Provisional results of Sierra Leone’s 2021 Mid-Term Census: Rebuilding trust will be difficult

Yusuf Bangura: Sierra Leone Telegraph: 3 June 2022:

The provisional results of the 2021 Mid-Term census recently released by Statistics-Sierra Leone have raised eyebrows among Sierra Leoneans. According to the results summary our population has increased by only 449,528-–from 7,092,113 to 7,541,641 between 2015 and 2021. (Photo above: Sierra Leone’s Statistician General – Osman Sankoh).

There have been massive population increases in some districts, especially those in the East and South, the strongholds of the ruling party; and population declines in Western Urban or the municipality of Freetown, as well as Bombali, Koinadugu and Port Loko districts in the North where the main opposition party draws its electoral strength.

Government advocates have long argued that the 2015 census undercounted people in the South and East and overcounted those in the North and Western Area, and have called for a new census that they believe will be credible. The 2015 census was conducted under the administration of the current opposition party, the All People’s Congress (APC). The victory of the Sierra Leone People’s Party (SLPP) in the 2018 elections provided an opportunity, in their view, to set the record straight.

In highly polarised polities the world over, census data is always contested as it is used by governments to allocate resources and parliamentary constituencies. An examination of the results for the 2004 census, which was conducted under an SLPP administration, and the 2015 and 2021 censuses (conducted by APC and SLPP administrations respectively) reveals a very troubling trend and casts doubt on the veracity of the 2021 results.

In 2004, our first post-war census, the North and Western Area accounted for 54.28 percent of the population, and the South and East 45.72 percent. The North and Western Area’s share of the population increased in the 2015 census to 56.2%, and the share of the South and East dropped to 43.8%. I characterised this ethno-regional demographic distribution and the electoral outcomes it generated as lopsided bipolarity in one of my papers on Sierra Leone’s elections.

Surprisingly, the 2021 census results have not only overturned the 56.2:43.8 ratio in 2015 between the North-Western Area and South-East, but has gone further to eliminate the difference that was recorded in the 2004 census under an SLPP administration. It is instructive to note that the census of 2004 was not challenged by any political party at the time.

The 2021 census has now recorded an outcome of almost parity between the two ethno-regional configurations. The North-Western Area is now believed to have only 50.01 percent of the country’s population and the South and East 49.99%. This manufactured parity is truly remarkable.

A similar trend can be observed when we examine the population shares of only the North vis-à-vis the South and East. In 2004, the North’s population was 77.27% of the combined population of the South and East. This figure increased to 81.38% in 2015. It is now only 66.39% of the South-East population in the 2021 census—much lower than even the figure for 2004.

The SLPP was stunned when, in 2018, it discovered that even though it had won the presidential election, the opposition All People’s Congress was going to call the shots in parliament. It had only 48 seats and the APC 68. The SLPP government engineered a process through the courts to cut the size of the APC parliamentary group and ensure some parity with the APC or dominance of the SLPP in parliament. Political observers have described the court’s decision to disqualify 10 members of the APC from parliament and award the seats to the SLPP without fresh elections as a judicial coup.

One of the reasons the SLPP fared poorly in the parliamentary elections of 2018, 2012 and 2007 was because of the first-past-the-post electoral rules that determine how MPs are elected. Even though the party improved its performance in the Western Area from 25 percent of the votes in 2012 to 33.5 percent in 2018, its voters were highly dispersed across the district and not concentrated in a few constituencies.

Under an electoral system of proportional representation (PR), the party should have received 33.5% of the seats or 9 out of the 28 seats in the Western Area instead of the 3 seats it received. Indeed, in 2012, it failed to win a single seat in the Western Area despite the fact that it had 25% of the votes. It is not surprising that the SLPP government has initiated moves to change the country’s electoral system to proportional representation.

