that ministries alone could do, versus tasks that semi-autonomous public agencies or the private sector could handle just as well. The reformers organized teams within each ministry. The teams identified which offices had responsibility for policy formulation, implementation, or regulation. Divisions, sections or posts that were "operational" and concerned service delivery were subject to the litmus test. Shifting functions to the private sector was not always as feasible as it was in other countries. As one reform manager put it, the Tanzanian government had long been in the business of selling "meat and tomatoes" when farmers and traders could do those tasks more efficiently. It might be possible to reduce the public payroll by spinning off some of these services. However, the private sector was developing slowly, and privatization was not always available as a short-term option. The easiest functions to shift to private firms were cleaning, messenger services, and building security. Private companies that could provide these services already existed, and the tasks did not include policy decisions, which would properly belong to the government. The reformers moved quickly to contract with private providers. Later they identified additional services, like cafeterias, that commercial firms might supply, eventually moving 48 non-core tasks out of the public sector, with varying degrees of success. If the private sector could do the job but the function served an important public purpose, the government transformed it into a semi-autonomous agency. The agencies remained part of the ministries but they were no longer under the permanent secretaries' day-to-day management. The hope was that these agencies, which could charge fees for their services, could help pay for themselves and cover part of their own wage bills. Each would have a governing board made up of government representatives and people drawn from business and civic groups. To reduce the number of employees further, the team chose a combination of voluntary retirement and layoffs. The reform program encouraged employees over 50 and those who felt their jobs were at risk to retire early, offering generous departure packages. Not everyone could apply for early retirement; professionals were prohibited because their skills were in short supply. Not all departures were voluntary. Management committees made lay-off decisions. Within a year of the program's start, 2,212 low-performing employees had been let go.7 Overall the civil service's size dropped from 355,000 in 1992 to 264,000 in 1998, a reduction of a little over 25% or about 90,000 people.8 Designing retrenchment packages Because the government was the largest employer in the country, any reduction in staffing was likely to have a significant effect on overall unemployment, human development and political support. The government realized that it needed to cushion the effects. To reduce anxiety and ensure that the program worked smoothly, the reformers launched an information campaign directed at civil servants and the public, explaining the purpose and method of the downsizing initiatives. The retrenchment strategy had several elements. All those who left the public service, whether they accepted early retirement or found their jobs eliminated, received a compensation package. One program provided start-up capital for small businesses, a form of "golden handshake." Another provided re-training to about 25,000 people. In some instances, people had the option of taking related jobs with private sector firms, which were then expanding to provide some of the services that had been eliminated through the reforms.9 The program also helped pay relocation assistance to those who wished to return to their homes in distant parts of the country. Counselors worked with employees who were laid off and provided advice about where else they could look for employment. The exercise was not cheap. Although salary savings helped to offset the cost of compensations packages, outside donors were instrumental in creating a retrenchment fund. From 1992-1998, donors contributed over US$90 million that the government could use to make severance payments to those laid-off.10 The system worked well on balance, but auditors later noted several shortcomings. For example, it was hard to correct mistakes in compensation packages because the office that made the disbursements issued vouchers manually and did not file the records by date. Consequently, it took a long time to retrieve information about specific payments. The Central Payments Office sometimes did not know when a change had taken place because there was no electronic indication. Supervision sometimes broke down, and neither the Civil Service Department nor the Central Payments Office followed the rules they had set out for themselves. Further, the people who staffed the process were often unavailable and worked more slowly than anticipated. Canceled checks were not always canceled immediately, and the payments office did not always pass information on returned checks back to the civil service department. Economic policy changes and favorable global market conditions accelerated growth at the same time public sector layoffs took place, enabling some of those let go to find jobs in an expanding new private sector. Annual growth in gross domestic product rose from less than 2% before 1995 to an average of 4% in the second half of the 1990s-and to over 6% at the end of the decade. Per capita income rose steadily as well. Rugumyamheto