Strategic ballot removal: an unexplored form of electoral manipulation in hybrid regimes

ABSTRACT

The literatures on electoral manipulation and rejected ballots have yet to engage one another in a compelling manner. This article provides the theoretical foundations for rejected ballots as electoral manipulation by exploring incidents of suspicious rejected ballot rates and practices around the world with a special focus on Zambia. Not only did the rate of rejected ballots in Zambia double between the 2015 and 2016 presidential elections, but a disproportionate increase was observed in president’s home province. Leveraging an original dataset, the article models a largely unnoticed form of electoral manipulation: the strategic rejection of opposition ballots by biased polling officials. Analysis reveals that more rejected ballots were associated with increased vote shares for the ruling party in the president’s home province, indicating probable electoral manipulation. Raising awareness around this difficult to detect, but likely pervasive, form of manipulation should help to improve electoral quality in hybrid regimes.

KEYWORDS:

• electoral manipulation
• spoiled ballots
• hybrid regimes
• ethnic brokerage
• electoral administration
• Zambia

Acknowledgement

This article benefited tremendously from the participants of the 2017 Concordia-McGill Universities African Studies Conference and University of Notre Dame Comparative Politics Workshop. A special thank you to Michael Coppedge, Jaimie Bleck, Mike Hoffman, Peter Johannessen, and Luis Schiumerini for their insightful feedback, and the three anonymous reviewers for their thorough comments.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

Paul Friesen http://orcid.org/0000-0002-9597-6780

Notes

1. Schaffer, The Hidden Costs of Clean Election Reform.

2. Alvarez et al., Election Fraud.

3. Birch, Electoral Malpractice; Simpser, “Unintended Consequences of Election Monitoring.”

4. Schedler, “The Menu of Manipulation”; Levitsky and Way, Competitive Authoritarianism.

5. Bermeo, “On Democratic Backsliding.”

6. Schedler, The Politics of Uncertainty.

7. Birch, Electoral Malpractice; Cox, “Authoritarian Elections and Leadership Succession, 1975–2004”; Cheeseman and Klaus, How to Rig an Election.

8. Van Ham and Lindberg, “From Sticks to Carrots.”

9. Albaugh, “An Autocrat’s Toolkit”; Raftopoulos, “The 2013 Elections in Zimbabwe.”

10. Zambian law defines a rejected ballot as:

one which: does not bear the official mark; is marked more than once; has the identity of the voter; is not marked by the voter; is not clear for whom the voter has voted for; or is not the official ballot paper issued for that election. (ECZ Voter Education Facilitator’s Handbook, 41)

11. Rejected ballot rates during the second multi-party era: 1991 – 3.1%; 1996 – 5.0%; 2001 – 1.6%; 2006 – 1.7%; 2008 – 1.3%; 2011 – 1.4%; 2015 – 1.0%; 2016 – 2.3%.

12. The 2016 ballot meets 15 of the 16 criteria set out by Norden et al., Better Ballots, indicating a strong design (See online Appendix I).

13. CCMG, “2016 Election Report,” 43.

14. BBC notes that, “high numbers of invalid votes can mean that officials are disqualifying ballots for the slightest imperfection, even when the voter’s intention is perfectly clear.”

15. Hanlon and Fox, “Identifying Fraud in Democratic Elections.”

16. Economist, “Suspicions of Vote Rigging in Honduras’s Election”; Hanlon and Fox, “Identifying Fraud in Democratic Elections.”

17. Cheeseman and Klaus, How to Rig an Election.

18. Danilova, “Ukraine Votes.”

19. Masinga, “IEC Dismisses ‘Vanishing Ink’ Conspiracy.”

20. Naahee, “JP, PPM Warn of Disappearing Ink Pens.”

21. One reviewer rightly pointed out that a lack of signal might include polling station officials either not engaging in strategical ballot removal at all or continuing to engage in the practice but based upon personal preferences rather than group-leader preferences. While the data cannot easily confirm one position over the other, the overall change in rejected rates would suggest that when polling station officials lack a clear signal, engagement in any form of ballot removal declines.

