IMPROVING THE POLICY PROCESS: GHANA TRIES TO BUILD SUPPORT FOR CABINET DECISION-MAKING, 2003 - 2008 SYNOPSIS From the 1960s to the early 1990s, Ghana's Cabinet-level policy management system deteriorated as multiple coups d'état produced abrupt changes in government. Many competent civil servants either left or were pushed out. Ministries submitted policy documents to the Cabinet that lacked essential information ministers required to evaluate the wisdom and feasibility of proposals. Ministries rarely cooperated with each other. But beginning in 2003, a newly formed policy unit in the presidency partnered with the Canadian International Development Agency to strengthen Ghana's policy management system. The unit helped coordinate policy planning between ministries and reported on implementation to the president. The Cabinet Secretariat introduced standardized formats to guide ministries in policy development and ensure that proposals contained all essential information. The Office of the Head of the Civil Service and the University of Ghana Business School worked together to train hundreds of civil servants in the practical skills of researching, writing, and communicating policies. By 2008, the new system was in place and the policy management process had improved, but sustaining the reforms through the tumultuous government transition that followed the country's 2008 elections posed additional challenges. Looking back on the effort, Samuel Somuah, who helped lead the Ghana Central Governance Project, underscored the importance of an effective policy management system by saying, "If there's one project every African country needs, or every developing country needs, it's this project." Jonathan Friedman drafted this case study based on interviews conducted in Accra, Ghana, during April 2013. Case published October 2013. INTRODUCTION Ghana's 2000 election marked the first transfer of power between civilian leaders in the country's history. "The previous government was in power 18 years, and this new government came in with strong democratic credentials," said Samuel Somuah, special adviser on information and communications technology (ICT) to President John Kufuor and project manager of the Ghana Central Governance Project, a government effort to strengthen Ghana's policy management system. But beneath the surface of that good news were challenges, Somuah said. "Because of the history of military governments, we did not have a system where we had really good people with very good policy-making skills who could take an issue and come up with a framework for tackling it." Kufuor's election was an important step in the consolidation of Ghana's new democracy. Jerry Rawlings, who came to power through a coup d'état in 1981, had initiated and won multiparty elections in 1992 and 1996 before stepping aside in 2000, when he reached the constitutional term limit. Kufuor's victory over Rawlings's vice president and chosen successor, John Atta Mills, marked for Ghana the first transfer of power between civilian leaders and the first time one political party had ceded control to another. Kufuor took office with high aspirations. He delineated an ambitious agenda built on five policy priorities that would drive his administration: good governance, modernization of agriculture, private sector participation, infrastructure development, and enhanced social services, including health and education. Circumstances forced a sixth priority: macroeconomic stabilization. Ghana was in the midst of an economic crisis in which the Ghanaian currency nearly halved in value relative to the US dollar. The government's deficit had reached unsustainable levels the year before Kufuor took office.1 Although Ghana faced significant challenges, international goodwill in the wake of the election and support for Kufuor's macroeconomic stabilization plan presented equally significant opportunities. Located between the West African countries of Ivory Coast and Togo, Ghana was roughly the size of the US state of Oregon, with a population of 25 million and an economy and workforce that relied heavily on agriculture. To achieve his goals, Kufuor needed a management system that could transform his priorities into well-designed, collaborative, practical, and effective policies. THE CHALLENGE Ghana's 1992 constitution and 1993 civil service law had elevated the role of policy making in Ghana's central government. The constitution empowered regional and local governments to take the lead in delivering services, leaving the central government to focus on formulating policies, managing finances, and supervising implementation. The civil service law outlined the institutional arrangement to support those functions. It set up directorates in all central ministries for policy planning, monitoring, and evaluation to formulate and supervise policies, alongside research and statistics, finance and administration, and human resources units. However, Ghana's policy makers found that it was hard to develop sound policies, because ministries could not recruit enough analysts with the right skills; neither did the government issue sufficient guidance to ministries or share information effectively. To execute their crucial and complex mandates, the ministerial policy units needed creative analysts with sharp minds. But it was difficult to find and retain the right people. The problem was not lack of talent, said Kwaku Appiah-Adu, head of a presidential policy unit set up in 2002, but the relative attractiveness of other government units. Some civil servants sought positions in human resources directorates because of the power they could wield in making decisions about recruitment, transfers, and promotions. Similarly, finance units were attractive because of the opportunities to influence procurement contracts. By contrast, "Policy units were perceived as brain work with relatively limited benefits," Appiah-Adu said. Some of the larger ministries had high-performing policy units, he said, but most did not. Officials in the ministry policy units also lacked rules and procedures to direct their roles in the policy formulation process. Previous administrations had developed a few guidelines, but not enough people knew about them. As a result, policy proposals often lacked essential information, which made it difficult for the Cabinet to make informed decisions. Frank Mpare, a longtime civil servant who became Cabinet secretary in 2005, said