Introduction

With almost 400 Members of Parliament, Uganda is one of the most overgoverned countries in the world. On Tuesday, August 4, Parliament passed a law allowing the NRM to create 43 new constituencies. This will potentially increase parliamentary representation to almost 570 MP’s! This has significant financial implications on the taxpayerwho is already overburdened by the expense of an overcrowded poor quality legislature largely considered to be a rubber stamp for the presidents whims which include the creation of yet another batch of new districts!

This is considered by many to be a form of gerrymandering. By definition gerrymandering is the manipulation of voting districts and constituencies in order to benefit the ruling power or party. The president has recently recalled parliament to “discuss” several new districts created just prior to an election. The creation of these new districts and constituencies was initiated by the Ministry of local government rather than the electoral commission and parliament as the constitution mandates. Their creation is a fulfillment of the presidents campaig promises during the last election!

The government claims that these new districts are in response to population growth.

Methodology

We hypothesised that the government of Uganda has created these new districts and constituencies for reasons other than population growth.

Census figures and numbers of districts in Uganda were obtained from UBOS and wikipedia for the years 1969, 1980, 1991, 2002 and 2014. population density for each district was calculated from these figures as were the absolute growths in population and number of districts at each census point.

We expected to demonstrate that despite the steady population growth, the population density per district had declined suggesting an excess of new districts (and constituencies) and by extension new MP’s and local administrators.

Free and opensource R Statistics software with the pander package was used to analyse the data. RMarkdown was used for documentation and reproducibility. Data and methodology are made available.

Results

The results demonstrate that the census to census absolute population growth differences between 1969, 1980, 1991 and 2014 were 3.1, 4.1, 7.5, 12.8 million people while that of districts were 12, 5, 18 and 55 new districts at each census. This suggests that the number of districts (and therefore constituencies and MP’s) has grown exponentially at a rate far in excess of the population growth rate! Of note is that the growth in new districts more than halved inn the period between 1980 and 1991 compared to 1969 and 1980 before trippling in the next decade 11 fold!

library(pander)
Districts = c(21, 33, 38, 56, 112) 
year=c(1969, 1980, 1991, 2002, 2014)
pop=c(9.5,12.6,16.7,24.2,37)
den=pop/Districts
census=cbind(year,Districts,pop,den)
census=data.frame(census)
pander(census,caption="Table 1: Population and administrative unit growth at each census")
Table 1: Population and administrative unit growth at each census
year Districts pop den
1969 21 9.5 0.4524
1980 33 12.6 0.3818
1991 38 16.7 0.4395
2002 56 24.2 0.4321
2014 112 37 0.3304

The population density in each district has dropped significantly since 1969 by almost 120,000 people (Table 1:)

plot(year,den,xlab="Census year",ylab="population density/district (million)",main = "Population density per district by year",lwd=6)
lines(year,den,lwd=3)

#Mean population density per district
mean.den=(sum(den)-0.3333333)/4
Expected_No_of_districts=37000000/(mean.den*1000000)
Excess_No_of_districts=112-87

The expected mean population density per district is 4.257098810^{5}(426,454) people per district while the expected number of districts is 86.9136512, resulting in an excess number of districts of 25.

Discussion

The population density per district has dropped by about 120,000 people since 1969 while the number of new districts has grown exponentially. This suggests that the rationale used to create these new administrative units and constituencies is false and means they were formed for reasons other than population growth including possibly gerrymandering.

New districts and representatives have become a “public good” gifted by a “magnanimous” president currying favour with the electorate. This latest round is a “fulfilment” of the presidents campaign promises during the last election. It affords him the campaign opportunity to travel to the new district, an opportunity denied other presidential candidates. The message is that if you vote the incumbent, he will return the favour “generously”!

Conclusion"

The results of this small study suggest that there is more to the NRM governments formation of new administrative units and constituencies than population growth!

More study is needed to better understand these reasons and their impact on the outcome of elections as well as on public expenditure on legislators.

References