Ghana’s Competitive Clientelism and Space for Long-Term Stable Policies

Given the close election results and the winner-takes-all nature of politics in Ghana, researchers have argued that the two parties are now characterized by a high degree of vulnerability, which in turn provides strong incentives for ruling elites in both parties to find strategies to ensure their political survival. This results in the distribution of state resources to political supporters and short-termism, which weakens the possibility of building a broad political consensus on any national development issues. Using the case of Ghana’s Right to Information Bill, this article will argue that there are conditions under which elite commitment to long-term development could be fostered and sustained in competitive clientelist political settings like Ghana.


Introduction
For social scientists studying the nature of state and politics in Ghana, there is both good news and bad news. The good news is that Ghana has earned the status of one of Africa's most stable democracies, as it has experienced only peaceful transitions of power since military rule ended in 1992 (Cheeseman et al., 2017). However, scholars studying patron-client politics have noted that Ghana's competitive politics show a very high level of clientelism, exacerbated by decreasing margins of electoral victory; this, among other things, has made politicians much more concerned about meeting the short-term goal of satisfying their supporters to ensure their own political survival (Abdulai & Hickey, 2016;Appiah & Abdulai, 2017;Ichino & Nathan, 2013). This is the bad news, because much of what economic development involves (in terms of policies) cannot flourish under Ghana's form of competitive clientelism.
For example, on public sector performance, Daniel Appiah and Abdul-Gafaru Abdulai (2017) argued that there has been a politicization of state institutions through patronage appointments, with adverse effects for state bureaucracy, because "one of the first acts of successive governments has been to dissolve the boards of state-controlled enterprises so as to appoint their own loyalists to those positions" (p. 11). On the effective governance of natural resources, Sam Hickey, Abdul-Gafaru Abdulai, Angelo Izama, and Giles Mohan compared Uganda (dominant party politics) with Ghana (competitive clientelism) and showed that Uganda seems to be governing oil according to national interests, while competitive clientelism and short-termism/rent-seeking have negatively affected Ghana (Hickey et al., 2015). Similarly, Monica Skaten argued that long-term stability is elusive in the oil industry due to the short-term incentives created by political polarization (Skaten, 2018). Using the case of Ghana's Right to Information (RTI) Bill, this article seeks to depart from much of extant literature regarding the impact of competitive clientelism on Ghana's development. Much of current research argues that the increasingly competitive character of elections in Ghana has incentivized politicians to focus more attention on the implementation of policies that have the potential to enhance their political survival in the short term. In contrast, this article argues that there is scope for long-term development policies in Ghana's competitive clientelism.
The article will focus on conditions under which elite commitment to long-term development could be fostered and sustained in competitive clientelist political settings like Ghana. We argue that under Ghana's competitive clientelism, there are two (sometimes opposite) sets of incentives in pursuit of political survival. One of these has been captured in the above literature: the goal of satisfying supporters. However, there is also the goal of satisfying non-core supporters (swing voters), which falls outside the scope of the need to appease supporters through patronage policies and is increasingly relevant to political calculation in Ghana due to the decreasing margins of electoral victory. This creates incentives that are continuous from one party to the next, and therefore opens up the scope for long-term structural policies with net benefits that accrue to a broad crosssection of society. Some of the continuous policies demanded by neutral voters violate the logic of patron-client politics. The RTI is one such structural policy. 1 Commentators have almost unanimously agreed that politicians from neither party in Ghana want the RTI, because it will make clientelism and the process of distributing resources to supporters difficult. However, our research shows that civil society groups (which we conceptualize in competitive electoral terms as representing swing voters), through collective action, have exploited Ghana's highly competitive political environment to realize the RTI. Ordinarily, the RTI goes against the logic of short-term distributional politics, but the fear of losing ground to the opposition (through the loss of swing rather than core voters) led to its gradual adoption. The theoretical input is to imagine a space for the introduction of long-term stable developmental policies in a competitive clientelistic system, even policies that may oppose short-term distributional politics.