Ghana needed a comprehensive manual for Cabinet documents. He was not alone in that view. Bridget Katsriku, head of the Public Services Commission, a government agency that supervised agencies' structures and personnel requirements, also said policies were not adequately researched and often lacked clear descriptions of problems, policy options, impact assessments, and implementation plans. She offered an example from her time as the top civil servant, or chief director, at the Ministry of Employment and Social Welfare, which was the ministry in charge of the promotion of employment plans and capacity-building programs. The Cabinet had approved a proposal to train and offer seed money to promising young entrepreneurs but did not consider the financial implications of the plan. Her ministry "didn't do the analysis to see if we had the money," she said. As a result, entrepreneurs "brought business plans and we had trainings, but we couldn't offer money." Lack of coordination was especially troublesome. Ministries tended to develop policy proposals on their own, without seeking input from other parts of the government. Appiah-Adu described ministries as having a "silo mentality"-a term meaning a cloistered administrative culture. Neither was there follow-through within and across ministries. Nana Dwamena, executive secretary of the Management Services Department, an in-house management consultancy for the central government, said nonimplementation of Cabinet decisions was a prime weakness and another example of a coordination failure. Information systems and communications systems further aggravated the coordination and implementation challenges: Many government officials did not have Internet access. There was no corporate e-mail system at the office of the president. And employees had to use personal e-mail accounts or rely on couriers and the postal service, said Somuah, the ICT adviser at the presidency. "The government didn't have an official Web site," he said. Judy Cavanagh, a Canadian project manager and former civil servant with experience in a Canadian cabinet office, added, "IT [information technology] systems were basic-no intraconnection, no secure e-mail system, nothing on document management, no tracking, no ability to track from beginning to end of decision making. Everything was manual." In many countries, a cabinet secretariat or a presidential policy unit facilitated consultation and coordination across government, ensured that policies were well designed and included impact assessments and action plans, monitored implementation, and reported progress to the Cabinet. No such system existed in Ghana. In the 1990s, the existing presidential unit monitored ministries' activities, but lacked the broad mandate required to manage the policy system from start to finish. Without significant changes in human capacity, systems, and infrastructure, Ghana's policy makers would continue to struggle to make government work. FRAMING A RESPONSE In 2001, the Canadian International Development Agency (CIDA) approached the chief of staff at President Kufuor's office with the idea for a project that would strengthen policy capacity in Ghana's central government. Policy management and "good governance" had become one of CIDA's overseas development priorities. CIDA had approached the Rawlings administration near the end of Rawlings's second term, but the proposed Ghana Central Governance Project was not taken up at the time because of the upcoming elections. "When the Canadians told us the ideas [in 2001], I was really excited about it and [so was] my boss at the time, Honorable Jake Obetsebi-Lamptey, chief of staff to the president," said Somuah, "because we recognized that a lot of the objectives of the project were things we could identify with." The project took shape through a series of discussions between the presidency and CIDA in 2001 and 2002. Somuah and others from the presidency traveled to Ottawa and Washington beginning in early 2001 to observe cabinet office policy management systems in action and to evaluate what might work in Ghana. The Canadian government financially supported the project. The Canadians and the Ghanaian government moved forward together in setting up project teams and hiring advisers. In early 2002, the president's office formed the Policy Coordination, Monitoring and Evaluation Unit (PCMEU) to conduct much of the day-to-day work the project envisioned. Appiah-Adu, a strategy and management specialist with PricewaterhouseCoopers LLP, a US-based audit and consulting firm, came on board to lead the unit. Appiah-Adu hired 12 staffers with expertise in engineering, economics (specifically, agricultural economics), legal affairs, and social sector policy. Through an open tender, the Canadian government invited bids from organizations that could provide technical support for the project. The Institute of Public Administration of Canada (IPAC), a network of public servants and academicians that had often advised foreign governments on public administration, won the contract. Kufuor's chief of staff together with the Canadian High Commission's director of international development led a steering committee. Presidential adviser Somuah and IPAC's Cavanagh worked together as project managers. The project formally launched in early 2003 and was initially intended to continue for five years. The project created working groups to address the three areas of weakness in Ghana's policy management system: weak capacity in ministerial policy planning units, lack of clear policy formulation procedures and effective oversight, and inadequate technological infrastructure to support an integrated planning system. Appiah-Adu led the first two project teams, and Somuah, with his IT background, led the third. The teams included officials from the presidency, eight strategic line ministries, the National Development Planning Commission, the Cabinet Secretariat, and the Office of the Head of the Civil Service. The Cabinet Secretariat was a key player in the project teams because it set the agenda for Cabinet meetings and organized preparation and follow-up. The Office of the Head of the Civil Service oversaw all chief directors, who were the top civil servants in the line ministries. Cavanagh explained the necessity of including the civil service head along with the Cabinet secretary as integral to the project to get buy-in from the chief directors. Unlike in Canada, "There