suggested that although the private sector was small, it was able to absorb the people laid off from the civil service as it grew into the areas government was leaving.11 Yambesi also reported that most workers who lost their jobs were reemployed elsewhere and many started new businesses, benefiting from the more open economic climate. Removing ghost workers While a major reform objective was reducing redundancy, Tanzania also faced a ghost worker problem. As the retrenchments went into effect in the mid-1990s, inadequacies in the payroll and record keeping systems, coupled with difficulties integrating the Ministry of Finance and civil service records, meant that some people continued to collect pay after termination, while new employees often failed to receive their paychecks. When the reform program started, Tanzania had no electronic personnel records. In most ministries, a manager had to request information about an employee from several people, each of whom maintained information in paper files. The human resources information, scattered among managers, was also separate from the payroll system, and early attempts to link the two through a simple software program had failed. As downsizing moved forward, the reform leaders realized they needed to improve record keeping and payroll management so they could keep track of the number of employees and the wage bill. The construction of a single, streamlined data program began in 1998, before the official assessment of the first reform initiatives; it would take several years for the data program to meet its objectives. At first, a management systems information unit, established by the reform team, organized workshops with permanent secretaries and directors of personnel administration from various ministries to discuss what they wanted the system to do. The unit started to collect the data the system would use, asking representatives in ministries and agencies around the country to complete forms it had designed for the purpose. At the same time, it developed details of hardware and software needs in order to issue a tender. The government advertised for bids through the Charles Kendall Group, a UK company, with a requirement that international companies bid with local partners. Lawson Software, a company based in St. Paul, Minnesota, won the bid. Lawson's South African subsidiary worked with a local partner, CATS Tanzania Ltd., and conducted a further set of workshops with the ministries to demonstrate the system and to ensure that ministries were ready for launch. Once the basic system was in place, reformers conducted a major audit of the payroll to check for accuracy and find ways to streamline the system. At the time, two lists of civil servants existed. One was in the Public Service Management office, the reformers' center of operations, and the other was in the Finance Ministry. These two lists were cross-referenced; the public management reform team identified records that matched in both lists and those that could be found only in one. Those entries on both lists were entered into the new combined human resource/payroll management system. The Public Service Management office then followed up with those individuals who appeared on only one list and sent requests to supervisors to verify employment.12 This step helped the reformers purge ghost workers and create a coordinated pay and employment list. In the first round, the reform team was able to identify about 14,000 ghost workers and remove them from the rolls. Moving functions to agencies As the reform program moved forward, Rugumyamheto and his team decided that some types of government offices merited a different status because they provided licenses or other services critical to the smooth functioning of a market economy. If these services worked poorly, investors and traders would face higher costs. Moreover, the services were important enough that private citizens and firms might be willing to pay user fees to help finance the services. In an effort to streamline government and increase its efficiency, the reform team decided these offices could become semi-autonomous agencies, governed by public-private boards. The team helped draft legislation to create the units and define their relationship to the rest of the civil service. This initiative drew on experiences in a number of countries, including the United Kingdom, whose "Next Steps Executive Agencies" program provided a partial model. The UK system, which began in 1979, aimed to improve efficiency and accountability by creating clear objectives and citizen contracts, or charters, for government services and by conferring greater flexibility in organizational structure and management practice. Although the agencies remained part of the government, they operated more as businesses did, and their executives reported to oversight boards. The UK had rapidly created 133 agencies encompassing 77% of the civil service. The changes were part of an experiment that many countries had embraced in "New Public Management," or NPM. In 1997, the Tanzanian parliament passed a bill creating executive agencies. The units, put in place in 1998, carried out functions such as business licensing, land lease and registration, road construction, dams, statistical services, weights and measures, the weather service, several training institutes, and aviation control. Under the new laws, the agencies would operate under an "eyes on, hands off" system, with much greater flexibility in day-to-day and strategic management. The relevant ministries