Men of the people? Democracy and prebendalism in Nigeria’s Fourth Republic National Assembly

ABSTRACT While Nigeria’s National Assembly has become increasingly assertive since the start of the Fourth Republic, lawmakers continue to be heavily engaged in corruption. Such rent-seeking practices are commonly regarded as antithetical to democratization. In this article, however, I argue that individual access to public rents actually supports legislative independence. Based on an original dataset of 8th National Assembly MPs (2015–2019), I find that membership of lucrative committees is positively associated to the number of bills and motions sponsored. I also show that while legislative initiatives may be self-serving, an important share also appear to strengthen accountability and the rule of law. This result can be explained by the fact that MPs access to rents is not determined by their loyalty towards the president, but rather by their relations with the Assembly leadership and the bureaucracy. Increased monetary power in turn strengthens lawmakers to play a more independent role in politics.


Introduction
While many Sub-Saharan African countries have (re-)introduced multiparty democracy in the 1990s, democratic consolidation has stalled across much of the continent. 1 One of the reasons behind lagging or interrupted democratization is the continued dominance of African presidents on the political process. Indeed, the key democratic institution to provide checks and balances on the executive, the parliament, has often been found too weak to engage actively in law-making and to hold the executive accountable.
Nevertheless, with the re-emergence of parliaments after decades of rubber-stamping in one-party regimes or nonexistence under military rule, scholars have noted substantial differences in power between them. Consequently, they have increasingly turned their attention to studying the factors that strengthen African legislatures. While rich case studies had already been conducted on different parliaments in Sub-Saharan Africa 2 , Barkan's edited volume has arguably been the first to inductively theorize why legislative power differs across the continent. 3 One of the major arguments of the book is that parliaments are better able to perform their core democratic functions of legislation and oversight if they are able to break free from dependence on patronage resources and ties to the executive via, among others, high personal salaries and a reduced focus on constituency service. Legislative power can develop over time if a coalition of change strengthens lawmakers' financial independence. According to Barkan, these coalitions are led by "reformers" who are committed to transforming parliament into an autonomous democratic institution, and "opportunists", who mainly join for higher salaries.
Barkan and colleagues' argument implies that democratic strengthening requires the crowding out of informal patronage politics. More recently, however, scholars have argued that a better understanding of the exact nature of patronage politics is needed to assess its relation to legislative independence. Opalo, for instance, has argued that clientelist relations between legislators and constituents do not necessarily undermine MPs parliamentary activities. 4 If MPs have substantial control over constituency development funds, as is the case in Kenya, they can build their own support base and increase their electoral security. This allows them to provide pushback in the parliamentary arena. Collord has similarly argued that clientelism may strengthen legislative independence if MP control over resources allows them to become patrons in their own right, based on the cases of Kenya and Tanzania. 5 Both Opalo and Collord focus on clientelist politics between lawmakers and their constituents, a form of patronage which has been regarded as compatible with democracy. 6 A different form of patronage politics exists in the form of intra-elite clientelism, however, in which elites are awarded political positions with the understanding that this position allows them to siphon off public money for their own private benefit. This form of patronage is generally regarded as antithetical to democracy. It may also be referred to as "prebendalism", a concept introduced by Joseph (1987) to describe the pervasive norm in Nigeria that public office comes with the opportunity to derive rents ("prebends") for one's personal use.
In this article, I contribute further to improving our understanding of the relationship between legislative power and patronage politics by investigating how prebendalism influences legislative assertiveness in Nigeria. The country forms an important case to explore this relationship because since the return to multiparty democracy in 1999, Nigeria's National Assembly has gradually evolved into an important player in the governance process. Private bill sponsorship has steadily increased over time and private bills have come to form about half of the bills passed by the Assembly. 7 Lawmakers have used their power to overrule the presidential veto, and the Senate famously rejected President Olusegun Obasanjo's third term bid in 2006. At the same time, however, prebendalism has remained a highly relevant concept to describe Nigerian politics, including in the Fourth Republic National Assembly. In parliament, this is in particular characterized by MPs influencing public procurement processes to gain inflated contracts for themselves and their allies.