was a separation between the Cabinet secretary and the head of the civil service, who had authority over chief directors and therefore was critical for implementing the project," she explained. "So you have to make sure they're on the same page in talking to chief directors." Chief directors wielded considerable power in their ministries, and to have any chance of success, reforms required their active engagement. The teams went on separate observation trips to Canada, the United States, and the United Kingdom and studied policy management systems in Germany and Malaysia. "We wanted best practices from around the world so Ghana could customize them to suit our local conditions," said Samuel Aning, social sector policy adviser in the PCMEU. "It was important in that it was not forcing down our throats a particular system of analyzing and implementing policy. ... They let us bring in people from the US, UK, and Canada. We found the Canadian model very attractive, but we sent people to the US, too. They run a presidential system like ours, so we wanted to find out how we could customize the good things we found in the Canadian system to suit a presidential system." The working groups identified plans of action for the project. By the end of 2003, the project had issued a reform road map called Framework for Ghana's Policy Management and Decision Making System. The project leaders decided not to ask the Cabinet to approve the framework for fear of delays, instead deciding to seek approval as necessary-one step at a time. The PCMEU proposed to strengthen the policy management role of the presidency by building a structure that would promote consultation among ministries during policy formulation and by introducing tools of performance management that would ensure ministries acted on the president's priorities. In addition to promoting the president's policy agenda, the strengthening of the presidency's policy role could improve the overall coherence of the central government's policies by focusing ministries on a common agenda. The Cabinet Secretariat would issue guidelines to line ministries regarding the submission of policy proposals, would exercise quality control, and would track implementation of Cabinet decisions. The civil service head would train hundreds of civil servants in policy planning as part of a larger effort to make professional positions in policy units more attractive. Somuah would lead an effort to improve the connectivity of the presidency and line ministries. Cavanagh, a former senior civil servant who understood the necessity of political will to achieve reforms, said she was encouraged by the level of Ghana's political support for the improvements. "It was a new government, and there was drive from the president's office as well as enthusiasm to get the project going," she said. Dwamena, who had worked for years in central policy-making roles, said the project was intended to fundamentally alter the way policy was developed. "It was seen as a major shift in the way we were going to do things," he said. GETTING DOWN TO WORK In 2003, the first year of the project, the project leaders got the word out across the government about the reforms. "Year one was spreading awareness of what executive decision making could mean, relationship building with counterparts, and the framework document," said Cavanagh. "Year two was focusing on the tools and continuing to raise awareness." The project leaders set out to improve the central government's IT infrastructure, streamline policy planning guidelines, and train policy analysts to better execute their mandates. Wiring government To improve policy planning, the presidency had to be able to communicate with line ministries and exchange documents quickly-and it lacked that capacity. "At the start, within the office of the president, there was no infrastructure to support communication between computers, nor servers to store confidential documents," said Somuah. "It was all paper based." The project contracted with an IT consulting company to run cables that would connect all the offices of Osu Castle, the president's office. The cables terminated in a secured data center constructed and equipped with multiple servers, e-mail exchange software, and firewalls for security. Officials could access the network remotely after authenticating their identities. To provide internet access, the data center was connected to a four-meter satellite dish in the courtyard. A gov.gh domain name, a corporate e-mail system for the office of the president, and the first official government Web site were all established as part of the ICT project. But to do their jobs effectively, policy planners needed more-sophisticated tools and networks. First, the project strengthened the records management system within the presidency. With funding from CIDA, the central governance project contracted a local IT services company to install a records management system for that purpose and sent a team from South Africa to train office of the president staff in using the new system. For the first time, officials at the presidency could electronically store, track, and retrieve all of the documents they used. Next, the project's IT group connected the presidency to line ministries. "We wanted to put in collaboration tools so people working on policy could share online," said Somuah. The project provided computer equipment for ministerial policy units and connected the policy units and the presidency so they could more easily share documents. The working group incorporated special tools to assist units within the presidency. For the PCMEU, the project set up what was called the castle portal, which enabled PCMEU staff to directly access information from line ministries relevant to the president's priorities, thereby leading to close monitoring and easier follow-up. The PCMEU also acquired a tool to help monitor loans from outside donors. "Of course, the Bank of Ghana is tracking, and the Ministry of Finance is tracking, but they're making sure payments are made and following the money," said Somuah. "We're looking at: Are we using the funds? And are we getting the outcomes we expect?" Underutilization of loans had plagued the government in prior years. For the Cabinet Secretariat, the project set up a system called CabTrack, based on a system the group observed at the Privy Council Office in Ottawa. CabTrack enabled the Cabinet Secretariat to track (1) documents that came to the Cabinet, (2) decisions the Cabinet made, and (3) the implementation of Cabinet decisions. Somuah said the IT upgrades would support all the other components of the project. "We made a major effort underpinning all that we were doing with appropriate IT applications and software," he said. "We supplied lots of computers and laptops to staff to make sure they had the tools." Introducing principles of performance management "We learned that in Canada, the prime minister gives a letter to the ministers with seven or eight key objectives that he wants you to accomplish within the years you'll administer in his government," Somuah said. In Ghana, though, the president issued only broad directives to ministers, without specific targets or deadlines. "Ghana was more like the Bible: [God says,] 'Be fruitful and multiply,' but God doesn't give us intervals," he said. Appiah-Adu, head of the presidential policy unit, led efforts to position the president's agenda as a driver of activity in line ministries and to make ministers accountable for their effectiveness in implementing that agenda. He began by working with President Kufuor to issue letters to ministers modeled on letters he had seen in Canada, referred to as ministerial mandates. Each letter contained 7 to 10 objectives for the upcoming year and was signed by the president and the ministers. Ministries then developed annual work plans that spelled out how they intended to accomplish their objectives. In addition to stating key objectives, the letters required ministries to report progress to Appiah-Adu's policy unit quarterly. Appiah-Adu developed reporting templates by using models from similar programs in other countries. The PCMEU trained the policy planning, monitoring, and evaluation departments of the various ministries as a means of entrenching the process and standardizing the reporting and the outcomes. After receiving the progress reports, he would then select a few projects, based on the president's priorities, for field verification. "Verifications would follow the president's priorities," he said. "We'd split into teams and go across the nation and check whether the level of success they put on paper was consistent with what we saw." At the end of each year, the president met with his ministers to evaluate their performance based on the policy unit's reporting. When it identified projects that were lagging, the policy unit intervened in the relevant ministries. Appiah-Adu gave a hypothetical example of a state-owned enterprise (SOE)-a subsidiary of the Ministry of Energy-that had not made adequate progress toward its goals. "We would call the heads of subsidiaries and the minister into a meeting and sit down. Then we would say, 'This SOE, these were the problems we encountered with your reporting or delivery.' And then we'd have the minister read the riot act to these subsidiary heads," he said. The policy unit could also use the power of the presidency to intervene in matters that cut across ministries. The unit would bring ministerial clashes to the president's chief of staff, who would address the matters at meetings with senior ministers. When necessary, the chief of staff would ask the president to attend. "We were able to cross boundaries," Somuah said. "If you're at [the Ministry of] Education, you can't cross boundaries to the Ministry of Trade, but we can cross boundaries." A common obstacle, said Appiah-Adu, lay in late disbursements from the finance ministry, in which case the presidency agreed to adjust ministerial action plans. Sometimes the finance ministry explained that there were no funds available. "Then the presidency has to better prioritize to focus our attention on projects that we can complete," said Appiah-Adu, "rather than on the thousand and one projects that we can't complete." The policy unit's monitoring activities ruffled some feathers, and its analysis did not always sit well, according to Aning. The PCMEU's first report ranked ministries based on performance, and that caused backlash from some of the ministries that were low in the rankings. The unit met with the head of the civil service and with directors of the ministry policy units to discuss their performance, said Aning, and the meetings helped bring policy makers on board. The PCMEU did not rank ministers' performance in subsequent years. Over time, some ministers looked more favorably on the policy unit's monitoring and evaluation activities. Aning offered the example of one minister who used the policy unit's verification report to shake things up in his ministry. "One ministry had brought their report, and we had analyzed it, and we had made verification visits to some of the sites and discovered discrepancies between what was reported and what we found on the ground," he said. "Based on our discussions with the minister, he recognized and acknowledged the gaps and was able to use what we had done for corrective measures." Somuah said the policy unit's reporting was essential in providing the president and the Cabinet with critical performance information that could inform decision making. Whereas the unit performed exceptionally well, he said, politicians too often did not take the information into account when making decisions. "Where we had challenges was with the feedback loop," he said. "PCMEU would do monitoring and find gaps. When you bring issues, you expect executive action, and it wouldn't happen. . . . Most often, the reasons things weren't fixed were political, not systemic." Developing a Cabinet manual The lack of rules and standards governing the formulation of policy documents meant the Cabinet at times considered and approved incomplete and deeply flawed proposals. Ministries produced policies without adequately consulting other ministries and omitted critical information that could determine the efficacy or feasibility of policy proposals. "What we wanted to do was improve on the role and responsibilities of the Cabinet Secretariat," said Nellie Mayshak, program manager at IPAC based in Toronto, "so that it moved a bit away from being just traffic controllers-just receiving Cabinet memos-but to do a little bit of analytical oversight." Previously, there had been some rough two- or three-page guidelines, according to civil servant Mpare, who became Cabinet secretary in 2005, but nothing that provided sufficient information to guide the policy formulation process. The Secretariat