For the study, we interviewed several frontline civil society leaders carrying forward the broad strategy used by civil society groups to introduce the RTI to Ghana. They were signed out for their influence. One of them led the civil society in their negotiation with the Ghanaian parliament about the RTI Bill. All of the interviews were carried out with the promise of anonymity. The interviews were conducted between June and July 2018 in Accra, Ghana. They were direct and intensive, focusing on the factors that make the RTI politically crucial for both parties and how they used Ghana's highly competitive political environment to realize the RTI. We also examined "gray" literature, media articles, and secondary literature. The materials helped to establish several connected themes: (a) Both political parties are opposed to the RTI and (b) the heightened competition between them, which is said to lead to short-term policies, is precisely what civil society used to promote and finally install the RTI. The analysis will show that to win swing voters, political elites in Ghana can be made to accept policies that are not in themselves in line with the logic of clientelism and are disliked by ruling elites on both sides of the divide.
In what follows, we start by discussing competitive clientelism and its implications in Ghana, after which we open up space for imagining stable long-term institutions. The second part offers an empirical exploration. It starts with an overview of the RTI from the perspective of civil society and that of political elites and describes the space available for civil society to persuade ruling elites to take its position. The analysis of the process that ended with the passage of the RTI in Ghana focuses on how civil society exploited competitive clientelism for its own ends, as it did when promoting transparency in the oil industry. The conclusion is that under competitive clientelism in Ghana, where swing voters are becoming very important, civil society is gaining increasing power to influence long-term developmental policies.

Competitive Clientelism and the Theoretical Space for Transformative Policies
One of the dominant analytical frameworks used by researchers to explain the failure of African countries to promote long-term economic development is the dominance of patron-client politics (Bayart, 1993;Mkandawire, 2001;Pitcher et al., 2009). However, in recent years, some scholars have argued that some forms of patron-clientelism is compatible with long-term development (D. Booth & Golooba-Mutebi 2012;Kelsall et al., 2010;Khan, 2010). This has led to several attempts to differentiate between different types of clientelism. For example, within the political settlement framework, Mushtaq H. Khan used the horizontal distribution of power (the strength of elite factions outside a ruling coalition) and the vertical distribution of power (the strength of lower-level factions within a ruling coalition) to create a typology of different patron-clientelism some of which are more developmental than others (see Khan, 2010, p. 65). Khan (2010) identified four categories within which further differentiation is possible: potential developmental, vulnerable authoritarian, weak dominant, and competitive clientelism (for more on the categorization, see Behuria, 2020;Khan, 2010Khan, , 2018. We are much more concerned with competitive clientelism here. Competitive clientelism is a distribution whereby a competition exists between elites over privileged access to a limited set of state resources that they can then distribute to their clients (Lust, 2009). In political settlement terms (within Khan's classification), the implication of competitive clientelism is that ruling elites will focus more attention on the implementation of policies that have the potential to enhance their political survival in the short term.
Plentiful research has been produced on the nature of competitive clientelism in Ghana's current republic (Oduro et al., 2014;Whitfield, 2011Whitfield, , 2018 and its implications (see Whitfield, 2018, p. 100). As summarized by Abdul-Gafaru Abdulai (2017), after the return to multi-party elections in 1992, a de facto two-party system emerged, or rather re-emerged, in the form of the National Democratic Congress (NDC) and the New Patriotic Party (NPP). This was a re-emergence because, as Whitfield observed, although the groups that make up the main political parties are fragmented, they are nonetheless arranged around the two main political traditions that have defined multi-party politics in Ghana since the 1950s: Nkrumah's political ideology, which adopts the broad outlook of socialism (NDC), and the Danquah-Busia coalition (NPP) (Whitfield, 2018, p. 100). The traditional strongholds of Nkrumah and Danquah-Busia are, albeit only approximately, retained in the new political parties. Between the return to multi-party democracy in 1992 and the last election in 2016, both political parties fluctuated in government after two terms of 8 years in office, with elections won by ever-smaller margins as illustrated in Table 1. Nevertheless, several observers have pointed out that Ghana's highly democratic multi-party system is underpinned not by democratic forms of politics but by a patronclient form of politics, intensified by decreasing margins of victory (Driscoll, 2020;Whitfield, 2018).
Given the close election results and the winner-takes-all nature of politics in Ghana, the two parties are now characterized by a high degree of vulnerability, which in turn provides strong incentives for ruling elites in both parties to find strategies to ensure their political survival in the short term. This has resulted in the distribution of state resources to political supporters, as each new administration dismantles (or redirects) the distributive policies of the outgoing administration. This has further consequences; for example, Sam Hickey, Abdul-Gafaru Abdulai, Angelo Izama, and Giles Mohan argued that it significantly undermines Ghana's capacity to pursue long-term structural transformation programs (Hickey et al., 2015).