22. Economist, “Suspicions of Vote Rigging in Honduras’s Election.”

23. In online Appendix II, historical rates of rejected ballots are tracked across several elections. In every election since 1991, Nyanja percentage is positively correlated with higher rejected ballot rates and statistically significant at p < 0.05 in five out of nine elections.

24. Hyde, “The Observer Effect in International Politics”; Asuka et al., “Electoral Fraud or Violence”; Ichino and Schüdeln, “Deterring or Displacing Electoral Irregularities?”

25. Birch, Electoral Malpractice.

26. McAllister and Makkai, “Institutions, Society or Protest?”

27. Power and Garand, “Determinants of Invalid Voting in Latin America”; Jackman, “Political Institutions and Voter Turnout in the Industrial Democracies”; Uggla, “Incompetence, Alienation, or Calculation?”; Denver et al., “Rejected Ballot Papers in the 2007 Scottish Parliament Election”; Pachon et al., “Ballot Design and Invalid Votes.”

28. Power and Garand, “Determinants of Invalid Voting in Latin America.”

29. McAllister and Makkai, “Institutions, Society or Protest?”

30. Power and Roberts, “Compulsory Voting, Invalid Ballots, and Abstention in Brazil”; Buxton, “Venezuela: Degenerative Democracy.”

31. Bratton et al., “Zambia at a Crossroads.”

32. Kriegler, “Report of the Independent Review Commission.”

33. Meyersson, “Is Something Rotten in Ankara’s Mayoral Election?”

34. Benton, “Configuring Authority over Electoral Manipulation.”

35. Meyersson, “Is Something Rotten in Ankara’s Mayoral Election?”; Hanlon and Fox, “Identifying Fraud in Democratic Elections.”

36. Posner, Institutions and Ethnic Politics in Africa.

37. According to the 2010 Zambian census, the primary languages are combined with similar linguistic variations are Bemba (41%), Nyanja (23.3%), Tonga (14.5%), Northwestern (6.6%), Barotse (6.3%), Mambwe (3.2%), Tumbuka (3.3%), English (1.7%), and Others (0.3%). Mambwe and Tumbuka are not officially recognized.

38. Posner, Institutions and Ethnic Politics in Africa.

39. Eifert et al., “Political Competition and Ethnic Identification in Africa”; Bratton et al., “Voting Intentions in Africa.”

40. Posner, Institutions and Ethnic Politics in Africa.

41. African Confidential, “PF Picks Two Candidates.”

42. According to Africa Confidential, Banda would support Lungu’s campaign financially, facilitate businesses connections, and campaign for Lungu in Eastern Province, while Lungu, should he win the election, would then clear Banda’s son of corruption charges, replace the current Attorney General and Public Prosecutor, and reinstate Banda’s presidential immunity (Africa Confidential, “PF Set for Victory”).

43. Under President Sata in 2011, the cabinet was composed of approximately 63% Bemba, 11% Nyanja, 5% Tonga, and 21% from all other ethnic groups. As of 2017, Lungu rebalanced the cabinet to 57% Bemba, 29% Nyanja, 0% Tonga, and 14% from other ethnic groups (Coded by author, four ministers left unassigned).

44. Breeze FM, “President Edgar Lungu Warns to Fire Lazy Workers”; Baldwin, “Why Vote with the Chief?”

45. Zambian Eye, “Chief Mpezani Assures Lungu of Victory.”

46. Other institutional differences included having casting one ballot in 2015 compared to five ballots in 2016, though boxes and ballots were colour coded (Commonwealth, “Zambia General Elections”). Also, although a constitutional amendment passed changing Zambia’s electoral framework from a one-round to two-round system, this legal change had no impact on the 2016 elections as a second round was not required.