studied Cabinet memorandums from Canada, Australia, and other Commonwealth countries and on that basis developed a comprehensive Cabinet Memorandum Manual for Ghana. The manual contained guidelines for the submission of proposals to the Cabinet Secretariat. A ministry had to state the problem that required a change in policy, show evidence of research, offer multiple policy options, and justify its preferred option. Policies also had to contain standard financial-, environmental-, and gender-impact analyses; implementation plans; and communication plans. If any of those components were addressed inadequately, Mpare said, the Cabinet Secretariat would return the proposal to the sponsoring ministry for further work. Mpare emphasized that the manual required ministries to be clear and accurate in their proposals. "Don't be too verbose, come up with the real facts, don't write things that don't really matter, then try to argue out certain things, options," he said. "Then, after coming out with concrete evidence, some attachments to confirm what you are saying, your conclusion, then tell Cabinet what you're asking them to approve." For Mpare, the foundation for policy proposals was the identification of a problem that really required the Cabinet's valuable time and attention. "The key point is to express the need, the justification" for the proposal, he said. "You must justify why this issue is coming to Cabinet." The manual encouraged coordination between ministries by requiring evidence of consultation. The civil service head established a process to help meet the requirement. He convened chief directors in such fields as social sector policy and infrastructure for discussions-before proposals went to the Cabinet Secretariat. The purposes were to iron out disagreements so that Cabinet ministers wouldn't have to and to solicit additional input from overlooked ministries. Mayshak, elaborating on the former, said the idea was to "save the president's time for strategic issues, to cut back on the time Cabinet takes reviewing memos." Joe Issachar, head of the Office of the Head of the Civil Service in the latter years of the project, offered an example of how those review sessions improved policy proposals. He gave the example of a proposal to expand a school child-nutrition program. One concern involved how to increase protein in children's diets. Fish was a possible source. But "the developing ministries, they forgot to consult the fisheries ministry," Issachar said. "However, at the chief directors review meeting, the chief director of the Ministry of Fisheries was present, and he gave them good input about providing protein in the meals." In addition to the Cabinet manual, Mpare and his staff developed a code of ethics for ministers and political appointees. Some of the ministers arrived at their jobs without prior experience at the top of government, and they would ask for basic information about how to conduct themselves and their work. The handbook provided instructions for ministerial etiquette when meeting with the president or foreign dignitaries, rules related to conflict of interest and receipt of gifts, descriptions of ministers' roles within the executive and in relation to other branches of government, and details of ministers' responsibilities within their ministries. It covered the "dos and don'ts of being a minister," said Mpare. The Cabinet Secretariat further addressed the issue of Cabinet document security. In the past, leaks of confidential Cabinet documents to the media had caused embarrassment to the government. The reform team took a number of steps to limit leaks, including placing unique identification numbers or marks on documents that would enable the Secretariat to trace documents back to their sources. Although the team also developed the manual at an early stage of the reforms, Cabinet approval of the document did not come until 2007, one year before the CIDA-funded project was scheduled to end. In Ghana, without Cabinet approval, the manual was impossible to enforce. The late approval of the manual meant ministries had less time to become accustomed to the new requirements than originally planned. "Slowness in decision making at senior political levels meant it didn't move as quickly," said Cavanagh. "So we had less time to apply these tools than we hoped." Building a policy planning cohort To apply new systems and procedures, the central government had to cultivate the skills of civil servants working in policy units. Part of the challenge involved elevating the status of policy work in Ghana so that policy units could attract the best and the brightest. "In Canada, policy shops are the ones with the smartest people, who vie for those positions," said Naomi Alboim, longtime Canadian civil servant in charge of policy development in a number of departments and professor at Queen's University School of Policy Studies. "Once you're a policy wonk, you want to stay a policy wonk. However, in Ghana, policy analysts were not highly regarded." The governance project's human resources working group developed a strategy to strengthen ministerial policy units. One aim was to distinguish the unique work of policy directorates within the civil service. For the policy units, the working group made recommendations to introduce job titles such as policy analyst and policy adviser to denote their distinct work, combined with professional development tracks within the standard civil service progression that would accommodate greater specialization. "Western models tend to have a specialty of policy wonk of some sort," said Cavanagh. "To do this policy piece, part of awareness raising was what that job meant in Western systems. After having seen the Canadian model, they wanted to get this classification." However, aside from the PCMEU in the presidency, the civil service did not adopt the new classifications because the changing of civil service rules would invite political complications. The core of the strategy was a program to train officials in the practical side of policy development. Some top civil servants received policy training at GIMPA (the Ghana Institute of Management and Public Administration), though sending larger numbers of civil servants to GIMPA was not feasible because of the expense. The governance project leaders considered developing and delivering the course internally but decided they did not have the capacity to do so. The University of