Why Run for Office in an Authoritarian Regime Ch 2

Aggregate Measures of MP Motivations and Role Orientations We turn first toward the aggregate measure of MP motivation to see whether the general theoretical expectations from Cameroon correspond with the survey results. Figure 2 reports box plots for each of the individual MP motivations, and the individual panels describe different scoring techniques. Panel (a) presents the full range of ranked responses and describes what percentage of the total points (60) was allocated to a specific motivation. Panel (b) addresses the potential for response fatigue and only looks at the distribution of the top two choices, or scores of 5 and 4, across the four motivations (this time allocated from the 36 available points). Finally, Panel (c) presents the distribution of a censored response range of only rankings 2, 3, and 4 (out of a total of 36 points available) in order to detect possible social desirability bias. Considering the full range of responses, Cameroon MPs ostensibly express strong ideological motivations to run for office (mean=30.8%). This includes strong agreement with statements about the value of working with other political parties to achieve legislation or the need to exert stronger oversight of the executive branch. Ideological motivations are followed by material motivations, such as the desire to obtain a good salary or deliver constituency-level resources (mean=25.0%). This is followed by prestige concerns like the desire to make public appearances (mean=23.4%) and career ambitions like making connections with key businesses for future employment (mean=21.0%). Considering just the most strongly expressed preferences, the ideological and material motivations become more strongly pronounced. Notably, in both scoring schemes the range of responses is quite wide, which is further explored in the next section. (a) Full Range of Responses 10 20 30 Material Ratio 40 (b) Top Two Responses 0 20 Career Ratio 40 (c) Censored Responses 60 0 Ideology Ratio 10 20 30 40 50 Prestige Ratio Figure 2: Box Plot of Four Different MP Motivations (Full Range of Responses, Top Two Responses, Censored Responses) There is strong reason to suspect such strong ideological motivations among Cameroon’s MPs. The strength of authoritarianism in Cameroon and the weakness of the legislature raises the possibility that this response is due to social desirability bias. In a censored sample, the score for ideological motivations drop sharply, while the scores for career and prestige motivations rise. However, the scores for material motivations do not actually change that much in the censored sample. This indicates that respondents are not consistently pushing material motivations to the lowest possible ranking, but rather frequently scoring them between 2 and 4. We conclude from this that social desirability bias is likely manifest in the expression of strong ideological motivations rather than weaker material motivations. In other words, the MPs in the survey are more concerned about conveying a sense of mission as a legislator than hiding their material motivations to run for office. Consequently, for the rest of the analysis we consider MPs who are more or less materially oriented rather than those who are more or less ideological.14 Figure 3 reports similar outcomes for our measure of role orientation. Contrary to some expectations regarding the impact of authoritarianism, Cameroonian MPs appear in line with other findings from Africa and exhibit stronger constituency servant role orientations (mean=30.4). Constituency servants express stronger support for statements like an MP must get development funds allocated to their constituency. Constituency servants are followed by partisans (mean=27.0), which would include agreement with statements like an MP must listen to the views of party leaders or an MP must donate finances to the party. The bottom two role orientations are public servant (mean=24.4) and entrepreneur (mean=18.4). Public servants would agree with images of the MP as someone who investigates government wrongdoing and consults experts before passing legislation. An entrepreneur is an MP who thinks that personal achievements and name recognition are important factors in their electoral success. These preferences become slightly more pronounced in the sample that only scores the top two choices. As with the measures of MP motivation there is significant variation within each category of role orientation. The emphasis on constituency rather than public servant role orientations is a bit surprising given MP’s emphasis on ideological motivations. It could be that constituency servant is the socially desirable role