maintained oversight of agency activities, but the operational details were left to the agencies themselves. The agencies operated differently from ordinary government units. The chief executive officer (CEO) was responsible for developing business and strategic plans, reporting annually on results, and accepting a yearly independent audit. The CEO could appoint and manage staff without reference to the Civil Service Department, although an approved performance appraisal system had to provide the basis for decisions. The agencies received part of their budget from the central government, but they could also raise revenues themselves from fees charged. They would have far greater latitude in how to use their resources, such as augmenting standard salaries through revenues. (By 2005, average pay in these agencies would be 57% higher than in the other parts of the civil service.) The executives of the agencies were retained through performance contracts of five years, subject to re-appointment if the executives met the criteria in their contracts. The decision to retain or dismiss an executive was vested in the relevant ministry, advised by a board of key private sector stakeholders and the relevant ministry Permanent Secretary and other officials. This board also monitored the agency's performance and reported to the ministry every quarter. Under the new law, the government created a total of 24 executive agencies. Most of these got their start in the first reform period that ended in 1998, as part of the effort to trim and reorganize the civil service. The goal was to create about 20 more, later. The executive agency initiative encountered several obstacles. Eric Shitindi, deputy permanent secretary for public service management, explained, "You find most of the people who are managing these agencies ... are people who are in the civil service. So if you continue managing the agency as a civil service entity instead of a business entity, then definitely your performance will not be very good." Finding the right levels for user fees and government budget contributions could be very difficult. Because the agencies could use revenues they collected to increase their own pay, there was some risk of charging high fees without much regard for what would happen in the future. Shitindi remarked, "We know how people can be motivated to charge a very high cost, because they think if they generate more revenue, they also have more pay." If they faced full market competition, they might go out of business. The continued support from the government limited that discipline. However, oversight by public-private boards could potentially minimize this problem. The relationship between the agencies and their home ministries was also a potential source of trouble. Shitindi observed that the agencies that did not perform well were often those that had a poor relationship with their ministries. Frequently the home ministry personnel did not understand the accrual accounting systems used by the agencies, contrasting with the traditional cash budgets the ministries maintained. In addition, ministries often made their contributions to agency budgets on a monthly basis, and the timing and amounts proved unreliable. Over time, poor relationships translated into decreased support that further undermined an agency's performance. Tanzania's internal evaluation conducted in 2004 noted several successes. For example, the time period to receive a birth certificate had shrunk from three years to less than three days, and the Business Registration and Licensing Agency cut the time to obtain a license in half. Land titles had always taken a long time to secure, and the slowness affected many people. Improvement in managing land titles generated popular enthusiasm for the reform movement. There were significant increases in performance in statistical services, the Institute of Accountancy, the Tree Seed Agency, and the Public Service College. The aviation authority enhanced airport services. The 2004 evaluation cautioned that these services might have improved if left within the ministries and managed differently. Nevertheless, analysis of 127 different indicators suggested that performance had indeed improved.13 Improvement, however, was not uniform. As a group, the agencies met their service delivery targets only 60% of the time in the first 5 years.14 The review suggested that performance had declined in a few areas, including some of the training programs. Physical infrastructure projects-water drilling, dams, and roads-got off to a slower start. Most of the agencies collected more revenue, but they were less successful in cutting costs, though some of the initial expenditures appeared to include one-time investments in improved infrastructure. Staff performance in the agencies was generally better, but vacancy rates were also high for some types of managerial and technical personnel. Although the agencies initiative was one of the more successful components of the larger civil service program, the reform team and observers both emphasized the incomplete nature of the reforms. Rakesh Rajani of Twaweza, a non-governmental organization devoted to improving access to information, acknowledged the success of the program but cautioned against complacency. He worried that satisfaction with limited success, measured mainly by improvement against a low standard, was an impediment to really making things work. As an example, he reasoned, "If your benchmark is how things used to be, certainly the Tanzanian Revenue