There are several variations of this. At its core is the weakened possibility of building a broad and lasting political consensus on any national development issues (Abdulai, 2017, p. 9), because as E. Gyimah-Boadi H. Kwasi Prempeh pointed out, the national interest has become "fragmented along party lines, with the result that each new administration has followed its own short-to-medium-term development agenda and spending priorities" (Gyimah-Boadi & Prempeh, 2012, p. 102). Therefore, as researchers have noted (see Appiah & Abdulai, 2017, p. 2;Skaten, 2018), one of the first acts of each successive government has been to dissolve the policies and layoff personnel of the previous government to create space to compensate supporters. This process is exacerbated by the increasingly expensive nature of electoral campaigns. Without fully objecting to this argument, we problematize it somewhat to create theoretical space for imagining long-term stable structural policies in particular areas within a system of competitive clientelism. In doing this, we do not necessarily diverge from the political rationale to which the original position adheres; instead, we imagine how rational politics create different incentives according to groups of voters.
We start from the assumption that supporters are not the only set of people who need to be satisfied in a competitive clientelistic system. In a de facto two-party system, there must be at least three sets of voters for competition to take place. The three sets of voters are core voters for Party A, core voters for Party B, and swing voters without allegiance or uninterested in either party. This classification is simplistic, but it is satisfactory for the theoretical purpose here. Satisfying swing voters entails a different set of political incentives from satisfying core supporters. This means that both Party A and Party B have two main incentives: satisfy core voters and satisfy non-core/swing voters (assuming that core voters are fixed).
Satisfying core voters for Party A and satisfying core voters for Party B are discontinuous incentives, which explains the fragmentation along party lines leading to short-termism and redirection of policies after political changeover, which is the focus of the literature on the relationship between competitive clientelism and long-term development policies. However, the goal of satisfying swing voters does not necessarily differ between the parties, resulting in a theoretical space for policy continuity after a change in government. The incentive structure of competitive clientelism is as follows: where S represents non-core or swing voters, leaving space for continuity regardless of the party in power. The more intense the competition between A and B, the more relevant S becomes. The decreasing margins of victory does not only intensify the patron-client form of politics but makes swing voters more relevant.
For example, in her more recent writings, Lindsay Whitfield (2018) called attention to swing voters (neutral) and showed how ruling elites bring together the distributional demands of core (loyal) and swing (floating) voters separately for electoral purposes. Most research on shortterm thinking and patronage appointments has not theoretically considered swing voters and the separate incentives that they generate. Whitfield (2018) defined them as people who are not "solidly committed to one party and whose vote can be influenced by the electoral strategies of individual politicians or the incumbent party's performance in government" (p. 122). Research has established that the percentage of swing voters in Ghana has almost doubled since the early 2000s (Whitfield, 2018). Weghorst and Lindberg's (2013) analysis of swing voters in Ghana showed that most of them are concerned with broader development issues in their community/the country as opposed to the patron-client type of politics that is dominant in party politics in Ghana. Indeed, by their very nature, swing voters, or at least some of them, do not follow the message of a single political party. Therefore, the main theoretical issues are who represents them in the eyes of politicians and how their demands are shaped and articulated. That is, what sort of incentives arise from the attempt to win over swing voters? As political parties cannot represent swing voters by their very nature, the importance of "third sectors" or non-political groups increases as they become more relevant. There are two ways in which swing voters can share and advocate for their concerns. The first is personally through direct engagement with politicians during campaigns and open engagement with duty bearers. This presents the obvious challenge of the accessibility of duty bearers and moreover, the fragmentation of concerns makes impact less likely. The second is through collective action, and this is where non-political groups comes in. Civil society organizations (CSO) are the main non-political groups in Ghana (Botchway, 2018).
There is no generally acceptable definition of CSOs, but they are generally seen as intermediary organizations mediating between the citizens and the state (J. E. Booth & Richard, 1998;Botchway, 2018, p. 18). CSOs refer to a group or entity that is neither in the state or private sector but is engaged in a public activity aimed at advancing the interests or values of its members and/or society at large. CSOs are a wide array of non-governmental and not-for-profit organizations that have visible presence in public life, "express the interests and values of their members and others, based on ethical, cultural, political, scientific, religious or philanthropic considerations" (quoted in Cooper, 2018, p. 6). As a collective, a civil society is typically composed of civic groups and actors, including nongovernmental organizations (NGOs), think-tanks, and activist groups with different ideologies, aims and objectives, membership, strategies, and geographical coverage. Most, or at least many, CSOs represent narrow interests of specific groups and their goals. However, when there is a common interest for which many of these groups mobilize, it not only transcends their differences, but provides a space for the articulation of and advocacy for structural reform proposals that benefit a broader cross-section of the society rather than just the interests of a specific group. For example, when more than a hundred CSOs came together to create a common platform for oil management in Ghana, the group succeeded in influencing government policies in ways that are difficult to imagine under competitive clientelism.