The co-presence of legislative power and high degrees of prebendalism requires further investigation. Following Barkan, it could be possible that MPs are on the one hand composed of reformers, who are dedicated to engaging in legislative activity, and on the other hand of opportunists, who are mainly interested in rent-seeking. In this article, however, I draw on MP-level data from the 8th National Assembly (2015-2019) to argue that lawmakers who have powerful patronage positions in parliament are also more likely to be engaged in legislative activities. Based on an original dataset of Nigeria's 8th National Assembly MPs, I show that there is a positive association between MP legislative activity, in the form of bill and motion sponsorship, and their position in the assembly's patronage system. The latter is measured by MPs membership of lucrative committees providing access to resource-rich Ministries, Departments, and Agencies (MDAs). Furthermore, I show that an important share of MPs' legislative activities can be considered to strengthen democracy. While many bills appear to be "porkbarrel" bills or instruments to create new institutions from which rents can be derived, key legislative initiatives have been lauded for strengthening the rule of law.
To understand the positive relation between "rent-seeking" and "rule-making" in the Nigerian parliament 8 , I argue that it is first necessary to understand the workings of neopatrimonialism in parliament. Indeed, while African presidents are commonly seen as the dominant actor in patronage relations and firmly in control of resource flows, Nigerian lawmakers' access to prebends does not appear to be dependent on their ties to the president, but rather their own connections to agents within public administration, and to the Assembly leadership, which rewards loyalty with chairmanship positions of committees tasked with oversight over resource-rich MDAs. Better access to prebends may then provide the security needed for individual lawmakers to take up a more dynamic role in formal parliamentary tasks, including the sponsorship of bills aimed at curtailing the power of the president.
The article proceeds as follows: Section 2 provides background to the case of Nigeria's Fourth Republic National Assembly. Section 3 discusses the nature of prebendal politics in the parliament. Section 4 makes use of an original dataset of MP behaviour in Nigeria's 8th NASS to quantitatively investigate the relationship between the prebendal value of lawmakers' positions with their activity in parliament in the form of bill and motion sponsorship. Section 5 evaluates bills passed by the Assembly and their contribution to strengthening the rule of law. Section 6 concludes and reflects further on the theoretical implications of the findings.

Nigeria's Fourth Republic National Assembly
Nigeria's Fourth Republic has been in place since 1999 and came about after a long period of predominantly military rule. Indeed, the first elected National Assembly (NASS) which was in place from 1999 to 2003 is also called the 4th National Assembly as it has only been the fourth running parliament since independence. Furthermore, the functioning of earlier parliaments had not been exemplary. As with the First Republic in general, the First Republic federal parliament has been regarded as dysfunctional. 9 The parliament sat no more than 54 days in any year and rarely initiated legislation. 10 There were only three working committees, largely for administrative purposes. The Parliament was kept in check by the ruling coalition and rarely performed its legislative and oversight roles effectively. 11 Its functioning was also hampered by defections from the opposition to the ruling party. 12 The Second Republic National Assembly appeared to perform even worse, with members preoccupied with monetary benefits and the role of parliament reduced to "rubber-stamping". 13 Executive dominance was high, with only one (trivial) memberbill passed in the Second Republic. 14 There were no legislative policy agendas nor clear oversight efforts as legislators seemed primarily concerned with scoring generous salaries and allowances emanating from the oil economy. 15 There were also widespread allegations that they received bribes to ensure the quick and easy passage of executive bills. 16 While a Third Republic parliament was installed, it never became functional before the annulment of the presidential elections and the return to military rule.
The Fourth Republic assembly is a bicameral parliament with a Senate or Upper Chamber and House or Lower Chamber. Every four years, and together with the presidential elections, 109 Senators and 360 Representatives are elected from single member districts. Senatorial seats ensure equality among states as 3 Senators are elected from every state, with 36 states in total, while 1 Senator is elected to represent the Federal Capital Territory (FCT) of Abuja. House members are elected based on the population in each state. Both chambers have to agree on any bill before it can be sent to the President for his assent and both scrutinize the budget. The Senate has the additional power to approve the president's nominees for political appointments.