Ghana Business School won an open tender to craft and deliver a policy-training course in consultation with the Office of the Head of the Civil Service-the agency in charge of training civil servants-and with Alboim's support. The curriculum and tools developed were the property of the civil service, allowing the training to continue beyond completion of the governance project. Kofi Nti, dean of the business school, worked with Alboim to develop a curriculum that would meet the needs of policy units. "We wanted to give them training that would make them have an applied feel for developing policy and doing policy analysis," he said. They spoke with officials in the presidency and the Office of the Head of the Civil Service, as well as with chief directors to identify gaps. After developing a curriculum and getting approval from the project steering committee, they recruited trainers with a variety of backgrounds, including university professors and consultants. They delivered six workshops for 94 civil servants in 2004 and an additional eight workshops for 144 civil servants in 2007. The course took six days to complete. Teachers guided civil servants through 11 steps of the policy cycle, from the identification of a problem through writing a Cabinet memorandum. "You go through the system: why policy analysis, the policy cycle, how do you craft options, deal with the problem statement," said Management Services Department executive secretary Dwamena, who participated in the course. "Then there were various things that you account for, like gender analysis, financial impact, environmental impact, communication plan, how to convey the message, and identification of areas of risk." The course differed from other policy training programs available to civil servants because of its focus on the cultivation of practical skills. Alboim said the course went beyond "chalk and talk"-mere lecturing. At each stage, students practiced skills. When learning how to present policy options, for example, students drafted sets of policy options to cover hypothetical issues. When learning about gender impact, students conducted sample analyses. "We had exercises for every segment of it. It was very new to some of the instructors and the participants-to learn by doing, actually apply what they were learning," she said. "There was always tension between theory and practice. They were more comfortable with the theory." Instructors pushed students from their comfort zones. One module of the course was devoted to briefing ministers on policy proposals, for which students had to role-play. "Some were quite anxious about doing that kind of role playing," said Alboim, "because it wasn't just doing it but getting public critique that made some people uncomfortable about it." Throughout, teachers referred to actual policies and statistics in Ghana to provide students with realistic experiences. "We tried to use whatever materials and real-life examples we could," Alboim said. One case study involved a donation from an external funder to support education for children with disabilities. In the actual case, Ghana had accepted the funds even though the donor had set inappropriate requirements. [Civil servant] Students struggled with Ghana's actual decision that "didn't acknowledge the real differences that made it difficult to implement it the way the donor expected it to be implemented," Alboim said. In 2008, as the governance project neared completion and the contract with the University of Ghana Business School was set to expire, civil service head Issachar sought to sustain the training program. "The training modules developed were for the civil service and not for the university. We wanted to continue to make use of them," he said. Nti, then retired from the business school, led a training of trainers with the Civil Service Training Center, the in-house civil service training school of the Office of the Head of the Civil Service. Trainers there were a mix of retired and active civil servants alongside full-time instructors. Beginning in 2010, the French embassy sponsored training for six batches of 25 civil servants each. The training center offered the course into 2013, but without external sponsorships, ministries had to pay for the course out of their training budgets, and registration slipped. The course continued to focus on hands-on skills. "You can go study policy at GIMPA," said Issachar. "For the practical training, writing Cabinet memorandums, you can go to the center." OVERCOMING OBSTACLES In early 2008, an election year, CIDA and the government of Ghana extended the project an additional 18 months to address an issue that had caused Ghana problems in the past. During the previous government transition, in 2001, the incoming and outgoing parties had squabbled over ministerial residences, government documents, and equipment. "It was terrible; the tension was too much. The outgoing party was not ready to go, and the incoming party was anxious to come," said Issachar. "In 2001, there were forced evictions from official residences and ad hoc car seizures. The resulting effect was a poisoned relationship and disagreements leading to confrontation instead of cooperation." Appiah-Adu, Issachar, and Mpare led a delegation to Ottawa to study how Canada managed its own transitions. Upon their return, they issued reporting matrices to ministers and chief directors to provide information the next administration might find useful. The reports included details about ongoing projects, plans for midterm and long-term projects, inventories of staff and equipment, overviews of responsibilities, key planning partners, and sample work calendars. The transition matrices for the preparation of handing-over notes were completed by early 2008. "We wanted to prepare ministers and the entire government machinery far in advance of the election, when you don't know if you will win or lose an election," said Issachar. Cavanagh said the process for getting the transition materials together in such a short time frame was "brilliant." Appiah-Adu said chief directors "bought into this big-time" and did an exceedingly good job in filling in their transition matrices, largely because "the Head of the Civil Service and the secretary to Cabinet were on them to do it." However, ministers did not match that performance. "The political guys didn't pull their weight very much-probably due to the fact that at that point in time, their