orientation that MPs want to present, and that this model of a legislator corresponds with their ideological motivations. It could also be that the framing of questions regarding role orientation is less sensitive than questions regarding ambition and motivation. We assume that a partisan or entrepreneurial role orientation are less socially desirable. However, there is little evidence to support this, and in fact the scores for entrepreneurial role orientations double 14 There is a negative correlation between material motivations and ideological motivations (corr=-0.48). in the censored sample. If respondents were consistently “punishing” the entrepreneurial role orientation with the lowest possible score, its aggregate should drop in the censored sample.15 From this we conclude that we can assess variation in constituency servant orientation with less risk of bias. 10 20 30 Partisan 40 0 20 40 Const. Servant 60 80 0 10 Public Servant 20 30 40 50 Entrep. Figure 4 Box Plot of Four Different MP Role Orientations (Full Range of Responses, Top Two Responses, Censored Responses) Exploring Variation in MP Motivations and Role Orientations While Cameroonian MPs appear to be materially driven and express stronger constituency service role orientations, the above results mask considerable intra-sample variation in responses. The literature on legislators highlights a set of individual-level and institutional-level variables, 15 Consider a hypothetically biased response where the five entrepreneurial statements receive a full range of scores that looks like this – “1, 1, 1, 2, 1” for an aggregate score of 6. If we dropped the low scores the aggregate should drop to 2. which we have hypothesized will have an impact on the extent to which MPs express material and constituency service orientations. In this section, we offer some preliminary insights into these factors and find support for a number of our hypotheses. Given the small size of our sample, a word of caution is needed regarding any statistical findings. Moreover, as noted above we have more confidence in the survey findings regarding material motivations than ideological motivations, which are likely influenced by social desirability bias. For these reasons, we also supplemented this analysis with a closer examination of the subset of MPs who fall along the extremes of some of our measures. Table 2 reports the results from an Ordinary Least Squared (OLS) regression, with each model varying MP motivation and role orientation. As noted in Model 1, opposition MPs exhibit weaker material motivations than ruling party MPs, which confirms one of the major hypotheses of this study and comports with findings from other authoritarian settings (Weghorst 2015). As expected, we also find that less lucrative occupations, which we term as “other,” are significantly more motivated by material incentives as compared to business elites.16 Likewise, MPs selected in primary systems are more materially driven than MPs selected centrally by the party. We find less support for the notion that political experience (time in the legislature), rural constituencies, electoral competitiveness, or district magnitude influences material motivations. Surprisingly, we find that the age when elected influences material motivations in the opposite direction we expected. Younger MPs are actually less materially driven than older candidates, which could be explained by their lower social status at the start of their careers. Without a material basis, younger 16 These occupations include school teachers, taxi drivers, small scale traders, and secretarial/administrative positions. MPs are less certain about their political careers, which could lead them to emphasize different motivations to run for office.17 On the other hand, there is little that correlates with any of the other motivations explored in Models 2-4. As expected, opposition MPs demonstrate higher ideological scores than regime MPs, but the F-score for Model 3 is not statistically significant. Consequently, the finding that MPs from single member districts are more ideologically motivated cannot be accepted at face value. There is some evidence that career motivations are more pronounced among professionals compared to more generic job positions. This might be because legislative service is a career endpoint for many government civil servants and that business elites are already advanced in their careers once they arrive in the legislature. For less lucrative occupations that are contained in the “other” category, career opportunities outside of the legislature are more constrained. By contrast, professionals – which includes doctors, lawyers, and academics – have the most opportunities