While Nigeria's prior experience with democratic rule and the parliamentary process has been weak, the Fourth Republic National Assemblies can be considered relatively powerful. 17 Parliament has the power to overrule the President with a 2/3 majority if he refuses to assent to a bill. This has happened in the case of the Niger Delta Development Commission Bill (2000), for instance. 18 Parliament can also not be dismissed by the President. The Senate President and his Deputy, as well as the Speaker and the Deputy Speaker are elected by the members themselves. Moreover, the chambers set up their own standing orders. Voting in parliament is done by voice vote. With the constitutional amendment of 2011, the National Assembly has also gained financial autonomy. 19 Nigeria's Fourth Republic is not only powerful on article, but has also constituted a separate force in Nigerian politics. The clearest example is the rejection of President Obasanjo's third term constitutional amendment in 2006. 20 Yet in general, the Assembly has been more active than previous ones. This can be seen from the significant number of private bills sponsored by lawmakers and the fact that private bills have come to constitute the majority of bills passed (see Table 1). In the 8th NASS, over 2000 bills were sponsored, with less than 100 coming from the Executive.
Given its history of military rule and the performance of previous assemblies, the relative strength of Nigeria's Fourth Republic NASS is particularly intriguing. What makes Nigeria an even more unlikely case for legislative power, however, is that this strength cannot be explained by low levels of political corruption in parliament, as discussed in the next section.

Prebendalism in the National Assembly
Nigeria's federal lawmakers have a reputation for rent-seeking behaviour. They have been considered among the highest paid parliamentarians in the world, for instance. 21 According to the 2017 National Assembly budget almost 5 billion naira was dedicated to salary costs for the House of Representatives, while 2 billion naira was set aside for the Senate. The total budget set aside for the House and Senate was around 50 and 30 billion naira, respectively. Substantial additional perks of the office come in the form of allowances to cover a range of costs including housing and fuel, as well as allocations for the purchase of an official vehicle. The amount of allowances received differs across rank, with the Speaker, Deputy Senate Speaker, Senate President, and Deputy Senate President, receiving substantially more. According to 8th NASS Senator Shehu Shani, Senators received on average 13.5 million naira each month for covering their offices' running costs. 22 While expenses normally have to be proven before costs are reimbursed, transparency is known to be low, allowing lawmakers to consider additional allowances as their own salary. 23 Salaries and allowances have tended to increase over time. An important hike occurred in 2008 when the total assembly budget rose from 65 to 105 billion naira. 24 By 2010, this budget stood at 150 billion. In 2015 and 2016 the budget was reduced to around 120 billion. It stood at 125 billion in the 2017 budget. The salaries and allowances of the 8th NASS have hence already been reduced somewhat. A positive evolution is also that the 8th NASS has published the budget on its website and increased transparency.
Substantial salaries and allowances generally increase legislative independence. 25 The latter is also strengthened by the substantial control that lawmakers have over the use of constituency development funds. 26 The budgetary amount allocated to such projects amounts to 100 billion naira a year. 27 The projects are included in the government budget and executed by the MDAs, but the lawmakers themselves decide on which projects to implement. Typical examples are infrastructure works, skills training activities, and the distribution of goods (e.g. school books, sewing machines, motorcycles). Yet even with these features present, Nigerian lawmakers still engage substantially in prebendal practices.
While lawmakers can accrue salaries and allowances during their four years in office, important additional benefits can be gathered through legislators' access to public revenues via the MDAs. 28 Indeed, lawmakers' oversight power over Nigeria's numerous federal MDAs is commonly used to extract additional revenues from the government. This takes the form of lawmakers cooperating withor bribing and extortingpublic servants to inflate the budget and be awarded with public contracts. For instance, current Minister and 8th NASS Senator Godswill Akpabio recently claimed that most contracts from the Niger Delta Development Commission (NDDC) were awarded to National Assembly members. 29 The actual implementation of these contracts is, in turn, poorly monitored. Control over contracts is often easier for constituency development projects 30 , with reports of corruption regularly reaching the media. 31 The allocation of civil servant job slots 32 to members of the Assembly also supports the practice of prebendalism: "they are able to fill the agencies with their cronies, get their projects into the budget, and get money for things that are not done" (Senate Clerk, September 2018). 33 Access to government revenue is mainly organized through the National Assembly's committee system. 34 Coveted positions include the chairman-and vice-chairmanships of the standing committees. The chairman, together with hisrarely her vice, regulates the workings of the committee and has most direct access to the MDAs. While (vice-)chairmen do not get higher wages or allowances, this access creates opportunities for informal rent-seeking.