focus was more on campaigning to win a keenly contested election," he said. In Ghana, as opposed to Canada, transition plans needed the force of law to have effect. However, the plans stalled at the Cabinet level, where approval was needed before any action could be taken by Parliament. The chief of staff to the president was the person who received handover notes from all officials when they left their government posts. Issachar said, "Unfortunately, the chief of staff didn't push the process that much . . . Here, as early as July or August, political activities and campaigns begin in earnest, and other engagements are set aside. Though we had the document ready in March, it was not approved or made public." With no plans in force, the postelection transition in 2009 turned into a rerun of the tumultuous transition of eight years earlier. "We had more problems, so that document didn't actually solve the problems," said Issachar. "Our experiences in 2001 and 2009 were not the best; they were terrible. There was intense animosity and disagreements." In 2012, the government finally passed a transition law. The content of the law derived from the work of the transition group in 2008 and further discussion at the Institute of Economic Affairs, a good-governance research organization. The law set up an independent administrator to oversee the handoff of government documents, equipment, residences, and the like. For instance, when an incoming minister requested information about the number of cars in the ministry's possession, as a result of the law the incoming minister would approach the independent administrator rather than his or her predecessor, thereby reducing tensions and sidestepping disagreements. In late 2012, just before the December elections, President John Mahama, who had replaced President John Atta Mills upon Mills's death in July of that year, appointed Issachar to the administrator position the transition law had created. "The act provides for the regulation of political transfer of power-to strengthen transparency and accountability of the assets and properties of the government and related matters. With this act, it will be different next time," Issachar predicted. ASSESSING RESULTS "We did our best; we made a solid contribution toward policy development," said Appiah-Adu, former head of the Policy Coordination, Monitoring and Evaluation Unit in the president's office. "With highlighting of PPMEDs [policy planning, monitoring, and evaluation departments], we let them take their rightful place in the civil service, with an emphasis on policy." The Ghana Central Governance Project created systems that supported the development of well-designed policies, encouraged collaboration between ministries, built the capacity of policy makers, and equipped them with the physical tools to do their jobs appropriately. The project produced outputs including the Cabinet Memorandum Manual, the ministers' handbook, and the training curriculum, all of which contributed beyond the conclusion of the project in 2009. Charlotte Morgan-Asiedu, director of human resources in the Ministry of Communications, who served on the project's working groups while director at the finance ministry, said ministries benefited from the standard Cabinet memorandum format. "It's a good product that streamlined the ad hoc way we used to do it," she said. Bridget Katsriku, chairperson of the Public Services Commission, who also participated in the project's working groups, concurred. "The policy format clarifies issues better than before: consider options, do analysis. By far, it's a better process," she said. Standards for the submission of policy documents were rigorously enforced. "When the rules were set, when Mpare was Cabinet secretary, the rules were clear, and he followed them to a large extent," said Dwamena. "All of the institutional mechanisms that were put in place he made sure were implemented. If the quality of a memo was not good, he sent it back. If there was a committee of chief directors who had to peer review Cabinet memos before they went, those mechanisms were in place." However, the transition in 2009 following presidential elections revealed that many elements of the reforms had not been institutionalized and relied too much on individuals. Approval of the Cabinet Memorandum Manual, delayed until 2007, left insufficient time for the new procedures to become fully embedded in Ghana's policy management system before the end of the president's term. Although the Cabinet manual remained in force after the 2009 transition, Mpare's successor as Cabinet secretary was less vigilant about securing compliance, according to Dwamena. Most officials from the Cabinet Secretariat were replaced by the end of 2009. Additionally, a new policy unit in the president's office, organized into divisions dedicated to policy coordination and to monitoring and evaluation, replaced the presidential policy unit. Most officials from the PCMEU had left by the end of 2009. Communication between the outgoing and incoming groups was difficult because of the turbulent political climate surrounding the transition. The new policy unit did not adopt much of what the PCMEU had produced in terms of templates and reporting. Smile Kwawukume, senior public sector specialist with the World Bank, said the presidential policy unit performed well but had been undone by the transition. "The design and implementation went well. We were excited such a process was developed, but with the departure of political heads, everything disappeared," he said. Dwamena lamented the fact that parts of what the project strove to achieve in the development of a better policy management system were not sustained. "This was an institutional agenda that would have supported any government. It was not a political thing," he said. "This is something that will help. It has no political coloring." REFLECTIONS The Ghana Central Governance Project made significant strides toward building the human capacity, systems, and physical infrastructure of Ghana's policy management system. Cooperation between the Policy Coordination, Monitoring and Evaluation Unit (PCMEU), the Cabinet Secretariat, and the Office of the Head of the Civil Service ensured that those in charge of policy management bought in to the project design. Nana Dwamena, executive secretary of the Management Services Department, an in-house consulting agency for the central government, said