One person, one vote? How many voters does each Tanzanian MP represent?

UPDATED 12/5/15 – See below

The job of an MP is to represent their constituents. But how many voters do they have to represent?

If there’s a big difference between the number of voters represented by different MPs, it can be unfair on MPs who have to represent a larger number of people. More importantly, it can be unfair on the voters. Those who live in constituencies with a smaller population would potentially have more influence over their MP – and therefore over the government as a whole. Those in constituencies with more voters would find it harder to make their voice heard.

On the other hand, there can be good reasons to have slightly different sized constituencies. Where there are very clearly defined communities – such as islands, for example – it can be appropriate not to split these up or to artificially combine them with a geographically separate community. And it doesn’t make sense to keep all constituencies exactly equal-sized by repeatedly re-drawing constituency boundaries as people move around the country and populations change.

The National Bureau of Statistics (NBS) recently published estimates of the population in each region, district, constituency and ward, across the whole country – both the total population and the estimated number of voting-age adults. This means we can see how fairly – or unfairly – the country has been divided into constituencies.

First, I’ve grouped the constituencies by region, and coloured them according to whether the constituency is predominantly urban or rural. The bars show the voting age population in each constituency, according to the NBS data. And second, I’ve also worked out the average number of voters per constituency for Dar, other urban and rural constituencies, and for constituencies won in 2010 by CCM and by opposition parties. In all these cases, I’ve left out Zanzibar, as the system for electing MPs is different on the isles.

People living in constituencies with bars that don’t reach the “average” line are better represented, while those in constituencies with longer bars are less well represented.

Access or download the full dataset here.

Some quick thoughts:

First, there is clearly a very wide range here. In Ubungo, Dar es Salaam, 575,000 voters elect a single MP, while in Pangani (Tanga), Mafia (Pwani/Coast) and Mchinga (Lindi), the MPs each represent less than 40,000 voters.

Second, the situation for voters in Dar is particular unfair, except in Ilala constituency. The seven constituencies with the largest number of voters per MPs are all in Dar es Salaam.

Third, there is only a small difference in the size of the electorate between rural constituencies and urban constituencies outside Dar.

Fourth, there is only a small difference in the number of voters per MP in constituencies won in 2010 by CCM and the opposition. Opposition MPs represent, on average, a slightly larger number of voting age adults (152,000) than CCM MPs (119,000).

Almost certainly, the main reason for the disparities here is that the population of some constituencies has grown much more quickly than others.

The National Electoral Commission has the power to redraw constituency boundaries, and to make the constituencies more equal. This data suggests that it might be time to use it.


UPDATE 12/5/15

Citizen fp 120515 new majimboSeveral newspapers reported today that the National Electoral Commission is indeed planning to announce several new constituencies. The Citizen made the story their front page lead.

From the Daily News:

the Director of NEC, Mr Julius Malaba, said demarcation of the constituencies will base on the population people in respective constituency as provided by the National Bureau of Statistics (NBS) and the number of current constituencies which is 239.

In urban areas, the director explained, a constituency will be made up of about 325,000 people while in rural areas the number has been set at 250,000.

He also called for applications and proposals on the new constituencies from stakeholders.

The Citizen’s article differs on some of these details.

4 thoughts on “One person, one vote? How many voters does each Tanzanian MP represent?

    1. mtega Post author

      I saw that announcement as well – remarkable that this blog has that kind of influence! 😉

      Sadly, the changes mentioned in the article are likely to widen the representation gap, by dividing relatively small, rural constituencies and leaving Dar untouched, rather than reduce it. Though of course there may well be changes other than those mentioned here.

      I will try to look again, once NEC has reached it’s final decision on boundaries.

  1. puza

    By default all Dar constituents are over represented since all the MPIGS lives there. On a serious note, the notion that MPigs should be selected based on the number of residents is not applicable on this case. For starter the mp really engage with the voters unless there is an election.
    Second their primary role is law making and to ensure that every citizen rights are upheld, which is irrelevant in this case MP are more interested inot raiding their saLary tHan haging a natonal gas law.
    And if the MPIGS are really interested in local development, the size of the area falling under their mandate has more influence on the challenges associated with stimulating the economy, having more people is a plus and not hindrance to the Mpig performance.
    And lastly tanzania government is centralized, now you can ignore the rest of my comments.

  2. Anthony

    This is very insightful piece Ben. It says quite a lot more of the underlying frictions of distributive politics in Tanzania than one could ever fathom. I think it will be interesting to extend this analysis by looking at the distribution of actual voters turn-out. For instance, Dar seems to be skewed, but what proportion of people in Dar actually voted relative to other regions? This will be very informative when one thinks of citizen agency. More so when we think along the lines of measuring MPs ability to deliver. I think you piece brings to bear lots of interesting questions, at least descriptively for now. For example how are social economic indicators of these constituents look like. I wont be surprised that levels of education and health outcomes to be correlated with voting outcomes. And what kind of economic activities these constituents have? I guess there are a lot of questions from your nice piece that will be very interesting to stimulate further debates in Tanzania.

    I am currently developing a research project on Tanzania’s MPs, distributive politics and their outcomes on governance, public policy and socio-economic outcomes in the country. Your piece is a timely ingredient!

Comments are closed.