Agency has done much better. If your benchmark is what it could be compared to what problems still continue to persist, then I think there's still a lot more to be done, and it could have certainly done better than it's doing right now. So I think it's a mixed record. I think it's a glass half full, half empty."15 The government's own evaluation supported this view.16 Limiting employment growth Although the 25% reduction in the size of the public sector work force was a sizeable achievement for the Tanzanian government, a one-time improvement meant little by itself. In many countries that cut personnel earlier or at about the same time, maintaining a cap on public service employment had proven difficult. Most saw their employment numbers climb again. By contrast, Tanzania's civil service managers tried to limit growth and generally linked increases in employment to expanded service delivery. Strong resource constraints sharply limited the government's ability to increase the size of the civil service.17 Desire to meet the Millennium Development Goals-poverty reduction targets set by the international community in 2000-created demand for more teachers and health workers. Beginning roughly in 2002, the government expanded teaching, health programs, police, anti-corruption work, and a few related regulatory functions. However, apart from professions like teachers and nurses, the size of the civil service stayed remarkably stable through 2008. In 2004, with a slight increase in size, Tanzania had roughly one civil servant for every 100 citizens-at the low end of the scale for low-income developing countries and well below the levels in some more industrialized countries.18 In a later phase of the reform program, the government adopted a more formal recruitment and employment system, which also helped maintain the cap on the size of the civil service. The chief secretary in the president's office, the head of the civil service, and the reform unit assumed greater control for monitoring and managing the size of the civil service. Managers could appoint people only to positions already budgeted and approved. The chief secretary was responsible for new posts, while replacement hiring had to win the approval of the reformers in the public service management office. Together with increased emphasis on merit-based hiring and promotions procedures, this high-level oversight strengthened discipline. The big question was whether future presidents would share the concerns of the reform team or increase the number of cabinet positions, departments, and employees in a bid for political gain. By creating a strong norm, the reformers hoped to contain these political pressures. OVERCOMING OBSTACLES The layoffs and reorganization met some resistance, although not as much as some anticipated. On the whole, politicians cooperated in the medium term, even though the payoff in terms of public enthusiasm and votes would not come quickly. Rugumyamheto, the permanent secretary who oversaw much of the retrenchment, suggested that the need for major downsizing in the 1990s was self-evident and therefore difficult to resist rationally. The sheer size of the civil service and the needs of the new government made downsizing non-negotiable, limiting the space for resistance from senior civil servants and politicians.19 To smooth the process, the reform leaders held seminars with legislators to provide more information about the process and goals. Developing rapport with members of the public and with the civil servants themselves was more difficult. The advocates of the initiative tried to win people over. "A lot of communication was done by the leadership, to try to explain [the need for reform]," noted Professor Bana. However, members of the public "thought that the public service [was] there to create employment for all," Bana observed. As a result, building a popular constituency for reform was challenging. At the same time, those who remained in the civil service also started to grumble. Some thought the changes made them work harder. Others noted that the reformers had promised better pay for those still in the public service, and although there were across-the-board increases in remuneration, improvement was sometimes nominal. Citizens also found that some of the services they used in the past were unmanned because employees had lost their jobs. Stretched thin, the reform team initially spent too little time trying to manage public expectations. There was no quick improvement in performance, either. The ultimate aim of the program was to make services more effective and efficient, but the initial focus of the program was strictly on adjusting size, organization, and record keeping. These steps laid important groundwork for introducing performance management, but they did not produce a rapid increase in effectiveness, except in some of the executive agencies that provided passports, licenses, or airport services. During one evaluation in which reviewers posed as citizens and asked for information, most civil servants were unresponsive and ignored the questions they received. The Ministry of Education performed especially poorly, as did the anti-corruption bureau, communications, and transportation.20 There was more work to do. Mkapa himself tried to engage restive public servants. He emphasized the need to balance concurrent needs. "I would lay it out to them and say, where do you expect me to get the money to meet the costs of all these requests that you are making to me? ... I remember going to one contingent