communication between those groups was crucial to the success of the project. "When the head of the civil service and the Cabinet secretary are engaged, the central governance will really be sustained. Because the head of civil service trains the staff and because the Cabinet secretary is the recipient of outputs, there's a need for improved collaboration: not just collaboration but a deeper interaction to ensure we're getting these results," he said. Joe Issachar, head of the Office of the Head of the Civil Service in the latter years of the project, said the long-standing relationships he and Frank Mpare, Cabinet secretary from 2005 to 2009, shared with chief directors made it easier to get compliance with the new rules. Specifically, he was able to command the attention of chief directors when calling for discussions among those chief directors on policy proposals about to be sent to the Cabinet Secretariat. "Communication was easy because of the rapport that existed between Mpare and myself and with all the chief directors. There was that camaraderie, and we used that effectively," he said. After the difficulties encountered in the 2009 transition, observers disagreed about how the reforms could have been better sustained. Some focused on staffing patterns, and others said the incentive structure was flawed. In the eyes of many, better engagement or integration with the ministries also helped reduce the impact of the changes. On staffing, Dwamena said the presidential policy unit should have included civil servants rather than officials on short-term contracts. "I stressed that a lot during meetings: let's have more civil servants brought in," he said. "The argument I heard was, they don't have the capacity. I said, 'Bring them in and let them make mistakes.' The design was right, but we should have made it imperative to be manned by civil servants." Christine Amoako-Nuama, head of the policy coordination unit that succeeded the PCMEU, said she tried initially to include civil servants in her unit with sustainability in mind but found that the civil servants posted to work in her unit lacked sufficient capacity, so she opted to bring in consultants from local universities and elsewhere on short-term contracts as well. Others said the incentives did not support a continual flow of top talent. Judy Cavanagh, who had helped manage Canadian support for the project, said the challenge in recruiting civil servants was not in their capacity but in the unattractiveness of working in the presidency. "We had discussions at the coordinating-team level all the time about sustaining PCMEU," she said. "The problem was, civil servants didn't want to come because it was perceived as risking their careers, since the presidency was seen as too political. Nobody wants to come to the center, where you're under a microscope and have no power or authority in managing programs or budgets." Several things challenged the effort to draw the ministries into the new process more effectively. The policy training component of the governance project provided hundreds of civil servants with practical skills that could be applied in their work. However, Bridget Katsriku, chairperson of the Public Services Commission, said the constant rotation of civil servants through ministries and directorates meant that policy analysts in line ministries didn't always get to apply their training for long. Cavanagh expressed similar concerns. "They weren't being drawn upon by their own ministries to apply their learning. . . . They needed to be able to come back and be used in their own ministries," she said. "That was a loss to Ghana, if they don't utilize the knowledge that people are acquiring elsewhere. If nothing else, they will carry the knowledge and skills of policy development forward into whatever job they have in the civil service." Dwamena countered that the idea of training only the policy and research units was flawed from the beginning. He argued that civil servants in all four directorates perform policy-making functions and should have been included in the training. "Policy is not for PPMEDs [policy planning, monitoring, and evaluation departments]. Every directorate does policy-HR [Human resources] is HR policy. . . . The thinking has always been policy is for PPMED, but the concept when we made the four line directorates, aside from administration, was that they'd be policy based." Samuel Somuah, a manager at the Ghana Central Governance Project, said the initiative was essential for Ghana and for other developing countries. An effective policy management system helps decision makers maximize returns on government funds and natural resources. He said, "If there's one project every African country needs, or every developing country needs, it's this project." 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 Before using any materials downloaded from the Innovations for Successful Societies website, users must read and accept the terms on which we make these items available. The terms constitute a legal agreement between any person who seeks to use information available at www.princeton.edu/successfulsocieties and from Princeton University. 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). 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. 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. A suggested citation format is as follows: [Document author if listed], [Document title], Innovations for Successful Societies, Princeton University, accessed at http://www.princeton.edu/successfulsocieties on [date accessed on Web] e. 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. f. 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. g. 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. Innovations for Successful Societies (ISS) is a joint program of Princeton University's Woodrow Wilson School of Public & International Affairs and the Bobst Center for Peace & Justice. The Woodrow Wilson School prepares students for careers in public service and supports scholarly research on policy and governance. The mission of the Bobst Center for Peace & Justice is to advance the cause of peace and justice through mutual understanding and respect for all ethnic traditions and religious faiths, both within countries and across national borders. References 1 "Ghana Country Profile: 2001." The Economist Intelligence Unit. 2001. http://www.eiu.com/FileHandler.ashx?issue_id=1204413120&mode=pdf.