for career advancement outside of the legislature. Models 5-8 report the OLS results for MP role orientations, which again only unevenly confirm the hypotheses set out earlier in the paper. Contrary to expectations, partisan role orientations are more pronounced among experienced legislators and among MPs with a business occupational background. A potential explanation is that it is not that partisanship declines over time, but that partisanship is in fact a key factor in longer-lasting legislative careers. Likewise, given constraints on business in Cameroon, one way that aspiring business elites demonstrate loyalty to the regime is through partisanship rather than strong constituency affiliation. On the other hand, as expected, opposition MPs are more partisan given their weaker ties to constituency service, although that is not picked up in the constituency servant role orientations presented in 17 In unreported models we interacted opposition affiliation with age, occupation, and selection process. In each model these variables maintained their independent significance, but the interactions were not significant. Model 6. In addition, electoral competition is a key factor that explains variation in role orientations, and it is significant in both Models 5 and 6. The higher the degree of electoral competition, the more constituency-oriented an MP becomes, while the stronger the regime’s electoral grip the more partisan an MP becomes. Other factors do not appear to differentiate degrees of constituency service role orientations, which is a strong orientation among the bulk of Cameroonian MPs. 1 Table 2: OLS Regression of MP Motivations and Role Orientations Age Elected Time in Legislature Government Occupation Business Occupation Professional Occupation Opposition Party Population Density, log (department) Election Result Single Member District Primary Selection Constant 2 3 Observations R-squared Prob>F (1) Material 0.11* (0.05) -0.05 (0.09) -3.17** (1.20) -4.34** (1.39) -3.28** (1.40) -2.88** (1.26) -0.41 Motivations (2) (3) Career Ideology -0.06 -0.05 (0.06) (0.05) 0.12 -0.04 (0.09) (0.09) -0.79 1.28 (1.33) (1.24) 1.54 2.15 (1.47) (1.40) 3.46** -0.50 (1.30) (1.24) -0.18 2.53* (1.09) (1.07) 0.23 0.01 (4) Prestige 0.01 (0.06) 0.00 (0.10) 2.80* (1.37) 1.12 (1.22) 0.58 (1.34) 0.16 (1.25) 0.22 (5) Partisan 0.05 (0.05) 0.27** (0.10) 1.77 (1.13) 3.23* (1.38) 2.07 (1.49) 2.72* (1.24) -0.08 Role Orientations (6) (7) Const. Public 0.06 -0.04 (0.05) (0.05) 0.03 -0.42*** (0.08) (0.10) -1.30 1.02 (1.06) (1.18) -2.09 -1.86 (1.31) (1.40) -1.21 -2.87* (1.29) (1.22) -2.12 1.31 (1.13) (1.25) -0.39 1.01*** (8) Entrep. -0.09 (0.05) 0.14 (0.14) -0.99 (1.39) 0.85 (1.21) 2.16 (1.15) -2.12* (1.05) -0.56* (0.29) -0.05 (0.03) -0.53 (1.05) 2.62* (1.27) 21.21*** (5.00) (0.30) -0.00 (0.03) -1.21 (1.13) -1.93 (1.28) 26.62*** (5.10) (0.28) 0.03 (0.03) 2.50* (0.95) -0.28 (1.00) 29.66*** (4.01) (0.30) 0.00 (0.03) -1.08 (0.87) -0.39 (1.50) 23.72*** (5.04) (0.27) 0.08* (0.04) 2.82* (1.13) -1.55 (1.22) 16.51** (5.04) (0.24) -0.13*** (0.03) 0.57 (0.96) 0.27 (1.17) 35.62*** (5.04) (0.28) 0.01 (0.03) -0.80 (1.14) 4.44** (1.34) 17.69** (5.88) (0.26) 0.03 (0.03) -2.34 (1.21) -3.10* (1.41) 30.97*** (4.82) 70 0.23 ** 70 0.21 * 70 0.16 - 70 0.11 - 70 0.21 * 70 0.23 *** 70 0.34 *** 70 0.24 ** Note: Ordinary Least Squared Regression for survey data. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05. Omitted category for occupation is “Other”. All models include unreported controls for gender and education. Models 7 and 8 present some interesting results regarding the less pronounced role orientations of public servant and entrepreneur. Time in the legislature is correlated with weaker public servant role orientations. This might be due to some process of adaptation and learning that some MPs go through as they realize the limitations of the ability of an MP to engage in public service in Cameroon. As noted previously, MPs elected at a younger age also tend to be less materialistic, which would correspond with this idea that early enthusiasm becomes tempered by legislative realities. Interestingly, MPs from less rural districts and those elected in primaries also express comparatively stronger public service orientations. By contrast, MPs from more rural districts and those elected in primaries are more likely to express stronger entrepreneurial role orientations (emphasizing their personal accomplishments in life). This raises the possibility that popularly elected candidates in more urban environments respond to a more educated electorate that demands stronger public service. However, the rural-urban distinction does not appear to shape motivations or decrease constituency servant role