Yet not all committees give equal access to public money, and (vice-)chairmanship and membership of some committees is more highly rated than others. At the start of the 8th NASS, a panel ranked all committees as A, B, and C to support the distribution of committee positions. 35 These committees are sometimes also called "juicy", "dry land" or "Siberia" respectively. 36 The "juicy" or "Grade A" committees are considered the most lucrative. These committees have oversight over MDAs with high budgets such as Health and Education or MDAs that generate substantial revenue for the government. The committee on FCT is also considered lucrative because of access to land entitlements. 37 Table 2 provides a list of the committees considered lucrative in the 8th NASS. 38 While it can be argued that access to these MDAs provides lawmakers with the opportunity to better serve their constituents, most see it as an opportunity to enrich themselves, as admitted by one lawmaker. 39 Given that certain positions in the Assembly are more conducive to rent-seeking than others, decisions on who gets what are crucial. These decisions are taken by the Assembly leadership, which includes actors of the majority and opposition. In the 8th Assembly this concerned the incumbent All Progressives' Congress (APC) and the Peoples' Democratic Party (PDP). The leadership is formed by the Speaker and Senate President, their deputies, the majority and minority leaders and their deputies, and the majority and minority chief whips and their deputies. Especially the positions of Speaker and President are considered powerful and the committee system is directly related to how they safeguard their positions. 40 One notable characteristic of the committee system is its expansion over time. While, the 4th Assembly (1999)(2000)(2001)(2002)(2003) started out with more than 40 committees in the House and Senate 41 , Speaker Na'Abba increased the committees of the House from 50 to 71. Na'Abba was famously adversarial towards president Obasanjo and mainly increased the number of committees to keep a sufficient amount of members on his side and prevent his ousting. 42 6th Assembly Speaker Bankole committed himself to a reduction and rationalization of the system, but 12 new committees were created nonetheless. 43 In the 7th Assembly, there were 56 standing committees in the Senate and 90 in the House. 44 The 8th Assembly counted 69 and 96, respectively. 45 The 9th Assembly further increased the House committees to 105. 46 Especially lucrative committees such as those related to the oil and gas sector, or education and health, have been split up over time.
The gradual expansion of the committee system allows the NASS leadership to retain sufficient support of the members. 47 Indeed, this objective motivates committee membership decisions rather than that key positions are awarded because of proven commitment to democratic norms or prior engagement in parliamentary work. In the 8th NASS, both President of the Senate, Bukola Saraki, and Speaker of the House, Yakubu Dogara, won the leadership against the candidates preferred by the APC. They mainly achieved this through promises made with regard to the distribution of committee positions. 48 As they had to rely on support from the opposition, many PDP members came to chair lucrative committees, for instance. Both politicians were members of the PDP before switching to the APC in the run-up to the 2015 elections. In July 2018, both returned to the PDP together with other Senators and Representatives.
Given lawmakers' high salaries and allowances, as well as the well-known prebendal practices within the National Assembly, it comes as no surprise that citizens have little trust in the federal parliament. This can be witnessed from consecutive Afrobarometer surveys. 49 Nevertheless, as patronage opportunities appear relatively independent from ties to the presidentan opposition member chairing a lucrative committee still has access to MDA contracts, for instancelawmakers could use their relative independence to play a stronger role in the country's governance processes. In the following section, I investigate how this power is related to lawmakers' formal parliamentary activities.

Prebendalism and parliamentary activity in the 8th National Assembly
The following analyses draw on an original dataset of MPs in Nigeria's 8th National Assembly. The dataset, auxiliary datasets documenting all bills and motions, and syntax are available as an online appendix (see supplemental data).

Independent variable: prebendal value of lawmakers' position in the committee system
To investigate the extent to which the prebendal value of lawmakers' position in the committee system affects the extent to which they engage in formal parliamentary activities, this value needs to be quantified. To do this I have collected information on committee chairmanship, vice-chairmanship, and membership for each lawmaker. This data is covered in the parliament's 1st session Votes & Proceedings books for both Senate and House. 50 Changes over time due to leadership decisions or to lawmakers passing away, losing court cases or bye-elections could only be taken into account when this was explicitly noted in the Votes and Proceedings or in news articles. This leads to some missing values occurring. Yet in general, interviews indicate that committee changes during the legislature do not occur often.