47. While an influx of new voters ahead of the 2016 elections seems an obvious source of increased rejected ballot rates, ahead of the 2011 elections, 1.5 million voters were also registered for the first time and again the rejected ballot rate was lower (Creative, “Electoral Peacebuilding in Zambia”).

48. ECZ, “General Elections 2016.”

49. Central Statistical Office, “Zambia Demographic and Health Survey 2013–2014.” The language categories included Bemba, Nyanja, Tonga, Lozi, Luvale, Kaonde, Luanda, and Other. For the Other category, constituencies were marked as having either a variant of one of the official national languages, or a distinctive language marked as “non-official” using the 2010 Zambian census data and the University of Leiden’s “Tribal and Linguistics Map of Zambia.”

50. Green and Vavreck, “Analysis of Cluster-Randomized Experiments.”

51. Angrist and Pischke, Mostly Harmless Econometrics.

52. Keele, “The Statistics of Causal Inference.”

53. The constituencies coded as 1 include: Chingola, Kalulushi, Luanshya, Mufulira, Nkana, and Solwezi West. Lusaka Times, “Mopani Lays off over 4,000 Workers”; Lusaka Times, “More Miners on the Copperbelt are to Lose their Jobs.”

54. See online Appendix III for determinants of PF support in both 2015 and 2016.

55. Lindberg, Democracy and Elections in Africa.

56. Not only did voters from Nyanja areas move en masse from MMD to PF following the Banda-back Lungu election campaign, but many local and national MMD politicians left party for PF, thus integrating the MMD-brokerage network.

57. See online Appendix IV for a replication of Table 4 using the absolute rejected ballot rates instead of the difference in rejected ballot rates. The models are extremely similar due to the relatively level distribution of rejected ballot rates in 2015.

58. See online Appendix V for causal mediation analyses for all ethnic groups.

59. See online Appendix VI for models used to estimate Figure 4.

60. See online Appendix VII for supplementary models comparing the different subsets of polling stations data with and without interaction effects. The interaction plots in Figure 5 were produced using the estimands from models 4 and 6.

61. The mean Nyanja percentage in the “dominant” constituencies is 95.6%, while the percentage is 28.6% in “diverse” constituencies.

62. 77,000 polling station staff were recruited to cover 7,001 polling stations (EU, “Final Report”).

63. Zambian Watchdog, “District Commissioners Imposing Cadres on ECZ Director.”; Zambian Watchdog, “Kafue Polling Agents are all PF Cadres.”; Zambian Watchdog, “Mulobezi DC Takes Over Recruitment of Election Officers.”

64. Daily Nation, “Remove Lungu, UPND Begs Court.”

65. EU, “Final Report.”

66. Ruiz-Rufino, “When do Electoral Institutions Trigger Electoral Misconduct?”

67. See online Appendix VIII for a list of domestic and international observation groups.

68. CCMG, “2016 Election Report.”

69. Simpser, “Unintended Consequences of Election Monitoring.”

70. Hyde, “Foreign Democracy Promotion, Norm Development and Democratization”; Birch, Electoral Malpractice.

71. EU, “Final Report.”

72. If one replaces the observed rejected ballot rate with the “natural rate” found using polling stations close to 100% PF vote, then the number of rejected ballots falls by 5945. Secondly, using fitted values from Table 2, eliminating the effect of Nyanja on rejected ballot difference (Nyanja from 23% to 0%), would results in a decline of 5734 rejected ballots.

73. The Varieties of Democracy (V-Dem) Clean Elections Index assesses the “absence of registration fraud, systematic irregularities, government intimidation of the opposition, vote buying, and election violence” (Coppedge et al., “V-Dem Dataset v7”).

Additional information

Notes on contributors

Paul Friesen

Paul Friesen is a PhD student in the Department of Political Science at the University of Notre Dame and a PhD fellow at the Kellogg Institute for International Studies. His research focuses on political behaviour, electoral integrity, ethnic politics and political parties in sub-Saharan Africa.