Thus, our emphasis is on civil society as a collective with common interest and not individual groups or organizations. Admittedly, in some situations, for purposes of political expediency or ideological resonance, political parties (most often the opposition), join forces with civil society to advocate for some reform. This in itself is recognition by political actors of the importance of the civil society front in articulating concerns of non-core voters. Given this, an effective campaign by civil society groups on a single issue has electoral implications, and therefore political elites in Ghana can be made to accept policies that are not in themselves in line with the logic of clientelism.
This way, the two main political parties will have to position themselves in relation to the demand of civil society groups to win certain swing voters or avoid losing them to the opposition. We argue that this partly explains how civil society groups have become powerful in Ghana in areas where they have effectively organized as a single voice. Indeed, as competitive clientelism deepens in Ghana, civil society groups are becoming more influential to almost the same extent that swing voters have become more relevant to winning and keeping power.
Finally, the problem arises of swing voters promoting long-term development policies that threaten short-term clientelistic policies at the same time. Here there is a contradiction between the two incentives; that is, swing voters demand policies that would render clientelism difficult or limit ruling elites' ability to distribute state resources to their supporters. Some long-term development policies can co-exist with short-term distributional policies, while others threaten them. Even in the latter case, there is scope for swing voters to use the highly competitive environment to pursue and eventually install such policies (as well as have them implemented). The RTI, for example, necessarily closes some space for clientelistic policies and is thus found equally threatening by the two political groups. However, the heightened competition between the two parties creates space for one party (likely the opposition) to entertain the possibility of implementing the RTI, which in turn leads the ruling party to entertain it. CSOs are the intermediary that played both parties against each other to create space for the RTI. Below we show how civil society exploited competitive clientelism to install the RTI in Ghana, even though neither party wanted it at first. In the concluding section, we explain how all of this provides a more nuanced understanding of competitive clientelism in Ghana.

Civil Society and Long-Term Policies in Ghana-Example of the RTI
Unsurprisingly, the country's military past limited the development of effective civil society groups, which flourished after the return of democracy. However, the discovery of oil and the consequent highly competitive election of 2008 transformed civil society in Ghana. After the discovery of oil, issues of transparency and accountability came to the fore, especially as part of an attempt to avoid the resource curse (King, 2009;Van Gyampo, 2010). This created the motivation for the development of a platform to harmonize the voices and actions of Ghanaian civil society groups in relation to the oil and gas sector. For example, on the question of monitoring the management of the oil industry, about 120 small civil society groups from different spheres (policy and governance think-tanks, research groups, human rights organizations, environmental and community-based organizations, gender-based groups, trade unions, faith-based groups, media, etc.) came together to create the Civil Society Platform for Oil and Gas (Annan & Edu-Afful, 2015; Debrah & Graham, 2015;Rundquist, 2014). This body has become one of the strongest stakeholders in the Ghanaian oil industry, advising and even compelling the government to promote transparency and accountability in its management of oil.
Several scholars have pointed out that it has successfully influenced the government to fashion policies to protect longterm national interests as opposed to the limited interests of ruling elites that is common in competitive clientelism (Annan & Edu-Afful, 2015). One example is the establishment of Ghana's Public Interest and Accountability Committee (PIAC) to ensure transparency and accountability in the management and use of petroleum revenue and investments. PIAC is a supervision committee responsible for monitoring and evaluating government compliance with the law and providing independent assessment and space for debating the management and investment of proceeds from oil and gas production. It is a third-party group statutorily nominated by professional bodies such as labor organizations, civil society institutions, private sector institutions, policy research organizations, traditional rulers, religious leaders, and other nonstate actors. Both parties at first saw PIAC as unnecessary and even a threat to the supremacy of Parliament (see Odijie, 2017). However, civil society persuaded ruling elites to accept it by enlisting the help of media and opposition MPs. As a result, revenue transparency has been one of the main successes of civil society in Ghana's oil sector (Van Gyampo, 2010). Generally, the international community now promotes Ghana as a model of good resource governance (Ofori & Lujala, 2015). PIAC and other transparency and accountability victories won by civil society necessarily avoid clientelism even though oil ought to exacerbate the use of state resources for distributive policies. This is a model of how a collective of civil society groups could wield influence in competitive clientelism where non-core voters are becoming more important in electoral calculation.