A point system was devised to quantify the value of each lawmaker's position in the committee system. 51 Specific value is attached to (vice-)chairmanship and membership of lucrative committees, as identified in Table 2. The following points were awarded and summed: 4 for a chairmanship, 5 for a lucrative chairmanship; 2 for a vice-chairmanship, 3 for a lucrative vice-chairmanship; 0 for a committee membership, 1 for a lucrative committee membership. 52 If lawmakers saw their positions change during tenure, quarter points were awarded based on whether the lawmaker was in the position for at least the majority of a session.
The final points per lawmaker correspond to qualitative insights on the workings of the 8th National Assembly. Former Senate President David Mark has a score of 3, while it is indeed widely known he has (deliberately) not played a significant role in parliament. Herman Hembe, who was known as an ally of Speaker Dogara before he had to be replaced due to a court ruling, has a score of 7. 53 Most representatives of Kwara state, allies of Senate President Saraki who switched to the PDP with him, have high scores. Finally, the point average is highest for the PDP (M = 5.64; SD = 2.21) and slightly lower for the APC (M = 5.06; SD = 2.33), corresponding to the argument that the Speaker and Senate President gave good positions to the PDP to gain the leadership of the Assembly. Lawmakers who belong to smaller parties generally have weaker positions (M = 4.33; SD = 3.04).
Finally, it is important to note that in the House, members of the leadership are not members of standing committees, which results in missing values for their prebendal position. Members of the Senate leadership are, however, also committee members except for the Senate President and his Deputy. The dataset includes a Senate indicator to address possible additional differences between both chambers.

Dependent variable: formal parliamentary activity
The dependent variable of interest is the extent to which lawmakers engage with their formal role and duties in parliament. To measure MPs' formal parliamentary activity, I focus on their bill sponsorship. While many private bills do not make it into law, bill sponsorship has become an established instrument to measure lawmakers' behaviour in the institution of parliament. 54 The focus on bill sponsorship is to some extent also driven by data availability as other measures, including attendance rates and speaking times, are not available. Nevertheless, bill sponsorship can also be regarded as a proxy for overall activity when we take into account that to sponsor a bill an MP at least has to be present and working in the Assembly, something not all of them do. The data on bills and their sponsors are derived from the Policy and Legal Advocacy Centre (PLAC) Bills tracking website. 55 National Assembly MPs also extensively engage in motion sponsorship. As a second measure of MPs' formal parliamentary activity, I hence also investigate motion sponsorship. Data on motion sponsorship is derived from the Votes and Proceedings and the Order Papers. No large date gaps occur, indicating that motion data is relatively complete. I only take substantive motions into account which address issues such as bad roads, environmental damage, etc., and exclude motions announcing the demise of important societal figures as well as procedural motions.
For each lawmaker I note how many private bills and motions (s)he sponsored in total. These measures include co-sponsorships, but variables restricted to single bill and motion sponsorship provide similar results. Table 3 shows that there is substantial variation in lawmakers' individual bill and motion sponsorship with many not sponsoring a single bill or motion, while others sponsor a large number. The correlation between bill and motion sponsorship is ρ = 0.390 (p-value < 0,001).

Control variables
I also include control variables such as age, gender, and regional zone. Additional variables indicate whether the MP has a Master degree or higher, whether the MP has already served a term in his/her respective chamber, and whether the member is part of the opposition. The opposition indicator is based on the platform the lawmaker won the election on, regardless of later switches. Another indicator registers whether a member has served a full term or had a shorter tenure due to his/ her passing away or the loss/late gain of a seat as a result of court cases and byeelections. 56 Control variables were gathered from the INEC candidate lists and MP biographies based on web searches. 57 As previous research in Sub-Saharan African settings has found that vote margin can be negatively related to parliamentary activity 58 , I include the 2015 vote margin as a control variable. This data is based on INEC 2015 election or repeat/bye-election results. As mentioned, I also

Empirical results
The results reported in Table 4 make use of linear regression analyses. 60 The prebendal value of a lawmaker is positively associated with his/her parliamentary activity. Lawmakers that take up a high-ranking position in the Assembly's informal rent-seeking system sponsor more bills and motions in general. Except for shorter tenure which has the expected negative effect, most control variables are not robustly associated with parliamentary performance. For instance, vote margin has the expected negative sign for bill sponsorship, but it does not reach statistical significance. For motion sponsorship the coefficient is positive and insignificant. Education level has a positive effect on bill sponsorship, but not motion sponsorship. The same applies to opposition membership. Legislative experience appears to have a negative association with motion sponsorship. This could be due to the fact that Assembly motions are rarely acted upon by the executive. Experienced lawmakers may hence choose not to invest in motion sponsorship. Finally, Senators clearly sponsor more motions than House members (note also the high R²), yet most of these are co-sponsored. The high degree of motion co-sponsorship in the Senate may be due to a norm of national unity and consensus within the Upper Chamber. The differential effects of the control variables in both models support the idea that bill and motion sponsorship can be regarded as different forms of parliamentary activity. Hence, the finding that lawmakers with more valuable rent-seeking positions are more likely to be engaged in both types of activities supports the association of informal prebendal politics with formal parliamentary activity. Note: OLS regression coefficients with standard errors between brackets. * p < .05, ** p < .01, *** p < .001.