orientations. There are limitations to this smaller-n statistical analysis, and the results also indicate a certain degree of indeterminacy. Therefore, we also look at a subset of exemplar MPs that either indicated strong preferences for or strong aversions to specific motivations and role orientations. We choose to focus on variation in material motivations and in partisan and public servant role orientations. This is because ideological motivations are so strongly pronounced in our sample and likely reflect bias, while very few respondents indicated primary support for career or prestige motivations (6.4%). Similarly, an overwhelming number of respondents indicated constituency servants as their primary role orientation (57%) and very few indicated it as their least pronounced role orientation (3%). On the other hand, nearly 20% of respondents ranked material considerations, as well as partisan and public servant role orientations, as their top or bottom motivation. This sub-sample reflects the MPs that are most amenable and most averse to these motivations and role orientations. Table 3 presents the results for these subsets of MPs, and further differentiates between the full sample and ruling party MPs. The top part of the table reasserts many of the findings from the regression analysis but does not reveal any clear-cut factor that differentiates MPs. The least materially motivated MPs are younger and are more likely to be opposition members, while the most materially motivated MPs are more likely to come from less lucrative occupational backgrounds and more likely to have been selected in a primary. Similarly, partisans are younger, in more lucrative careers, and more oppositional but are more likely to be selected from uncompetitive districts in primaries as compared to the least partisan MPs. Clear public servants are more oppositional, urban, and selected in primaries as compared to MPs with the lowest public servant scores. However, the only factor that appears to clearly differentiate is the odd fact that the least partisan MPs were all elected from the ruling party in multimember districts and through a centralized selection system. Looking at the smaller range of regime MPs the least materially motivated MPs in are all employed in comparative more lucrative careers, competed in fairly safe constituencies, and were selected through a primary to compete in a multimember district. All of the partisan ruling party MPs were employed in lucrative careers and elected in very safe constituencies. Public servants are also predominantly urban and elected in multimember districts. Table 3: Subset Examination of Exemplar MPs Along Certain Motivation and Role Orientation Scores All MPs Age Elected Time in Legislature % “Other” Profession % Opposition Population Density Election Result % SMD % Primary N Ruling Party MPs Only Age Elected Time in Legislature % “Other” Profession Population Density Election Result % SMD % Primary N Material Motivation Highest Material Motivation Lowest Partisan Orientation Highest Partisan Orientation Lowest Public Servant Orientation Highest Public Servant Orientation Lowest 52.2 10.1 33% 22% 1091 54% 11% 44% 9 47.3 8.5 8.3% 42% 1142 61% 17% 25% 12 46.5 10.6 12.5% 31% 478 68% 5% 50% 16 50.8 5.4 40% 0% 661 54% 0% 0% 5 47.0 8.4 25% 58% 2179 55% 25% 58% 12 47 10.8 15% 23% 108 65% 23% 38% 13 50 10 29% 1397 56% 14% 57% 7 46.9 7.1 0% 1881 69% 0% 14% 7 46.2 11.6 0% 669 75% 18% 55% 11 50.8 5.4 40% 661 54% 0% 0% 5 48.6 9 40% 5085 60% 0% 40% 5 45.6 11.3 20% 118 69% 30% 50% 10 It is worth looking even close at a number of ruling party exemplars that buck expectations the most from an authoritarian context like Cameroon. We first identify MPs who exhibit the weakest material motivations and the strongest partisan identifications. In our sample these are identified as CM4 and CM25. CM4 was an early party activist in one of the CPDM’s urban southern strongholds and was president of the local youth wing. When he first ran for office in 2002 at the age of 31, he was pushed by the youth wing of the party. Since entering the legislature, he has also undertaken significant national responsibilities within the party and seen as a key organizer in the southern region. CM4 also rates prestige motivations above ideological and career motivations and ranks constituency service as his weakest role orientation. Respondent CM25 entered the legislature at the age of 51 after nearly twenty years as a barrister in the business center of Douala. CM25 was likewise active at local levels of the party, in particular with the women’s wing. She rates prestige as the primary motivation that led her to run for office, and also rates constituency service as her weakest role orientation. 