The strength of civil society from 2008 onward led to the highlighting of other concerns; this is when the RTI came to the fore in Ghanaian politics, although the Institute of Economic Affairs (IEA) had first drafted the RTI Bill back in 1999 (Interview, June 11, 2018). The IEA lobbied the new NPP administration, headed by John Kufuor, to look at the bill, which it did. The Executive Branch of the government redrafted the bill in 2002 (Government of . The draft Executive RTI Bill was subsequently reviewed in 2003, 2005, and 2007, but was never laid before Parliament. The idea of the RTI received a boost in 2007-2008. Using the oil model, organizations across all regions of the country came together to create the RTI Coalition (whose members currently number around 80 CSOs). Along with the RTI Action Campaign Group, the RTI Coalition is currently one of the most visible interlocutors with the state on the status of the RTI Bill. For civil society, the RTI, once passed, "will promote total transparency and accountability in Ghana through the requirement for political operatives to disclose all transactions to the media and public." In political terms, civil society groups view the RTI "as a way of combating patronclient politics" (Interview, June 11, 2018). In legal terms, they view the RTI as a fundamental human right guaranteed by Ghana's 1992 Constitution. Indeed, Article 21(1)(f) of the Constitution states that "All persons shall have the right to information subject to such qualifications and laws as are necessary in a democratic society" (1992 Constitution of Ghana). As one influential player explained to us, "the Constitution provides the broad strokes necessary to establish open space for the appropriate legislation. The bill, when passed, will therefore give substance to a constitutional provision" (Interview, June 29, 2018).
Numerous observers and commentators have pointed out that elites from both of Ghana's political parties are opposed to the RTI. As Dr. Franklin Oduro of the Ghana Center for Democratic Development stated, "If there's one, or two or three things that the two main political parties [NPP and NDC] align, agree to, then, it is this RTI that they don't want. I think that is what it is" (quoted in Star FM, 2018). This is logical from the perspective of the nature of clientelism in Ghana. Inter-and intra-party competition has created a considerable need for political financing, as securing and remaining in power have become increasingly expensive endeavors in which contributions made by party sponsors have to be recovered through appointments and/or business contracts (Bob-Milliar, 2012;Ichino & Nathan, 2013, p. 9). This process of recouping such investment is corruption in formal language, and this is what civil society wishes to expunge through the RTI, which opens information created and/or held by political elites to public scrutiny (Interview). Therefore, there was no real support for the RTI from either political party at the very beginning (several interviews).
As a side note, it is crucial to observe that some commentators in Ghana have remained constant in their condemnation of the RTI Bill as irrelevant and potentially even a threat to the very nature of politics in Ghana. A prominent example is the NPP Member of Parliament Kobina Tahir Hammond, who represents Adansi-Asokwa in the Ashanti Region (a safe NPP seat). While in opposition in 2015, Kobina Tahir Hammond argued against the RTI (Ghana Web, 2018). In 2018, while his party was in government, he described the RTI as "a recipe for disaster [. . .] a recipe for bad governance," (Ghana Web a, 2018) on the grounds that "Ministers can't operate, governments cannot proceed effectively, governance cannot take place meaningfully with this sort of albatross hanging over our heads" (Ghana Web b, 2018).
But to come back to the conversation, despite the lack of actual support from both political parties, competition between the two parties created a space for civil society to advance the promotion of the RTI and generate support. Civil society became effective in this regard after different CSOs came together to create the RTI Coalition. In terms of how ruling elites position themselves to gain and retain power, civil society represents swing voters. In Ghana, the party not in government, seeking a coalition to win the next election, was always open to joining civil society in its call for the RTI, a move that the ruling party interpreted threatening to its survival. All of our interviewees made the following observation (in different words; we use here the phrasing of one respondent): "the political opposition always pushes for the bill but as soon as they gain power, they become hesitant to pass the bill" (Interview, June 18, 2018). This "pushing for the bill" by the opposition can be seen as political opportunism, whereas hesitation about passing the bill after gaining power more closely resembles political realism. However, to the degree that political opportunism places the opposition in good standing, the ruling party will have to take some action. The crucial point is that once civil society has managed to get the two political parties into this position of competing for attention, it has the potential to achieve its goals even if those goals are antithetical to the logic of clientelism. This is how the RTI Bill was ultimately passed in Ghana, as narrated below.

