|Title||The Political Economy of Elections in Lagos|
|C1 Background and Explanation of Rationale||
The current literature on electoral violence overlooks the fact that individuals in many developing countries are embedded in clientelistic relationships that regulate their participation in politics as well as their access to resources (Scott 1969; Piattoni 2001). These clientelistic relationships — the personalized networks that facilitate the mutually beneficial exchange of material benefits for political support — are known to shape the behavior of both politicians and voters (Kitschelt and Wilkinson 2007). But the role of clientelistic relationships in mediating individual experiences with electoral violence has been neither theoretically developed nor empirically tested. In this context, the proposed research contributes to our understanding of electoral violence in two ways. First, we examine whether individuals who belong to clientelistic-like associations are exposed to electoral violence at different rates or in different forms. Second, we examine how exposure to violence affects individual attitudes and behaviors toward politics.
To pursue such questions, we focus on the experiences of traders in Lagos, the commercial capital of Nigeria that is often affected by violence during local and national election campaigns. We study traders in particular because they are enmeshed in economic and political relationships that allow us to directly assess how clientelism affects their exposure to violence as well as how such exposure influences their subsequent behavior. Traders in Lagos often participate in market associations whose leaders provide clientelistic-like services with varying degrees of success in shielding them from official predation (Grossman 2016). In leveraging the variation in market associations, we seek to assess whether traders in associations with stronger leaders are less likely to be victimized by election violence.
|C2 What are the hypotheses to be tested?||
To examine if clientelistic relationships affect how traders experience electoral violence, we draw on Grossman’s (2016) finding that traders’ interactions with government officials are significantly mediated by their ties to market leaders. The influence of these market leaders extends to the political arena, allowing them to mobilize support among their members for politicians competing for local offices. We thus expect that the type of market association to which traders belong (strong vs. weak or engaged vs. disengaged) will significantly affect whether or how traders experience violence or other forms of electoral malfeasance during elections. We therefore examine the following hypotheses:
H1: Traders under a market association (compared to traders who are not under an association) are more likely to have interactions with vote brokers in their commercial area.
H2: Traders under strong associations (compared to traders under weak associations) are less likely to report experiencing or witnessing election violence or intimidation.
H3: Traders under politically engaged associations (compared to traders under politically dis-engaged associations) are more likely to have interactions with vote brokers in their commercial area.
Additionally, we examine respondents’ political attitudes and behaviors after being presented with a prompt reminding them about the possibility of election violence in Lagos.
H4: Respondents who receive the violence prompt are more likely to fear becoming a victim of violence or political intimidation.
H5: Respondents who receive the violence prompt are more likely to agree that vote monitoring is likely.
H6: Respondents who receive the violence prompt are more likely to justify use of violence.
H7: Respondents who receive the violence prompt are less likely to plan to vote in upcoming elections.
H8: Respondents who receive the violence prompt are less likely to say they intend to vote for the local ruling party.
H9: Respondents who receive the violence prompt are less likely to trust electoral institutions.
For hypotheses H4-H9, we expect heterogeneous treatment effects based on market association membership as well as partisanship. The effect of the violence prompt should be weaker among members of strong or engaged market associations if they feel that their market leaders can insulate them from the violence in the first place. Similarly, members of the local ruling party should be less affected by the violence prompt if they expect to be shielded from violence otherwise directed at opposition supporters or swing voters.
|C3 How will these hypotheses be tested? *||
We will pursue two strategies for the empirical analyses. First, the analyses of H1-H3 will be conducted using linear regression. These analyses will use the self-reported measures of association strength and engagement. All regressions will be estimated without control variables as well as the following controls: income, party affiliation, ethnicity, and gender. We will employ commercial area fixed effects.
Second, the analyses of H4-H9 will compare average responses among respondents randomly assigned to treatment and control conditions using standard two-sided difference-in-means tests with an alpha of 0.05. We will also estimate treatment effects by regressing the outcomes on the treatment condition along with the measures for heterogeneous effects, namely, market association type and party affiliation. Other controls will include income, ethnicity, and gender. We will also employ commercial area fixed effects in these analyses.
|C5 Scale (# of Units)||1,025|
|C6 Was a power analysis conducted prior to data collection?||No|
|C7 Has this research received Insitutional Review Board (IRB) or ethics committee approval?||Yes|
|C8 IRB Number||PRO-FY2018-289|
|C9 Date of IRB Approval||11/22/2017|
|C10 Will the intervention be implemented by the researcher or a third party?||Researchers|
|C11 Did any of the research team receive remuneration from the implementing agency for taking part in this research?||not provided by authors|
|C12 If relevant, is there an advance agreement with the implementation group that all results can be published?||not provided by authors|
|C13 JEL Classification(s)||D72|