The democratic quality of Nigerian MPs' formal parliamentary activities
While the previous section revealed a positive relation between lawmakers' position in parliament's informal rent-seeking system and MPs' formal parliamentary activities, this section traces whether formal activities can indeed be associated with strengthening democracy. Indeed, private bill sponsorship may easily focus on "pork-barrel" bills and strengthen clientelism. 61 In Nigeria, lawmakers are also known to promote bills in order to create new public agencies which allow for additional rent-seeking opportunities. 62 Or, bill and motion sponsorship simply serves the function of gaining media attention without actual commitment to the legislative process: "Many legislators propose bills, but only for first reading, when media picks up on it. You hardly see them pushing their bill to second reading." (NASS Legal advisor, August 2018). According to data from the PLAC bills tracking data, 1492 private bills were introduced in the 8th NASS House, while 331 of these were passed. 63 695 private bills were introduced in the Senate, with only 200 passing. These numbers indeed indicate that from the large number of private bills introduced only a limited proportion is passed. Moreover, not all these bills were turned into law as the President did not assent to a substantial number of 8th NASS bills. 64 Quite a number of the bills introduced in the Assembly also appeared geared towards the creation of new regulatory agencies, as well as medical centres, universities, and polytechnics in MPs' own constituency. Nevertheless, an inspection of the bill titles indicates that only 20% of House bills and 30% of Senate bills are officially categorized as "establishment bills". An additional 10% of House bills and 5% of Senate bills could have the purpose of creating new rent opportunities based on their title. The majority of bills hence does not appear to have a clear rent-seeking function. 65 Establishment bills are, however, significantly more likely to pass in both chambers, perhaps because more lawmakers can agree to "horse-trading" politics than to other types of bills.
These general statistics appear to suggest that not all bill sponsorships serve prebendalism or publicity-seeking. Furthermore, when we repeat the previous analysis of bill sponsorship and restrict this measure to bills that do not have an establishment purpose and bills that have actually been passed, we still see a positive relationship with lawmakers' prebendal position. The effects are somewhat weaker, however. Motions as well appear predominantly geared towards addressing problems in MPs constituencies (e.g. bad roads, insecurity) or to bring potential corruption cases to light. Hence, they do not appear to be directly associated with narrow private interests ( Table 5).
The categorization of bills into establishment and non-establishment bills may say something about their instrumental purposes, but not their democratic quality. Some institutions may strengthen democracy and the protection of human rights. This includes perhaps the expansion of the National Commission of Refugees to cover IDPs and migrants. 66 Others, however, may undermine it. A remarkable example of the latter is the Non-Governmental Organization Regulatory Commission or "NGO bill" which was introduced as a private bill by the Deputy Majority Leader in the 8th NASS House, but failed to pass. The bill was not only criticized for duplicating the work of existing institutions and creating legal confusion, but also for giving the Commission the prerogative to refuse NGO registration if their work was deemed to be counter to the national interest. 67 This rule could have easily been used to undermine critical pro-democracy movements. Some bills have been acclaimed by civil society organizations for strengthening democracy, however. 68 The most important example is the 2018 Electoral Amendment Bill, which, among other provisions, sought to make electronic voting possible and reduce the potential for fraud, to change the electoral timetable and separate National Assembly from presidential elections to increase legislative independence, and to improve the transparency of internal party primaries. 69 Another lauded initiative was the "Not-Too-Young-To-Run" Bill which reduced the minimum age to hold an elected office. The Presidential Transition Bill sought to provide a legal framework for turnover to ensure stability. The 8th Assembly also passed the Petroleum Industry Governance Bill, which was a long time in the making 70 , to increase transparency and accountability in the corruption-struck petroleum sector. Finally, the 8th Assembly Senate passed the Sexual Harassment In Tertiary Educational Institutions Prohibition Bill to redress abuse in the sector. Among these bills, only the Not-Too-Young-To-Run Bill became law.
