Managing Expectations during Elections to Prevent Violence

Following are my notes on the discussion we had a few days ago about the electoral process, its challenges and the improvement we could bring to facilitate participation and minimize occurrence of errors and fraud. After a reminder of the ideas presented previosuly, I will built on those thoughts about elections to offer my perspective on managing expectation during elections in a global context.

First, let's recap briefly what Alex, Ian, Tux and A series of tubes have proposed so far: Series of Tubes think we should have:" Receipts to print with a randomly generated verification code. This way, the table could be publically analyzed to verify that the reported totals are correct AND each individual could verify that their vote is both included in the table and is correct." . Ian thinks: "the e-voting system will benefit from impovements in its auditability, transparency and accountability especially if such improvements are cost effective. From a developing country’s perspective, electronic voting is just one and may be a minor aspect of the larger issue of ensuring free and fair elections and strengthening democracy. Still, the frequent use of money, muscle and state power to influence election outcomes is a much bigger challenge to tackle." Finally Tux warns of the perils of investing in big innovations that have not been fully tested and Alex proposes that votes "be updated and compared with physical records every hour or so to ensure maximum accuracy".

In our following conversation, we clarified the following points:

-Elections are measurements, therefore we all should expect measurements errors and be transparent about them.
-Elections are not only about determining the choice of the constituents but also about providing trust to the same constituents that their votes are accounted for properly.

About preventing voting errors or fraud:
- A consensus was reached that combining a paper trail of voting record and a stable digital streamlining of the process.
Paper trail provides the critical physical evidence of a casted vote. Digital streamlining helps limit the intermediary steps in vote count so that the probability of fraud is reduced.
- Search for statistical hints that suggest that polls might have been fabricated.
- Spreading the voting period over three days instead of one to make sure everyone has a chance to vote, unclog voting poll and prevent wearing out of poll volunteers.
- The importance of exit polls as a mean of verification of the official count.
- In fact, holding opinion polls in between election cycle helps make sure that there are no sudden outrageous difference between ratings and voting poll.

Those are the different interventions we discussed that help prevent elections from being stolen and increase voters' confidence in the electoral process.

And yet all of them taken together still won't ensure that the voting process is error-free.

Of course, each intervention aimed at increasing the reliability of elections also implies an additional significant financial burden. The question becomes how big of a margin of error would a society be willing to accept given their specific political context.

Mass election fraud have plagued so many countries for the past five years. Just last week at the European Parliament as he discussed African elections, Mr Charles Milupi bemoaned that "Elections, which should be the ultimate arbiter of political differences as well as the guarantor of peace, stability and security, have become a major source of conflict and political violence".

The past surges in violence in Kenya, Guinea and more recently in Gabon were due in part because evidences of wrongful procedures were witnessed but still left unresolved.

Whether the elections may have provided the right results (doubtful at best) were rendered pointless because many suspected fraud before the elections started and further evidence confirmed those doubts. It may not be entirely justified but the onus is on African nations to show additional guarantees that their electoral processes are fair and produce an acceptable outcome.

One technological solution to better monitoring in countries with limited poll monitoring capacities can be provided by grassroots communities crowd sourcing interactive platform like Ushahidi via basic text messages.
Here is an example of the Kenyan-born project used for the vote report project during elections in India.


Erik, co-creator of Ushahidi recently discussed the pros and cons of crowd sourcing election monitoring.
There are surely set back to trust regular citizens with the task of monitoring elections but the circumstances of the developing world are different from the western world so our democracies require different solutions. An important feature of the ushahidi platform that helps sort out the signal from the unavoidable noise stemming from such data collection is a digital information filter called swift river.

The process of noise filtering is a bit complicated and explained in the previous link but here is a short video that makes sense of how it works:

<
An Introduction to the Swift River initiative from Ushahidi on Vimeo.


This is a prime example of information technology providing the first step in moving towards elections that are more trustworthy and less prone to massive fraud, especially in developing regions where monitoring capacity is scarce.

Instead of resorting to nonconstructive statements that argue that some countries are just not mature enough yet for democratic elections, one could provide ways of limiting the possibility of fraud and prevent possible post-electoral violence in many countries facing elections this year on the African continent.