Tanzania’s Biometric Voter Registration process is now proceeding at full pace. It had previously been delayed for several months while waiting for voter registration kits, which meant it was impossible to conduct the constitutional referendum as scheduled at the end of April (as predicted here). Now, there are concerns that the process is happening too fast, not giving people enough time for everyone to register:
“The machines are just too few,” complained one resident of Geita. “This situation will cause more chaos because there are very many people remaining yet there are only two days left.” And according to the Daily News, people in Arusha “have now resorted to spending nights at the registration centres in order to beat the jam”.
This led me to think: what proportion of the voting age population needs to be registered in order for an election to be considered legitimate?
There is no universally-agreed answer to this question. And voting (and registration) is not compulsory, so people are free to decide against registering themselves. But the ACE Electoral Knowledge Network, a project of the International Institute for Democracy and Electoral Assistance and others, puts it like this:
“There is consensus that for a system to be considered democratic and representative, it must provide equal opportunity for everyone to participate in an inclusive voter registration process.”
In other words, there’s no strict percentage cut-off point. What matters is that Tanzania’s estimated 24 million eligible voters should all have a fair chance to register if they wish to do so.
However, it is possible to compare Tanzania’s voter registration record to that of other countries:
We can see that voter registration in Tanzania was relatively low in 1995 and 2000 (63% in both cases), but much higher in 2005 (94%) and 2010 (95%). The chart also suggests that there was a similar improvement across the region.
Looking a little wider, the average voter registration as proportion of voting age population in all African elections since 1995 is 83%. In Tanzania it is 79%.
Next, just for Tanzania, lets combine the figures for voter registration and votes cast:
Here we can see again the improvement in voter registration in 2005 and 2010. But we can also see the dramatic drop in the percentage of voters who actually used their vote. The orange section of the chart – registered voters who did not vote – gets bigger in more recent elections.
Incidentally, the high level of voter registration in 2005 and 2010 suggests that some of the decline in turnout (previously discussed here and here) is because people who were never very interested in voting may have registered for other reasons – such as to get a trusted form of ID card.
So what would be a fair figure for voter registration in 2015? What figure would suggest that the National Electoral Commission (NEC) has done a good job and given everyone who wants to register a fair opportunity to do so?
I can’t definitively answer that question. But the figures from 2005 and 2010 suggest that Tanzania should be able to register over 90% of the voting age population. And to stay ahead of continent-wide trends means a minimum of 80% registration. Anything less than that would not look good.
The data used in this post comes from the Voter Turnout Database of the International Institute for Democracy and Electoral Assistance.