Politics fights back: Civil society uses competitive clientelism.
On the whole, throughout civil society's campaign to get the RTI Bill passed, the ruling party usually took the bill seriously during election periods and episodes of highly publicized protest that included opposition MPs, but not during political scandals. The RTI legislative bill was laid before Parliament by the NDC government on February 5, 2010, in part in response to a highly publicized civil society march to Parliament on January 27 (Modern Ghana, 2011). In the 2012 election, it became a campaign point for the NDC, as the party that had brought the bill to Parliament. In contrast with the original atavistic bill prepared by the IEA, and the idea conceived by civil society, the bill submitted to Parliament in 2010 was watered down by the NDC. By creating several escape clauses, this watering down made the bill not only potentially political but also clientelistic.
For example, the information exempted from the RTI Bill placed before Parliament included information from the Office of the President and of the Vice-President (as well as the commissions and agencies under it); information relating to the Cabinet; information on the internal workings of agencies, parliamentary privilege and tax; and information relating to third parties. The 2010 bill also gave Cabinet members power over its review and implementation, opening up the potential for conflicts of interest and control by the government. Another problematic aspect of the bill was its coverage of government agencies only, not public bodies (such as political parties or private institutions doing business with the government). In essence, contracts awarded by government agencies, an area stressed by civil society, could in principle be exempted according to the 2010 bill (several interviews). Civil society groups were somewhat divided as to whether to improve or pass the bill. "We do not want the bill to pass in its current form because it doesn't yield much benefit," a senior member of a civil society group said to us (Interview). Another senior member of civil society who was deeply involved in the negotiation argued that the bill "presents a good starting point from which future amendments can be made" (Interview). The first set of amendments won by civil society groups reduced the possibility of the bill's becoming an idiom for clientelism. For example, the initial provision regarding the central role given to the Attorney General under the Ministry of Justice in supervising requests was replaced (MyJoyOnline, 2016b).
From 2011 to 2016, there were constant demonstrations and media campaigns to get the bill passed, but all to no avail. Civil society groups were bent on exploiting the power of protest, and the ruling party usually gestured toward proceeding with the bill in Parliament at times of protest. Sometimes, the ruling party lost its patience with civil society groups, which then turned to the opposition. For example, civil society groups marked the first anniversary of the RTI Bill's arrival in Parliament with an angry demonstration before Parliament demanding to know why the bill had still not been passed into law (Modern Ghana, 2011). In their protest, they petitioned both the Majority Leader and the Minority Leader of Parliament, and the Majority Leader, Cletus Avoka, responded by stating that "nobody can torpedo Parliament to pass a law" (Modern Ghana, 2011). The Minority Leader, Osei Kyei Mensah-Bonsu, was much more receptive to the protesters' claim for the importance of the bill. This divergence, as we have already hinted, explains the subsequent positions taken by the ruling party and the opposition: The opposition showed more interest in the bill before becoming the ruling party.
The divergence in attitude was most clearly exhibited by Joe Osei Owusu, the current first Deputy Speaker of Parliament under the NPP's administration. While in opposition from 2012 to 2016, Joe Osei Owusu was highly committed to the RTI Bill and collaborated with civil society on the matter on several occasions. For example, he chaired some technical meetings on the RTI Bill organized by civil society groups and supported their demand for the bill to be passed. He also became notorious for labeling the former NDC president John Dramani Mahama a hypocrite (Modern . When Mahama said in interview that he was frustrated by the failure of the RTI Bill to progress through Parliament, Owusu responded by stating that "any time the Executive considers something a priority, it gets done, so if it were their priority, it would have been done" (Ghana Web, 2015). When Joe Osei Owusu became a Speaker, after the NPP won the 2016 election, the RTI Coalition thus regarded him as an ally. However, in his first interview on the bill in his new government role, he suggested that the legislation was not critical: "for me, if you ask me, the more critical bills that are still pending are the Property Rights of Spouses and the Interstate Succession bills. These are things that affect the everyday people" (Ghana Web, 2014). He also argued that the Constitution already provided for the RTI, and that persons who felt information was withheld from them should seek a court order if the information they wanted were of public interest . Likewise, when the NDC left government in 2016, it became a staunch promoter of the RTI Bill.