This brief overview suggests that the SLPP has devised three strategies to advance its goal of parliamentary dominance. While the call for a change from first-past-the-post to PR is legitimate, the other two strategies (using the census to obtain parity and working with the judiciary to manufacture a majority in parliament) are brazenly dubious and high-handed.

The Big Lie: Are there more people in Western Rural than Western Urban?

The figures that really hit me in the census results are those for Western Urban and Western Rural. Does Statistics-Sierra Leone really believe that there are more people in Western Rural (662,056) than Western Urban or the Freetown municipality (606,701)? This must be the joke of the century.

There were 1,055,964 people in Western Urban ´and 444,270 people in Western Rural when the 2015 census was conducted. In other words, Western Urban or the municipality of Freetown had two and half times more people than Western Rural in 2015.

The really disturbing point about the Western Urban figure is that it is even lower than the figure recorded in the 2004 census (772,873), which was conducted under an SLPP government—the party of those who are currently in power. Are we to believe that the 2004 figure was also inflated by the Tejan-Kabbah SLPP government?

It really beggars belief that Western Rural has more people in the 2021 census than the highly overcrowded municipality of Freetown. Is this what will now be taught in our schools or used to allocate resources and constituencies?

Will Western Rural now have more constituencies than Western Urban? Anyone who is familiar with the settlement patterns in Western Urban and Western Rural will easily conclude that the 2021 census result as it relates to these two areas woefully fails the eye test. If we exclude the Waterloo-Grafton axis, Tombo and Goderich areas, much of Western Rural consists of sparsely populated villages.

Concern for the integrity of Stats-SL should have driven those in charge of administering the census to press the pause button when confronted with such a gross anomaly or data that offended common sense.

A census should aim to count an entire population, not those who are only willing to be counted. When the level of abstention or variance is shockingly high as in this case, it serves no useful purpose to validate the results unless if they’re meant to advance a political project.

And it makes no sense to defend the results and blame those who boycotted the census if the final product is unusable from a development point of view. It is spurious to argue that the final result captures those who made themselves available during the census period and is, therefore, an accurate reflection of the population of Sierra Leone and its different geographical units. This is clearly a botched job. It is the responsibility of those who believed they could count everyone despite widespread concerns and protests to deliver results that are unimpeachable or reflect the true population of the country.

What all this suggests is that a census should not be conducted in an environment with high levels of mistrust and threats of boycott from large sections of the populace. Ignoring the headwinds and stubbornly forging ahead politicises the institution entrusted to organise the process. This can only be a recipe for disaster.

It may explain why the World Bank and the European Union were critical of the process last year. The World Bank even withdrew its funding and affirmed that there was ‘insufficient time to satisfactorily address all pending actions, including evaluation of the pilot census, the field operation plan for the enumeration, and ensuring enumerators were adequately trained’ before the scheduled census date of 10 December 2021.

And the European Union described the Mid-Term census itself, barely a year before an election, as unprecedented. As it stated, “the use of an optional census to provide data which might be used to change electoral boundaries shortly before an election is not conducive to the political atmosphere or good electoral practice.”

Public trust in the census from large sections of the population nosedived after those interventions. Statistics-SL should have paused and tried to rebuild public confidence before proceeding to conduct a census that had ´boycott’ and high probability of dubious results written all over it.

It’s difficult to avoid the conclusion that this whole exercise has been a huge waste of public funds. No census has been as politicised and dodgy in Sierra Leone’s history as this one. Rebuilding trust in a census exercise is going to be very difficult after this experience.

Let me make one final point in concluding this piece. Reputations are won and lost on issues that advance the public good. The lesson from this sad episode is that state institutions should always be shielded (and must be seen to be shielded) from partisan politics, especially in a highly polarised ethno-regional setting like ours.