of the uniformed services and there was a mantra of needs. So I said, 'Fine. My budget this year is this much. How would you distribute it so that you can meet the costs of all these needs you have outlined, because if you were to take all this money and meet your needs it means your father, who is back in the village, will not find any medicine when he goes to the health center. So what do you want, medicine for your father or this stuff?' That kind of openness and truthfulness helped very much to enforce public confidence in the government." An increase in economic growth and reduction in inflation-generated uncertainty gradually helped ease the pain of the reform movement. GDP per capita in constant USD, rose from $288 at the start of the reform period to $318 by the beginning of the second phase in 2000. Inflation dropped from 26.9% in 1995 to 7.6 in 200 (and to 5.3 the year after).21 Civil service salaries rose by about 75%, on average. Mkapa noted, "In the first years here, I think it was very difficult to persuade [people] that my reform obsession was justified, because it involved tightening the belt for everybody, including the civil service." He continued, "In Swahili when we play cards there's a point where you don't win anything, you say kapa-na kwenda kapa. My name is Mkapa, you see. So they asked, 'What is this kapa economy that this man is building around the place?' But I was patient enough to go on. Within three years everyone forgot about the kapa economy because they had more in their hands and it is much more than they had ever had before." ASSESSING RESULTS In 1998, the government sponsored a major symposium to evaluate the results of the Civil Service Reform Programme. The first phase of the reform program had generated some clear successes. The civil service was 25% slimmer and the country's newly growing economy could support the wage bill, which had fallen below the average level for Sub-Saharan Africa and for Africa's low-income group.22 Some of the executive agencies had enhanced their services in ways that made the public happier and helped make it possible for the economy to grow. The government had started to develop the information system needed to manage the civil service more effectively, although most of the improvements still lay four years in the future. Within the context of this success, Yambesi, the reform secretary, remained dissatisfied. He noted that the reform team "had taken measures to reduce the cost of running government, but the glaring issue which was coming out was that services were not improving ... So what was wrong? [T]he objective was actually to improve services, to improve access to service."23 The Tanzanian reformers noted emerging challenges, too. The downsizing was important, but it also created difficulties. For example, Rugumyamheto observed, "By reducing numbers just on the basis of fiscal considerations, it means we didn't take into consideration human resources planning and succession planning requirements." One of the medium-term consequences was a generation gap that began to threaten the capacity of the civil service later, about 10 years after the initial cuts took effect. Many of those who took the compensation packages and left voluntarily during the mid-1990s would have moved into the middle and upper management within a few years. As a result, 10 years later, in 2008, the service suffered a critical shortage of skills and experience in the managerial ranks. There were fewer candidates for higher positions at just the time when improved management was crucial for efficient operation and effective reform implementation. In response, reformers later created a fast-tracking program. High-potential employees in junior to mid-level positions were allowed to advance through the ranks more quickly than would normally have been the case. REFLECTIONS Tanzania's civil service reform team generously shared their reflections on what worked and what they would do differently. They offered these insights both during the reform movement and with the benefit of hindsight. They wrote about their program and hosted visits from aspiring reformers from other countries. They discussed the need for high-level leadership support, continuity of personnel to carry out the changes, and a sense of ownership on the part of ministries and managers. By the late 1990s, it was clear that greater ownership of reform within the ministries was one of the top priorities. Creating an independent reform secretariat within the Civil Service Department in the Office of the President had enabled quick implementation of reforms because team members were able to operate outside the bounds of normal civil service structures. However, this isolation proved increasingly problematic. The secretariat relied on managers in the Civil Service Department to carry out initiatives, but it had not cultivated a sense of ownership and had not conducted extensive capacity development either within the department or in the ministries themselves. The pay differential between the secretariat's staff and the rest of the civil service caused discord. Members of the secretariat, including very recently retired civil servants, were paid several times more than their colleagues in the regular civil service. Others grumbled that the secretariat was too closely aligned with donor interests. It was hard to produce broad change in the performance of the civil service without more engagement from the ministries themselves. In the next phase of the Tanzania reform program, building a sense of ownership and boosting performance would become main objectives. To move forward, the reformers also altered their own organization and operating procedures. 