Given that these initiatives have been regarded as strengthening democracy by external actors, including Nigerian civil society and opposition members, it is interesting to take a closer look at the lawmakers behind these efforts in the Assembly. Table 6 shows the private bills that were behind each of these initiatives. I also investigate the position of the sponsors behind the bills according to the prebendal value measure used in the previous section. The findings indicate that a stark majority of the lawmakers involved also held lucrative positions in the NASS committee system. This supports the view that lawmakers successfully engaging in prebendalism are also more likely to actively use their formal parliamentary role to support democracy and the rule of law.

Conclusion
The National Assembly has become an active player in Nigerian politics. Existing arguments could relate this relative strength to lawmakers' high formal salaries and control over constituency funds. 71 In this article, however, I have argued that prebendalism in parliament, or the private use of public revenue, also explains this legislative power. Based on individual-level analyses, I demonstrated that MPs' access to lucrative committees in parliament is positively related to their engagement with formal parliamentary tasks including bill and motion sponsorship. I also showed that an important share of these bills and motions appeared to benefit constituents and the broader citizenry as well as strengthen democratic accountability. Hence, the co-presence of legislative assertiveness and prebendalism in the Nigerian parliament cannot be explained by a distinction between "good" lawmakers, who adhere to democratic norms, and "bad" ones, who are mainly interested in accruing rents. Many MPs rather appear to embody both sides.
MPs' push for more voice and accountability through legislative means is not necessarily surprising. Indeed, reducing the power of the presidency can often also advance the personal careers of lawmakers. The proposed amendment to change the electoral timeline is a good example of how reducing executive dominance over the electoral process is often directly related to increasing individual MPs independency. Supporting popular initiatives to improve the rights of women or young people, is also a way MPs can garner support from broad groups in society. MPs legislative initiatives may hence serve their own interests, but at the same time promote democratization.
The question then is how prebendalism allows MPs to put constraints on presidential power. Opalo 72 and Collord 73 have both already argued that legislative independence is related to the means at lawmakers' disposal. Where Barkan first stressed that high salaries and a reduction of clientelist constituency relations strengthened parliament, both scholars argue that clientelist politics can actually strengthen MPs independence as long as MPs have substantial control over patronage resources. This control allows MPs to become patrons in their own right and build personal constituency support. Increased electoral safety in turn leads parliament to be more assertive towards the chief executive. Prebendalism could increase legislative independence in the same way. While prebendalism often serves private interests and those of elite allies, some share of rents may be channelled towards constituents to garner personal support. In addition, however, prebendalism may also allow MPs to safeguard their position within their parties. Turnover rates are high for the National Assembly. Many MPs lose their seat in party primaries and this may in particular apply to assertive lawmakers. 74 Access to personal finances to gather sufficient internal party support and pay off party delegates during primary elections may also explain the link between prebendalism and legislative independence in Nigeria. 75 The proposed mechanisms here of course require further research. First, a more comprehensive content analysis of parliamentary bills may be needed to further improve our understanding of the types of initiatives Nigerian (and African) MPs propose and to what extent they strengthen democracy. This would require an extensive analysis of bill texts as well as an analysis of the existing legal frameworks bills address. Second, further research is required to assess the effect of both parliamentary activity and prebendalism on individual MPs' political careers. Finally, this article has shown that Nigerian MPs have access to public revenue outside of the president's control, which appears atypical for African countries. The reasons for this need to be delved into more. While Opalo argues that legislators' access to resources is determined by how parliaments developed under one-party rule 76 , his theory cannot apply to Nigeria where military rule did away with the institution of parliament entirely. The answer may be sought at the other side of the patronage relationship, however, namely the MDAs. Joseph (1987) has suggested, for example, that military rule in Nigeria has strengthened the bureaucracy in the country to an important extent.