  1. I have followed the discussions and comments on the Mid-term Census report of 2021 with much emphasis on statistical science which is quite good. However, let us use common sense and remove the political lens if we are patriotic and mean development for our country.
    Our past censuses have been paper base and we migrate to real time digital process of data collection for the first time. With common sense, which of the two approaches do we need to support?. At the start of the process, statistics Sierra Leone said they cannot account for 59 % of the population for comparison because the data were no-longer available. Common sense has shown that paper based data are difficult to archive, because they are large and quick for the data to be missing.
    Another common sense we need here is that it is the responsibility of government to conduct census, and if the last census was conducted without any hidden agenda, why do you prevent your supporter from being counted? I think it would have been the responsibility of the opposition to encourage their supporter to come out in their number to be counted. So with the current result, whether sample or not, do you expect to cancel this result where political parties clearly cause people to boycott the process? Do you want political parties in the future to dictates the affairs of running government rather than the rules. We hope that SLPP will not stay in power for ever, but assuming that the go to opposition, can they prevent the running of government activities by preventing their supporters in their strongholds not to participate.
    Teachers were verified throughout the tenure of the last administrations and even current administration. what is wrong if government wants to verify the population of the country?. We agree that there was a reduction in the population of western urban, but can that stop us from moving ahead?
    Let us use common sense, that census was called, people went out to collect data, and those that resisted were not counted, but we cannot annulled the result because of them.

  2. Yes, there has never been a census where each and everyone is counted. That is why the term statistically significant is brought in to account for the margin of error – which is normally 3 to 4%. That is, a count is statistically significant when the estimate is equal to or more than 96% correct. Of course, statistics is not an exact science. But when a greater portion of the POPULATION is deliberately omitted (or disregarded) then, irrespective of the technology used, the outcome is not a CENSUS. It is inference used on a SAMPLE to assume an estimate. This is the case in the 2021 Mid-Term population and household census of Sierra Leone. And this is the argument that the Opposition should bring on board in order to win their case.

  3. I am not qualified to comment on the legal basis of the census process. I can only comment on the scientific integrity of the surveys.

    The objective was to count everyone in 2021 using the best statistical practice and equipment. All national censuses try to achieve that objective, but no country, with more than a million people, has ever been able to count everybody. It is impractical.

    South Africa just completed a national census in May 2022, and I was not counted even though census officials visited my security complex. Census officials were on TV in South Africa narrating challenges in accessing homes. The bottom line is that not everyone was counted, but that will not invalidate the census results. The government acknowledged that some people may not have been counted. People that were not counted were asked to complete the census form online. Obviously, not everyone can complete an online form.

    Efforts to discredit Statistics Sierra Leone are intellectually uninformed and a devaluation of the currency of common sense.

  4. The provisional results of the 2021 Mid-Tern population and household census amid the arguments and counter-arguments of their authenticity, it is in one’s best interest to look at the true definition and fundamental purpose of a census. What does CENSUS mean in government? With reference to arguably the greatest democracy, “the US census counts every resident in the United States. It is mandated by Article 1, section 2 of the constitution and takes place every 10 years.” In the UK, another great democracy, “the census is a survey that takes place every 10 years. It gives the most accurate estimate of all the people and households in England and Wales.” And, in statistics, “a census is a study of every unit, everyone or everything, in a population. It is known as a complete enumeration, which means a complete count.”

    On the other hand, what does a SAMPLE mean with reference to a census survey? “A sample is a subset of the unit in the population that represents data of a selected group of people in the population.” Consequently, if the argument, that a great number of the population of Sierra Leone was not enumerated in the 2021 census, as a result of the unwillingness of the residents that were unconvinced as to the credibility or legality and purpose of the census, then the whole exercise represents the data of a selected subset of the population. In effect, a sample of the population was taken, organised, and after structural analysis or manipulation, it was inferred to as the total population of Sierra Leone.