While political positioning, to which politicians are accustomed, may explain the efforts made by the opposition (vs. ruling) party to emphasize its support for the bill, the refusal to pass the bill shows that ruling elites did not truly want it (several interviews). The NDC made some gestures toward passing the bill, which turned out to be tactics for delaying until at least the 2016 election. For example, as well as making statements on the parliamentary floor and in his presidential speech, then President Mahama tabled the RTI Bill before Parliament on November 12, 2013 (Ghana Business News [GBN], 2016). Following parliamentary procedure, the bill was then referred for review to a Select Committee on Constitutional, Legal and Parliamentary Affairs. The committee held two meetings (in May and September 2014) to review the content of the bill and make recommendations for consideration by Parliament. It took the committee a year to hold two meetings with stakeholders in civil society to review the content of the bill and proffer recommendations to Parliament (Ukaigwe, 2017). On December 17, 2014, the Select Committee submitted to Parliament a report containing recommendations for amendments to the bill; and the motion for a second reading of the bill came 6 months after the submission of the committee's report, on June 25, 2015, under pressure from civil society. The second reading was concluded within a month (on July 24, 2015), and following parliamentary procedure, the bill preceded to the consideration stage. However, it was never in fact considered, despite its repeated appearance on the parliamentary agenda during that session. This parliamentary process partly relieved the pressure placed on the NDC by civil society for a while. However, by December 2016, when the government made a real attempt to pass the bill, it was clear to all involved that what had come before had been merely lip service.
In 2016, during a transition in government, the outgoing NDC predicted that the incoming NPP government might end up passing the bill (even if subsequently to delay its implementation), and therefore calculated that it would be better for the NDC to do so. The NDC therefore sought to pass the bill during the last weeks of its parliamentary tenure. When the NDC government was defeated in the December 7, 2016, general elections (with roughly 1 month left in power), the bill became its top priority for passage, leaving the implementation to the next regime. The speed with which this unfolded was remarkable given the earlier delay. Civil society groups attributed this to the NDC's desire to "claim credit for the passage of the bill" (Interview), or perhaps to prevent the incoming NPP from claiming credit. The NPP promised to rescind any passage of the bill or walk out of Parliament (MyJoyOnline, 2016a). The then minority leader, Osei Kyei-Mensah-Bonsu, argued that the speed with which the government attempted to pass the bill, just a few days before the hand-over, contravened the Presidential Transition Act 2012.
Two days before the expiration of the tenure of Parliament, the Speaker of the House, Edward Doe Adjaho, reiterated his preparedness to have the bill passed into law, stating (quite correctly) that "The Presidential Transition Act is not [a] Parliamentary Transition Act" (MyJoyOnline, 2016a).
This 2016 episode can be best explained in terms of the NDC's desire to claim credit and bring certain voters into its fold. Indeed, a senior NDC politician opined that had the party passed the RTI legislature before the election, it might have won. Even more importantly, this episode made the implementation of the bill more likely, as the NPP now had to pass it because the NDC had shown commitment to it (several interviews), despite evidence that it did not want to. Far from preventing the passage of the RTI (which could in principle close the space for clientelism), competitive clientelism between the two political parties led to its creation. The idea, therefore, that a given policy is not compatible with competitive clientelism should be revisited, because competition makes political parties vulnerable to groups that could be seen as representing votes.
It took 14 months of constant media protest for the NPP government to lay the bill before Parliament on March 23, 2018 (Ghana News, 2018). On gaining office, previous supporters of the RTI started to claim that the bill was less crucial after all (as in the case of Joe Osei Owusu above). In March 2018, the bill was referred to the Joint Committee on Constitutional, Legal and Parliamentary Affairs and Communications, whose report appeared in a few weeks; the second reading was held on June 7, 2018. Consideration of the bill began in July 2018 but was subsequently halted by the Majority Leader. Speaking to the press, the Majority Leader in Parliament, Osei Kyei Mensah-Bonsu, disclosed on July 28, 2018, that the bill could not be passed imminently because it contained too many clauses for examination (Mensah-Bonsu, 2018). It is not clear whether this was another delay tactic; the NPP may have been using this as an excuse, given that top NPP officials had adopted the stance that the bill lacked relevance. However, at the beginning of 2019, the NPP government started to take the bill seriously. The Parliament of Ghana finally passed it on March 26, 2019, and Presidential assent was given 2 months later on May 21, 2019. However, the implementation of the RTI has been postponed to December 2020. This has already led the minority group to call for its immediate implementation.

