REPORT ON THE PROVISIONAL RESULTS OF THE 2015 POPULATION AND HOUSING CENSUS OF SIERRA LEONE

SUBMITTED BY SLPP TECHNICAL COMMITTEE ON 2015 CENSUS

Background:

Following the release of the provisional 2015 census report published by Statistics Sierra Leone (SSL) on March 31, 2016, the Sierra Leone People’s Party (SLPP) convened a stakeholders’ meeting on Tuesday 6th April, 2016 at the Head office, Lightfoot Boston Street in Freetown and constituted a Technical Committee to examine the report. The following members were appointed to serve in the Committee.

1. Prof. Herbert Bob Kandeh
2. Prof. Lawrence Kamara
3. Alhaji Mohamed L Kamara-Kolleh
4. Moses Williams
5. Ing. Francis Lahai
6. Mr. Jacob J Saffa

 

1. Terms of Reference (TOR)

The committee was mandated to look into the following issues:

• Whether the provisional census report is credible to be used by national and international stakeholders for evidenced-based policy formulation and decision making in the private and public domain.
• Determine whether the SLPP and other interest groups were excluded in the technical advisory structure and mainstream proceedings of the census program during the tenure of operations up to the production of the 2015 Provisional Population and Housing Census report.
• Evaluate the reliability and acceptability of the provisional census document to guide preparations for delimitation of national constituencies and the conduct of public elections by the National Electoral Commission (NEC) in particular the forthcoming Presidential, Parliamentary and Local Council elections scheduled for 2018.
• Assess whether the provisional 2015 census figures for the period 2004 - 2015 announced were exaggerated indicating extraordinary growth with large increase in the national population compared to previous censuses covering same number of years.

 

1. Methodology

The Committee held a series of meetings to address the TOR. In the initial meetings, the Committee agreed that the technical committee focuses its work on the exclusion argument, the credibility of the total provisional figure using demographic and economic arguments. The Committee reviewed both past and provisional census figures for Sierra Leone and used demographic statistics obtained by SSL. Thereafter, sub-Committees were set up to work on the various aspects. The various drafts were consolidated and shared with members of the Committee through email. Comments and contributions were made and the draft revised accordingly.

 

The Committee compared population projections for previous years published by SSL to 2015 provisional figures to estimate the variance. This gives an idea of how the provisional estimates varied from the projected. This variance analysis is strongly supported by demographic data obtained by SSL from the 2008 and 2013 Demographic and Health Survey (DHS) and migration statistics as well as economic arguments.

1. Findings

 

1. SLPP Exclusion from the 2015 Census Operations

 

A new Statistician General (SG) was appointed in 2006. Since he was appointed during the SLPP administration, with the advent of the APC in 2007 he felt threatened and relied on staff at SSL who were APC to enable him survive the change of administration. The protection was however short-lived because in 2009 an article was published claiming that SSL had been dominated by people from the South and East Regions since the administration of the first SG in 2000 and that the process was being intensified. An investigation was launched by the Deputy Minister of Finance, while the SG was suspended. He however never returned to work although he was paid up to early 2010 before the end of his contract in 2012. The Deputy SG was then confirmed as SG, without an advertisement for the position and without due process as the SG was not first nominated by the President for approval by Parliament

 

Meanwhile the investigation committee submitted its report and the report was never tabled at the SSL Council so the SG remained suspended. DFID also sent a consultant to assess the capacity of SSL and recommended that the SG be reinstated and the MOFED Minister wrote a letter to SSL Council to the same effect all to no avail

Around the time the confirmation of the new SG was taking place there was no Statistics Council which is supposed to supervise SSL as the previous Council’s term (2009-2012) had expired. During the interregnum the Minister of State for Economic Development in the Ministry of Finance and Economic Development supervised SSL

A new council was appointed in early 2014 and the five nominated members (Chairman and Regional Representatives), were screened and approved by Parliament although the new SG had failed to go through the same vetting process. This council was short-lived as instead of their three year tenure they were hurriedly removed after only six months and replaced by the present SSL Council. No eyebrows were raised by either Parliament or the general public about the hasty replacement of the SSL Council

The weeding out process which had been nurtured since 2008 now took centre stage and witnessed the departure of several seasoned staff members including statisticians and demographers

1. The Census Process

UNFPA has a long standing population and development program with SSL and this provided support to SSL for updating the enumeration maps and general cartographic work even before the census process started in earnest.

The mapping teams were not answerable to the District Statisticians and as was revealed later by the Paramount Chiefs did not update traditional authorities on mapping progress nor did they hold an exit meeting before leaving the chiefdom. Although there were rumours of mapping irregularities these were not investigated or substantiated.

Furthermore District Census Officers (DCOs) were appointed even though there were already District Statisticians in place in the various Districts. The DCOs reported directly to the SG. They received applications for field staff and managed enumeration in their respective Districts

Also a census technical committee was established in January 2014 with the SLPP conspicuously left out. It took several public pronouncements and appeals to the international community for the SLPP representatives to be finally invited in mid-July 2014. SSL claimed that they had been sending invitations but that SLPP failed to respond to the invitations. There is no evidence to support that claim.