    Thus, by all accounts, the 2021 population exercise undertaken by Statistician General, Professor Mallam Osman Sankoh, to give an authentic estimate of the population of Sierra Leone, was not in line with the true definition of a census. And therefore, his effort was not only full of anomalies, but it was projected to misinform, and to a certain extent incite the people of Sierra Leone for his political and financial gains. It was also a complete waste of needed resources, and the Professor should be investigated and held accountable. It is not surprising a lot of employees with northern and north-western surnames lost their jobs at Statistics-Sierra Leone (SSL), at the start of the present ruling SLPP government in 2018. Within the overall scheme of things, the manipulation of the population distribution across the country was a grand plan.

  5. My respect and admiration for Dr. Yusuf Bangura for his academic achievement has been seriously dented. For the simple reason that his analysis of the Mid term Census report is far from professional balance. I happen to have read his analysis after listening to submissions from Dr. Richard Conteh and Dr. Denis Bright on radio Democracy. One thing I noticed that was glaringly exposed by the radio host on 98.1 FM was that both men argued from hind sight instead of using facts and figures. The radio host kept asking for evidence of their opinions. Sadly, both men did not make reference to any data from Statistics Sierra Leone.

    Unlike Dr. Bangura, who made reference to records but deviated from those facts and exposed his political orientation. To subscribe to the erroneous view that the population of Eastern Province are cooked up figures remain laughable. If the results of the percentage population distribution by district and Province from 1985 to 2004 and then 2015 all strongly give evidence of the Eastern Province much more populous than the Western Area (Rural and Urban) districts. What were we really expecting that the flooring population of Western Urban is a determinant factor to justify that Western Area is much more populous than Eastern Provonce? I am not sure whether Dr. Yusuf Bangura and others questioning the number of people living in Eastern Province want to tell me that they are not aware of the activities going on in the two regions. The population in Eastern Province is residential, while the one in Western Urban is partly residential and partly flooting.

    From logical perspective, are these men saying that because the population of Eastern Province is more than the Western Area and therefore the Mid Term Population and Housing Census of 2021 is incomplete and unreliable? To conclude using that weak inference is a fallacy. Honestly, I am not disappointed because I never expected any opposition supporter to glorify the report of the Midterm Population and Housing Census.

  6. In the aftermath of a criminally orchestrated census, with predefine partisan political and regionalism objectives, with results proving the worst ever, defying common sense, not to mention optics, PAOPA sympathizers have now turn to a spin game, concocting stories in explaining their fraud census result, talking about citizens have travel to their place of origin on the census night, citizens not wanting to be counted, and a host of other nonsensical utterance.

    To start with, the census was carryout for over a period of 3 months, with enumerators visiting random establishment unannounced, without prior notice or communication to the locals, and showing up randomly at a given time of the day, not midnight or dawn. So the question is, in the entire country, are South-Easterner’s the only citizen to have ceased all work and activities of their lives, staying put at their homes constantly, for over 3 months, just so they can be counted? Now it is a fact that majority of citizens who flooded Freetown after our civil war came from the North, so is PAOPA telling us now the population of Freetown have all along been imploded by South-Eastners, with their over 3 months migration back to their origin just to be counted, now halving Freetown population? Absolute garbage!

    Now, it’s an indisputable fact that South-Easterners are among the citizens who cherish their homelands, hence the least migrated within the nation. Travel in every region of the nation, you will sure to find Limbas, Temnes, Loko, Susu, Fullahs, etc, but the Mende tribe, hardly leave their homeland. So i am not sure where this concocted census night migration came from. If anything, the population in the North should have quadruple if we are to buy this ‘cock and bull, story of census night migration, since majority of the migration tribes i mention above have Northern origin. As per the digital nature of the census making it the best, oh please! I suggest you educated yourself regarding digital data inputting/harvesting. Humans input the data, not the computer itself. What stop me to input 10 family members for a household that has only 3 members? It’s sad that someone of you come across as educated, but politics have corroded your thinking faculty!