Conclusion: RTI as Potential Idiom of Clientelism?
The above analysis has shown that to win swing voters (represented here by civil society groups), political elites in Ghana can be made to accept policies that are not in themselves in line with the logic of clientelism and are disliked by ruling elites on both sides of the divide. The real issue concerning long-term development in Ghana is how third sectors exploit the heightened competition between the two parties to promote certain goals. The promotion of transparency in the oil sector may be regarded as one such strategy; when oil was discovered, scholars imagined that it would end up as an idiom of competitive clientelism. This is halftrue. The same heightened competition that led the government in Ghana to pursue quick production to use the propaganda of exporting oil for electoral gain (Hickey et al., 2015) also led the government to institute one of the most transparent and accountable oil sectors in the world. The RTI policy is definitely antithetical to the way ruling elites distribute benefits in Ghana, yet civil society has been able to exploit competition to bring it to fruition.
Scholars have generally agreed that in Ghana's competitive clientelistic system, policies have become an idiom of political clientelism due to the politicization of institutions through patronage appointments. Therefore, the institution of policies such as the RTI can be corrupted by the desire of the NPP and NDC to refashion them according to their political interests. Here we must distinguish between the pursuit of short-term goals/institutions and the punctuation of longterm structural policies by political restructuring every 8 years. The literature tends to collapse both possible strands into the pursuit of short-term policies in general, which is only limitedly helpful, because some policies cannot be restructured every 8 years; transparency and accountability provisions are a good example. The pursuit of short-term policies in the form of the selection of economic focus/sectors for promotion is different from altering the management style of a single agreed continued policy. It is possible to imagine "compromised continuity" and hence long-term policies in the latter context, while the former is discontinuous by definition.
The Kosmos and EO Group controversy is an example of compromised continuity (Phillips et al., 2016). The NDC came into power in 2009 and decided to purge the oil sector of the perceived influence of the NPP by investigating the nature and terms of engagement of the firms that had discovered oil in Ghana, because some of the individuals involved had ties to the NPP. A logical effect of such a politicized/depoliticized process is a final balance, whereby institutions become somewhat neutral because an implicit compromise has been reached. While the NDC therefore set out to expunge the influence of the NPP, the NDC did not permeate the oil sector with its financiers/influence, given both the full glare of civil society and the full knowledge that the NPP would reorganize the sector to take out NDC financiers. The outcome was continuity. This space for compromise and continuity in sectors in which civil society has an active interest contravenes the idea of nonstop short-termism and discontinuity propounded in the literature.
The NDC made amendments to the RTI bill that were hostile to the NPP. One such amendment, which the NPP found to be very problematic and subsequently revised when it took over power, was the provision to bring chiefs under scrutiny.
This clause was popular with civil society, as one of our interviewees argued in 2018: In the last bill, we had a provision which I have described as the very first formal and legislative attempt to bring chiefs under public scrutiny. That provision has not found expression in the 2018 bill. In the 2016 bill, it was there to allow us to examine chiefs and regional rulers who are custodians of royalties from gold mining, etc. Take the Ashanti region as an example. A provision was made in the 2016 law to enable subjects and citizens to question their chiefs on how much they received and how they were utilising the monies. This provision does not exist in the current law.
Many regions with the most powerful chiefs claiming rents in Ghana are NPP strongholds such as the Ashanti Region. As some of the chiefs have customary power and are thus electorally useful, the NPP is reluctant to reduce their power. The NDC (and its progenitor) has battled against chiefs in Ghana since the Nkrumah era. Therefore, the possibility exists that the NPP is currently shaping the current bill to protect its supporters. The outcome of such a process is likely to be a compromise, which will set the RTI on the path toward becoming a long-term institution as opposed to one of clientelism and short-termism, wherein the NDC amends the law (to include traditional rulers, for example) and the NPP expunges it on gaining office.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD
Michael Ehis Odijie https://orcid.org/0000-0002-9012-5979 Note 1. Broadly, UNESCO (2018) defines freedom of information (indistinguishable from Right to Information) as the right to access the information created and held by public bodies and sees it as one aspect of the fundamental right of freedom of expression.