1 United Republic of Tanzania. President's Office. Civil Service Department. Public Service Reform Programme Strategy & Action Plan, 1998-2003, v. 1 (December 1998), draft for discussion, p. 4. 2 The World Bank Salutes Reform Efforts in Tanzania, Panafrican News Agency (Dakar), 24 February 2001. 3 Kithinji Kiragu and Rwekaza Mukandala. "Public Service Pay reform: Tactics Sequencing and Politics in Developing Countries: Lessons from Sub-Saharan Africa," Draft, January 2003, p. 136. 4 United Nations, Division for Public Administration and Development Management. United Republic of Tanzania Public Administration Country Profile, July 2004, p. 4. 5 Graham Teskey and Richard Hooper. 1999. Tanzania Civil Service Reform Programme: Case Study. Joint DAC Informal Network/ACBF Workshop on Institutional and Capacity Development: Harare, Zimbabwe. Available at http://unpan1.un.org/intradoc/groups/public/documents/CAFRAD/UNPAN010583.pdf 6 Kiragu interview. 7 Emmanuel Lukumai. "The Implementation of Civil Service Reforms in Tanzania, 1991-2000," Master's Thesis, Department of Administration and Organization Theory, University of Bergen, Norway, June 2006, p. 41. 8 J. A. Rugumyamheto. 2004. "Innovative Approaches to Reforming Public Sector Services in Tanzania," Public Administration and Development, 24(5): 437-446 passim. 9 Hooper and Teskey. passim. 10 Teskey and Hooper, p.7. 11 Rugumyamheto Interview. 12 Kiwango interview. 13 United Republic of Tanzania. President's Office-Public Service Management. "State of the Public Service Report 2004," 23 June 2005, Dar es Salaam. Issued by perm sec PO-PSM, p. 17. 14 United Republic of Tanzania. President's Office-Public Service Management. "State of the Public Service Report 2004," 23 June 2005, Dar es Salaam. Issued by perm sec PO-PSM, p. xi 15 Rajani Interview. 16 Monitoring and Evaluation Unit, President's Office - Public Service Management, Do Executive Agencies Perform Better, January 2005. 17 Morin Interview. 18 United Republic of Tanzania. President's Office-Public Service Management. "State of the Public Service Report 2004," 23 June 2005, Dar es Salaam. Issued by the permanent secretary, PO-PSM. 19 Rugumyamheto Interview. 20 United Republic of Tanzania. President's Office-Public Service Management. "State of the Public Service Report 2004," 23 June 2005, Dar es Salaam. Issued by perm sec PO-PSM, pp. 64 and 69. 21 World Bank. World Development Indicators online. 22 United Nations, Division for Public Administration and Development Management. "United Republic of Tanzania Public Administration Country Profile, July 2004, p. 4. 23 Yambesi Interview. Innovations for Successful Societies makes its case studies and other publications available to all at no cost, under the guidelines of the Terms of Use listed below. The ISS Web repository is intended to serve as an idea bank, enabling practitioners and scholars to evaluate the pros and cons of different reform strategies and weigh the effects of context. ISS welcomes readers' feedback, including suggestions of additional topics and questions to be considered, corrections, and how case studies are being used: iss@princeton.edu. Terms of Use In downloading or otherwise employing this information, users indicate that: a. They understand that the materials downloaded from the website are protected under United States Copyright Law (Title 17, United States Code). This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. b. They will use the material only for educational, scholarly, and other noncommercial purposes. c. They will not sell, transfer, assign, license, lease, or otherwise convey any portion of this information to any third party. Republication or display on a third party's website requires the express written permission of the Princeton University Innovations for Successful Societies program or the Princeton University Library. d. They understand that the quotes used in the case study reflect the interviewees' personal points of view. Although all efforts have been made to ensure the accuracy of the information collected, Princeton University does not warrant the accuracy, completeness, timeliness, or other characteristics of any material available online. e. They acknowledge that the content and/or format of the archive and the site may be revised, updated or otherwise modified from time to time. f. They accept that access to and use of the archive are at their own risk. They shall not hold Princeton University liable for any loss or damages resulting from the use of information in the archive. Princeton University assumes no liability for any errors or omissions with respect to the functioning of the archive. g. In all publications, presentations or other communications that incorporate or otherwise rely on information from this archive, they will acknowledge that such information was obtained through the Innovations for Successful Societies website. Our status (and that of any identified contributors) as the authors of material must always be acknowledged and a full credit given as follows: Author(s) or Editor(s) if listed, Full title, Year of publication, Innovations for Successful Societies, Princeton University, http://successfulsocieties.princeton.edu/ Andrew Schalkwyk and Jennifer Widner Innovations for Successful Societies (c) 2012, Trustees of Princeton University (c) 2019, Trustees of Princeton University