By the time SLPP started attending these meetings in mid-July, the field mapping had been completed and the pilot census also completed. These are two significant components of the census process in which SLPP had not participated at all

The importance of inclusion became important during those first meetings as the SLPP now challenged the mapping results and the distribution of EAs and submitted a proposal for cartographic inspection and verification

The Ebola outbreak led to the subsequent postponement of the census from December 2014 to December 2015. This enabled a review of the field maps and visits by the UNFPA mapping consultant to all chiefdoms to get Paramount Chiefs to validate the list of localities that had been mapped. So in fact if Ebola had not struck and the census postponed in September 2014 there would have been no review and the census would have proceeded as diabolically planned

The experience of challenging field results at the technical committee now forced a procedural change in technical committee meetings, to PowerPoint presentations instead of the previous circulation of documents. And then in the final analysis total closure as all communication to the SLPP representative ceased from mid-September 2015. In essence the SLPP did not take part in the following crucial stages of the census

• Design of questionnaire
• Training and selection of mapping field staff
• Updating of field maps
• Training and selection of Pilot census field staff
• Selection of pilot census areas
• Implementation of pilot census
• Evaluation of pilot census
• Training and selection of field staff
• Distribution and return of completed questionnaires
• Field enumeration
• Compilation of provisional results
• Post enumeration survey which assessed over/under count
• Coding and editing of questionnaires
• Data entry for final results

In effect SLPP only took part in the review of the questionnaire after the pilot, cartographic verification and the justification for postponing the census because of Ebola. So apart from the launching of the provisional results which the SLPP was now invited to witness, chances are that the SLPP will again not be invited to participate in the processing and analysis of the census data but will be guaranteed a seat when the final results are launched in December 2016

1. Exclusion of Competent Sierra Leoneans Statisticians/Demographers

Another aspect of the exclusion is the deliberate exclusion from the census technical committee of some Sierra Leoneans with census experience who were present in the country during the entire census process (to name a few):-

• Professor Joseph Kamara  (Former Statistician General 2006 – 2010)
• John Kposowa (Census Mapping and GIS Expert in SSL, and South  Sudan)
• Bartholomew Bockarie (Census Mapping and GIS Expert in SSL, Sudan and Afghanistan)
• Moses Williams (Retired SSL Director of Demographic and Social Statistics)
• Mwaluma Gegbe (former SSL Manager  of National Accounts)
• Philip Amara (former SSL Director of Economic Statistics)

This was a tremendous loss to the nation in terms of census experience. In fact Paul Sengeh (former UNICEF M&E Specialist and Prof. Bob Kandeh (Statistician  General 2002-2005 and Census Advisor in Sudan, Botswana, Gambia, Uganda, Mozambique and Swaziland) only took part as a result of representing the SLPP in the technical committee. Sierra Leone cannot rapidly develop if it continues to marginalise its own talent.

1. Presence of a Census Technical Advisor

The Presence of three UNFPA Census Technical Advisors (Operations, Mapping and Data Processing) should not be bandied about and used as justification for the technical committee not validating the provisional results, as the current best practice is for short term assignments except if SSL is confirming that its staff do not still have the capacity to undertake a census after fifty years. In 2004 there was no need for a long term CTA in operations since the SG was a qualified demographer with survey experience under his belt, but there were short term support missions and long term advisors in cartography and data processing.

1. Professionals and not Politicians

Another exclusion strategy adopted was to further marginalise the SLPP representative on the technical committee with the argument that he is now a politician, thereby ignoring his tremendous wealth of census technical experience. The argument has even been used to justify exclusion from the recruitment and field enumeration. So he cannot even monitor and observe as was done for the census mapping verification.

1. National Census Advisory Committee

This committee is required by law (according to the Statistics Act 2002). This broad based committee provides the forum for consensus building and validating the census process and results before results are announced.

There was very little evidence of its existence and functioning for the 2015 census. It seemed to have been replaced by the census technical committee which became a hodgepodge of everybody instead of statisticians, economists, demographers and people with census and survey experience

One can therefore conclude that there was a deliberate policy at the SSL to exclude certain individuals and the SLPP from the census operations. The SLPP should therefore not accept the provisional results of the 2015 census. Furthermore the SLPP should insist that the results should not be used for electoral boundary delimitation as it is now apparent from all the above that this was the design all along hence the exclusion. SSL should make available to the census technical committee the results of the post enumeration survey so that a technical assessment of census coverage can be undertaken and with the participation of the SLPP this would open the door for a reconsideration of the SLPP’ s position on the 2015 Census Provisional Results.

B. Credibility of Provisional Figures

Comparing projected figures obtained from SSL to provisional figures supported by demographic statistics obtained from SSL and corroborated by economic arguments, it is evident that the provisional results are flawed. Projected population for 2014 obtained from SSL was 6,348,350. Using population annual growth rate of 2.22% between 2010-2015 obtained from the UN Population Division, Department of Economic and Social Affairs, projected population for 2015 was 6,489,283 compared to 2015 provisional figure of 7,075,661 (See Annex 1). This means that the provisional figure exceeds the projected figure by 586,378. Below are the demographic and economic evidence to support the position of SLPP that the provisional census results are wrong.