  7. Thanks to the almighty for blessing our nation with Professor Osman Sankoh who has displayed his PATRIOTISM by performing his constitutional duty. He successfully collaborated with the Finance Minister, the Minister of Economic and Planning Development, as well as the United Nations Population Fund (UNFPA) and the United Nations Economic Commission for Africa (UNECA) to name but few. Amid all the challenges, NOISES AND DISTRACTIONS, he successfully completed this UNPRECEDENTED ( ONLY) Mid Term DIGITAL Census which has finally corrected the mistakes of the past. We hope and pray that this process will continue to improve in 2025 and beyond based on the fact that a solid foundation has finally been built on DIGITAL AND MODERN TECHNOLOGY.
    The link below shows your true character:

  8. I am a big fan of Yusuf Bangura. His academic discuss on this forum, about historical and social issues, have been very credible.
    Statistics is a science, but not an exact science. It depends on the sample population, the data collection methods, the variables, and the conditions of the survey.

    The survey design and methodology used for the 2021 census followed good international survey practice. Data was captured in real time for the first time in Sierra Leone. The question determining the place of residence of the respondent was ‘where did you sleep the night before you participated in the survey?’

    The Chief Statistician is one of the most experienced demographers in the world. He has participated in, and analysed demographic surveys in many countries in Africa and Asia. His design and analysis of census data have been peer reviewed by several international statistics journals. His science is well respected globally.

    So where did we go wrong? Did people refuse to be counted? YES Did people move to their ‘home districts’ to be counted as been resident there? YES (Regional WhatsApp groups were appealing for people to travel home so that they can be counted in their districts).

    Is the survey outcome a true reflection of the survey methodology? YES. Is the survey outcome a true reflection of the number of people in Sierra Leone today? Probably not, and the reasons for that are not scientific.

    Yusuf Bangura’s review of the political and historical context is great. His attempt to doubt the demographic principle that informed the census and scientific outcome, I am afraid, is wrong.

    This is a clear case of blaming the messenger.

  9. First was the weaponizatin of our staple food rice .Now the one directionless government of Bio has once again demonstrated how far he is ready to go to weaponised census figures for his political advantage.The story of the Sierra leone preliminary census results is similar to the story of Nigeria’s 1962 census that never gets old. Now we know why Bio was pulling all the stops in the middle of the Covid19 pandemic to organised this census.Any one who question the logic of it was invited by the Siera leone police to explain themselves .Dr Femi Cole, Dr Samura , and Dr Bright.”In 1962 Southern politicians seeking to end the North’s dominance of Nigerian politics decided that the only way to do it was through the organisation of a census for Nigeria. Population figures at the time determined not only Parliamentary representation , revenue allocation but employee service in the public sector.

    In may 1962 the first census in Nigeria was held after independence .Preliminary results was quiet clear as to what has happened.The North population have grown from 16million to 22milion since the last census that was taken in colonial times in 1952.An increase by 30%.But in some parts of the South and West the population have increased by 200%. Though the results were not made public , the Northern leaders were not about to take those results lying down .A new census was held in 1963 and the results were an eye opener .An additional 8 million people were discovered in the North which was added to the previous year’s census which once again tilt the scales in favour of the North with 31 million people.”To me we might just be in the same census terrain that Nigeria finds itself back in 1962. I visited my village not long ago in the North .After many years of absence I couldn’t recognise half of the strange faces that have moved in to my village and called it their home .

    And many of the Fulani men including some of my distance relatives have two to three wives .And the population of the village have explored beyond recognition.So I don’t know which parts of the country the census official visited and got this skewed results that only lead to one thing to bolster Bio’s electoral chances .At the moment is the only trick that he can pull out of his rabbit hat to stand any chance for a second term .When you failed to honor your promises and lied to your people , you have to restore to desperate measures. Desperate times called for desperate measures .The moment of truth is finally catching up with Bio .You can’t hide from the facts Mr President.

Leave a Reply

Your email address will not be published.


This site uses Akismet to reduce spam. Learn how your comment data is processed.