Finally, it remains important to nuance the relationship between prebendalism and democracy as such. While this article has argued that access to rents in parliament is associated with legislative strength, the pervasive norm of prebendalism in Nigeria may still, on average, undermine democracy. For instance, only few of the proposals made by lawmakers to restrain executive dominance actually make it into law. Indeed, the electoral amendment bill and the petroleum industry governance bill have still not come into force. Yet lawmakers' prebendal practices undermine citizens' political trust and lead to a loss of legitimacy of parliament as an institution. Over time, these negative effects may still prove more important for the future of democracy in Nigeria than improved legislative independence. 42. Uganwa,Nigeria Fourth Republic National,182. 43. Lewis,"Rules and Rents," 200. 44. Obadan,"The National Assembly and," 159. 45. Information at 20 June 2017. The number of committees can be altered during tenure. 46. Avitogo, "House Committees: APC Gets". 47. Fashagba, "Legislative Oversight under the Nigerian," 455 ; Nwozor and Olanrewaju, "Oiling the Legislature," 177-8. 48. Chioma, "House of Reps: Breakdown". 49. www.afrobarometer.org. 50. The Speaker also installed a system of "co-opted members" whereby certain committees could expand their membership beyond the stipulated list. These could not be included here. However, members who are given official positions could be considered more powerful than those who were only co-opted later, which is why I do not expect this to distort findings substantially. 51. The idea for a point system was inspired by a conversation with a lawmaker who argued that the position of a Speaker is worth a 7, while a Senate President is worth an 8, the same as a governor. 52. Chairmen and vice-chairman of lucrative committees are not awarded an additional point for membership of that committee. 53. e.g. Shibayan, "Finally, Dogara Swears in". 54. see Schiller, "Senators as Political Entrepreneurs". 55. http://placbillstrack.org/. 56. MPs who had a shorter tenure do not get their prebendal value points reduced based on the sessions they were active in as the shorter tenure indicator already compensates for this in measuring their performance. 57. The following websites were used in particular: shineyoureye.org, kyg.nigeriagovernance.org (now no longer operational), nigerianbiography.com, and nass.gov.ng (now updated to 9th NASS). Education data was taken as given, regardless of the occasional certificate scandals uncovered by Nigerian media. See e.g. Oketola et al., "Dino Melaye's Certificates Scandal". 58. Wegmann and Evequoz, "Legislative Functions in Newly". 59. While the Assembly counts 360 House and 110 Senate members, the dataset includes all members who have won a seat during 2015-2019, even if this seat was lost at a later time. 60. Linear regression statistics are preferred over negative binomial analyses because of the straightforward interpretation of the coefficients and R² goodness-of-fit statistics. Negative binomial regressions provide highly similar results, however, with generally higher significance levels for the coefficients (see online appendix, see supplemental data). 61. See e.g. Crisp et al., "Vote-seeking Incentives and Legislative". 62. Suberu, "Legislative Development and Decadence," 225. 63. Bills that were consolidated with other bills that passed were coded as "passed". Note that the number of passed bills is calculated per Chamber and private sponsor. 64. Akinwale, "8th Assembly Introduced". 65. For the categorization, bill titles were consulted and the bill texts where these were available on the PLAC website. 66. Jimoh, "The 8th National Assembly." 67. PLAC, "Factsheet on the Non-Governmental". 68. Akinwale, "2019: CSOs Urge N'Assembly"; Commonwealth Observer Group. "Nigeria General Elections"; Iroanusi, "Nigeria: Sign Electoral Amendment"; Jimoh, "The 8th National Assembly." 69. PLAC, "Factsheet on the Electoral". 70. Suberu, "Legislative Development and Decadence," 229. 71. Barkan, Legislative Power; Collord, "The Legislature: Institutional Strengthening"; Opalo, Legislative Development in Africa. 72. Opalo, Legislative Development in Africa. 73. Collord, "The Legislature: Institutional Strengthening". 74. Suberu, "Legislative Development and Decadence," 226. 75. This mechanism is partly supported by the quantitative data. Lawmakers who attempted to gain a higher elected position had on average a higher prebendal value (6.21) than members who did not participate in elections (4.83), and members that went for re-election in the same position (5.25). The effect of prebendal value on these choices in more extended models is not robust. Yet, this may be due to relatively small effect sizes, measurement error, and other confounding factors. 76. Opalo, Legislative Development in Africa.