18 We can also identify MPs with weak material motivations and the strongest senses of public service. Once again, our sample identifies two respondents – CM6 and CM17. CM6 was elected in 2013 at the age of 51 from a heavily urban constituency and regime stronghold. She had a lengthy career in government service and was added to a multimember electoral list as part of a women’s quota implemented in 2013. She rates career motivations more strongly than material or prestige motivations and ranks constituency service as her second most pronounced role orientation. CM17 comes from the same constituency, after a career as a business executive. He was more of a party activist than CM6 and held a position on the central committee long before he was elected to parliament. Unlike CM6, he was elected in a primary in 2007. CM17 likewise rates career motivations above material and prestige motivations and ranks partisanship above constituency service. TBD Conclusion This paper has offered one of the first attempts to examine motivations to run for office and MP role orientations in an African authoritarian setting. Much of the literature from African and authoritarian politics has made assertions about the strong material motivations that drive individuals to seek office, and this literature tends to see MPs as strongly beholden to partisan networks for their success. Given the difficulty of gathering quality data on MPs in authoritarian settings, these assumptions have only been empirically tested among a small range of cases. Moreover, much of the literature on authoritarian politics does not address the potential for 18 In fact, CM4 and CM25 are the only two respondents who rate constituency service as their weakest role orientation. legislatures to become contested arenas that are populated by individuals with diverse motivations and role orientations. There is now extensive literature from advanced democracies about the impact of individual and institutional-level variables on the kinds of candidates that emerge to run for office. Moreover, emerging literature from electoral authoritarian settings has begun to explore variation in MP motivations and behavior (Weghorst 2015, Reuter and Turovsky 2014). This study contributes to this literature with novel survey data from the case of Cameroon, a longstanding and fairly repressive authoritarian regime. The paper confirms the strong material motivations of most MPs, but contrary to expectations MPs in Cameroon also view themselves primarily through the lens of constituency service. This corresponds with survey findings established in Africa’s new democracies (Barkan et al. 2010). The study has also confirmed findings from other African authoritarian regimes that suggest that opposition MPs are more likely to seek non-material benefits from higher office. While we cannot conclusively say that opposition MPs chase ideological benefits in Cameroon, they do express lower levels of material motivations. Likewise, opposition MPs are likely to be somewhat more partisan in their orientation, which is to be expected if they are not as equipped as a regime candidate to deliver constituency resources. Opposition MPs likely emphasize partisan bone fides as an alternative to constituency service. The study has also tried to assess whether standard variables in studies of MP motivation and role orientation can explain some of the intra-sample variation. The answers to this question are mixed. Statistical evidence suggests that at the individual level, as expected, less lucrative occupational backgrounds influence the propensity toward material motivations, but age appears to be a factor that actually increases material motivations. On the other hand, more experience in the legislature corresponds with stronger partisan role orientations, which seems to indicate that while MPs express constituency service as their role orientation, the more successful ones can better navigate partisan networks. As expected, at the institutional level candidates selected in primaries are more materially motivated, and higher levels of electoral competition increases the salience of constituency service role orientations over partisanship. Candidate selection processes have not apparent influence. Importantly, a closer examination of exemplar MPs, and especially those with the lowest levels of material motivations, exposes more questions than answers. While opposition MPs are overrepresented in this subset of MPs, there are also a significant number of ruling party MPs. There are no clear discerning factors that explains the propensity toward lower material motivation or stronger public service role orientations. Weak ruling party material