Mathematical, demographic and economic arguments all support the position of the SLPP that the provisional results obtained from the 2015 Population and Housing census are flawed.

1. Demographic Arguments

 

1. Fertility is Declining

Generally, levels of fertility have declined over the years. Total Fertility Rate (TFR) was 6.3 children per woman in 1985 but declined to 6.1 in 2004. It further declined to 5.1 in 2008 and was 4.9 in 2013 (2008 and 2013 DHS).This clearly confirms decrease in fertility rate over the period 1985 and 2013.

 

1. Mortality 

Overall, mortality in Sierra Leone is considered very high and is among the one of the highest in the world. Although there has been slight decreases infant and under-five mortality over the years, these have not sufficient to warrant a significant increase in population. Below is a table showing trends in both infant and child mortality.

 

As shown in Table 1, between 1999-2003 and 2004-2008, infant mortality decreased by 16.4% and under-five mortality slightly decreased by 14.5% within the same period. Between 2004-2008 and 2009-2013, infant mortality decreased by 27.6% compared to 19.6% drop in under-five mortality within the same period. The decreases in mortality rates are far below the 42% increase in population between 2004 and 2015.

Table 1: Trends in Infant and Child Mortality between 1999 and 2013

 

Source: SSL - 2013 DHS

 

1. Life Expectancy at Birth by Sex

The average number of additional years a person could expect to live if current mortality trends were to continue for the rest of that person's life is commonly cited as Life Expectancy at birth.  Running mortality tables using the indirect techniques, the life expectancy at birth for a Sierra Leonean is 48.4 years. Disaggregated by sex, life expectancy for males is 47.5 years, while that for females is 49.4 years. These figures are considerably higher than the life expectancies in 1985 that were recorded as 37 years for men and 40.3 years for women.

 

1. Migration

Although there is no empirical data on migration, anecdotal evidence suggests that more people migrated out of Sierra Leone than those who migrated into Sierra Leone. There are no economic opportunities attracting other nationals to Sierra Leone. There has been a substantial slowdown in alluvial diamond mining that attracted nationals mostly from Guinea, Mali, Senegal and Gambia. During the war period thousands of Liberia refugees resided in Sierra Leone. Most of these have been resettled to Liberia following the end of the war and return of stability.

Out migration of Sierra Leoneans is experienced every day. Hundreds of Sierra Leoneans travel to neighbouring countries and western countries monthly.

1. Trends in Population Increase and Percentage Changes

The trends in population figures and percentages changes presented in Table 2 below illustrate that the percentage change in population between 2004 and 2015 exceeds percentages for previous inter-censal periods.

 

Table 2: Trends in Percentage Changes in Population

 

 

Between 1963 and 1974 (11 year period), percentage change in population was 25.4% but slightly increased to 28.5% between 1974 and 1985 (11 year period). Between 1985 and 2004 (19 year period), percentage change in population was 41.6% compared to 42.2% between 2004 and 2015 (11 year period). This is sufficient evidence to prove that the 2015 provisional figure is wrong.

Trends in inter-censal growth rate of the population of Sierra Leone between 1963 and 2015 as illustrated in Table 3 below validates the SLPP position that the provisional results are wrong. Since 1963 to 2004, annual growth rate varied from 1.8% to 2.3% per annum. This compares reasonably with estimates of growth for neighbouring West African countries with comparable demographics. According to trends in annual growth rates from the UN Population Division, Department of Economic and Social Affairs, between 2010 and 2015, annual growth rates for Nigeria was 2.67%, Liberia was 2.58%, Ghana 2.39% and Guinea 2.71%. With decline in fertility rate, marginal decline in mortality and no evidence of substantial in-migration should mean reduced growth rate.

Table 3 below presenting trends in annual growth rate in Sierra Leone shows growth rate of 3.2% between 2004 and 2015 which deviates significantly from earlier trends. This 3.2% also significantly exceeds 2.22% annual growth rate for Sierra Leone for the period 2010 and 2015 obtained from the UN Population Division, Department of Economic and Social Affairs.

Table 3: Trends in Population Annual Growth Rate 

1. Economic Arguments

Migration is generally caused by pull and push factors. Poverty is one key factor that cause movement of people from their communities to communities of better economic opportunities or reduced poverty. The provisional results do not support this theory. For instance, it is common knowledge that the bulk of internal migration is towards Western Urban which account for over half of the urban population and has an incidence of poverty estimated at 20.7% compared to national average of 52.9%. Yet, as seen in Annex 1, the provisional census result was less than the 2015 projected population figure by 13,695 for the Western Urban. Also, Bo where the incidence of poverty was 50.7%, provisional census result was less than projected figure by 94,463 compared to Bombali with incidence of poverty of 57.9% where the provisional results exceeded the projected figure by 101,074. In the same vein, Tonkolili with the highest incidence of poverty estimated at 76.4% had a provisional figure higher than the projected figure by 86,183.

(C) Politico 28/06/16