motivations can be found among MPs of different ages and experiences in the legislature and from rural and urban constituencies. All ruling party MPs with weak material motivations were selected in multimember districts and in constituencies where the ruling party has won over 60% of the vote. However, they also exhibit a variety of role orientations and alternative motivations to run for office. We focused on non-material partisans and non-material public servants and found that some were motivated by prestige and others by career ambitions. Closer examination of their career trajectories opens up the possibility that where specific types of legislators emerge from might be more idiosyncratic, and merits closer interrogation of the individual MPs. Nonetheless, a key takeaway from the paper is that even under the most difficult of circumstances there are MPs who buck theoretical expectations about authoritarian regimes. This sheds new light on the authoritarian legislature and presents it more accurately as an arena of contradiction and tension. While authoritarianism in Cameroon is empirically quite strong and the legislature extremely weak, individual MPs do emerge seeking something else than just immediate material benefits from office. This is particularly true for opposition parties, but importantly also for the ruling party. Better understanding these MPs provides new insights into how authoritarian institutions function, and where there are potential pockets of legislative efficiency, policy agenda setting, and even political resistance. In the aggregate authoritarian institutions like ruling parties and legislatures might serve to elongate authoritarian tenure. However, in their actual daily practice they are home to a diverse body of individual legislators, reflective of unique life experiences, operating within structural and institutional constraints. Supplementary Materials Appendix 1: Summary Statistics Appendix 2: Correlation between MP Motivations and MP Role Orientations Appendix 3: Regression Analysis Using Top-Two Choices Appendix I: Summary Statistics Mean 24.94 20.95 30.79 23.43 30.37 27.01 24.41 18.35 46.93 9.49 28.57% 20.63% 23.80% 31.74% 23.8% 807.05 61.38 28.57% 47.6% Material Motivation Career Motivation Ideological Motivation Prestige Motivation Constituency Servant Partisan Public Servant Entrepreneur Age Elected Time in Legislature % Opposition % Business % Professional % Government % Other Population Density Election Result % SMD % Primary SD 4.59 4.55 4.22 4.13 4.50 5.09 5.21 5.43 9.06 5.81 4.55 4.07 4.29 4.69 4.29 1861.85 15.29 4.55 5.03 Appendix II: Correlation between MP Motivations and MP Role Orientations Material Career Ideology Prestige Constituency 0.49 -0.15 -0.25 -0.13 Partisan -0.25 0.04 0.35 -0.09 Public -0.14 -0.08 0.18 0.08 Entrepreneur -0.01 0.16 -0.27 0.12 Appendix III: Regression Analysis Using Top-Two Choices Motivations (1) (2) Material Career Age Elected Time in Legislature Government Occupation Business Occupation Professional Occupation Opposition Party Population Density, log (department) Election Result Single Member District Primary Selection Constant 0.30 (0.15) 0.06 -0.21 (0.13) 0.05 -0.19 (0.13) 0.06 0.10 (0.13) -0.17 Role Orientations (6) (7) Const. Public Servant Servant 0.22 0.05 0.01 (0.15) (0.15) (0.15) 0.78** -0.15 -0.63 (0.22) -11.73** (0.22) 0.74 (0.22) 6.12 (0.23) 4.87 (0.26) 4.70 (0.21) -5.76 (0.36) -0.79 (0.28) 1.35 (3.52) -5.07 (3.56) -5.23 (3.60) 3.52 (2.93) 6.78* (4.10) 1.93 (3.82) -6.06 (3.86) 2.09 (3.44) 1.95 (3.11) -7.44* (3.06) 6.42 (3.33) -0.69 (3.21) 1.71 (3.23) 3.00 (3.23) -3.81 (3.84) -8.39* (3.51) 6.72* (3.64) -6.23 (3.67) -1.06 (3.07) 6.34* (2.71) 0.95 (4.04) 7.01 (3.61) -7.39* (4.02) 5.70 (3.09) -3.75 (3.18) -0.94 (2.75) 0.19 (2.71) -0.11 (2.78) 0.87 (3.52) 0.17 (3.17) -2.36** (3.58) 2.67** (2.56) -0.21 (0.73) -0.11 (0.78) -0.05 (0.79) 0.10 (0.74) 0.05 (0.77) 0.28* (0.74) -0.43*** (0.77) 0.15 (0.62) 0.00 (0.09) -0.77 (0.08) -7.68** (0.08) 9.50*** (0.07) -1.05 (0.12) 6.44 (0.09) -1.48 (0.08) -0.82 (0.08) -2.85 (2.80) 6.51* (2.75) -2.64 (2.33) -3.19 (1.99) -0.68 (3.37) -4.42 (2.97) 5.35 (3.40) 5.11 (2.96) -5.63 (3.25) 24.69 (15.44) (2.86) 36.10** (12.83) (2.99) 32.94** (12.28) (3.32) 6.26 (11.02) (3.46) -13.03 (15.42) (3.47) 72.30*** (13.08) (4.22) -0.42 (14.87) (3.21) 35.70** (13.21) (3) (4) Ideology Prestige (5) Partisan (8) Entrep. -0.32* (0.14) 0.21 Observations 70 70 70 70 70 70 70 70 R-squared 0.23 0.23 0.22 0.15 0.17 0.26 0.23 0.14 Prob>F ** * ** ** *** * Note: Ordinary Least Squared Regression for survey data. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05. Omitted category for occupation is “Other”. All models include unreported controls for gender and education.