OUTSIDE ASSESSMENTS: MEASURING CORRUPTION INCIDENCE IN GHANA. INWENT ALUMINI EXPERIENCE WITH GOVERNANCE ASSESSMENT

1
OUTSIDE ASSESSMENTSMEASURING CORRUPTION
INCIDENCE IN GHANA.INWENT ALUMINI EXPERIENCE
WITH GOVERNANCE ASSESSMENT USING INDICATORS TO
INCREASE POLITICAL ACCOUNTABILITY, INFORM POLICY
PROCESS AND POLITICAL DEBATE

• GOVERNANCE ASSESSMENT WORKSHOP,
• SAFARI COURT HOTEL, WINDHOEK, NAMIBIA
• 2-5 NOVEMBER, 2009
• UNDP, INWENT NID
• JOHN YAW AMANKRAH
• LABOUR STATISTICIAN AND DEVELOPMENT PRACTITIONER,
  GHANA
• amankrahj_at_yahoo.co.uk.

2
OUTLINE

• INTRODUCTION BACKGROUND
• METHODOLOGY
• MAIN FINDINGS
• LESSONS LEARNED, CONCLUSIONS AND RECOMMENDATIONS.

2
3
1.0 INTRODUCTION BACKGROUND

• Presentation provides findings of a study on
  corruption incidence in Ghana, conducted in 2008
  for the former Office of Accountability, of
  former President Kufuors government following a
  course in Applied Statistics in Governance and
  Human Rights from the InWent Institute at Bonn,
  Germany in 2007.
• Using insights gained from the 2007 InWent
  training in governance assessment and recognising
  the methodological challenges in measuring
  perception of corruption, this study was
  essentially an attempt to assess the level of
  knowledge, magnitude and involvement of
  corruption at the household and enterprise levels
  in Ghana.
• It was the first time such a comprehensive study
  on corruption has been undertaken in Ghana since
  Ghana joined other countries to introduce various
  initiatives to address the corruption challenge
  in Ghana which is one of the key commitments
  adopted by the AU, NEPAD, APRM.
• The objective of the study was to provide
  baseline indicators on corruption in Ghana
• following increased political focus given to
  addressing corruption
• to inform The Office of The President in tackling
  the issue of corruption in Ghana
• to monitor progress of poverty reduction and
  government s accountability and probity agenda.

4
2.0 METHODOLOGY- Scope and Coverage

• The study was done at two levels at both the
  household and enterprise levels and across
  sectors nationwide-.
• Household probability sample survey of Ghanaian
  households who have knowledge and experience of
  corruption through their dealings with officials
  in both public and private sectors.
• Establishment-based survey of a random sample of
  officials in the public and private sector
  institutions as well as media houses in urban
  centres who interact mostly with people from both
  the public and private sectors in their working
  life. These officials include top management as
  well as middle and low-level officials working in
  Ministries, Departments and Agencies who are
  administrators, secretaries, accountants,
  auditors and procurement officers. 
• The time frame considered for the study was a
  period of 12 months preceding the study in 2008.
  The study was done in 103 sampled communities
  across all the ten regions of Ghana while a
  sample of 230 establishments was also canvassed
  for the study.
• Data was collected from a total of 2,502
  households respondents and 193 officials working
  in 85 establishments in both public and private
  sector as well as media houses in the country
  were can.

5
METHODOLOGY (CONTD.)-Sample Design

• The sample used in the study was a
  household-based probability twostage stratified
  sample covering the entire four ecological zones
  of the country.
• A two-stage sample design was adopted.The first
  stage of sampling involved constructing a frame
  for the selection of communities with a
  population of not less than 200 people which were
  stratified across the entire country. In all, a
  total of 103 communities were selected
  proportional to the total population in each
  region using systematic probability proportional
  to size selection.
• For the second stage, a listing exercise was
  undertaken in all the 103 selected communities to
  enable two adult members in each of the selected
  1545 households to be selected for the study.
• In addition, 131 establishments and 39 media
  houses located in the urban centres in Ghana were
  randomly selected using the systematic sampling
  method. In each of the establishments two
  officials in the following five categories of
  positions were targeted for the study (1) top
  management, (2) accountant, (3) auditor, (4)
  procurement and (5) Others comprising clerks,
  drivers, messengers, typists and secretaries.
• Note Ghana was ranked 69 by Transparency
  International s 2007 Corruption Perception Index
  (CPI 2007)on a list of 180 countries with a score
  of 3.7 out of maximum score of 10.

6
METHODOLOGY(CONTD.)-Survey Instruments

• Two types of instruments were used to collect
  information using a direct interview technique
  supplemented by the use of key informants survey
  to check for validity, consistency and
  reliability. These instruments were household and
  enterprise questionnaires.
• The questionnaires were designed to provide
  ordinal scores for corruption and also to provide
  cardinal scores of corruption. Ordinal scores
  were used to measure the level of perception at
  both the household and enterprise levels.
  Cardinal scores were used to measure both the
  direction, magnitude of participation and
  involvement of corruption as well as knowledge of
  corruption at both the household and enterprise
  levels.
• The study also used a variety of statistical
  tools and analysis to measure
• (1) the direction and magnitude of participation
  and involvement of the respondents in corruption,
  and
• (2) and knowledge of corruption.
• Survey data on these variables were analysed from
  2,502 households respondents across the ten
  regions and the four ecological zones of Ghana
  coastal, forest, middle belt and savannah.
• At the workplace level, the study also analysed
  survey data from 193 officials working in 85
  establishments in both public and private sector
  as well as media houses in the country.

7
3.0 MAIN FINDINGS

• Level of perception of corruption
• About one out of every two of Ghanaian households
  respondents (52) as against about one out of
  every three respondents in workplaces (31.1)
  think that the level of corruption is high i.e.
  over 70 percent.
• 24.2 of the youth aged 18-35 think that the
  level of corruption in Ghana is high i.e above
  70
• More men (26.3) than women (25.7) think that
  the level of corruption is high.
• There is a strong inverse relationship between
  how people in households think about the level of
  corruption and the level of education. The lower
  the level of education the higher the level of
  corruption is thought to be.
• About four out of every five respondents who
  rate the level of corruption to be high have not
  been educated beyond the secondary school level.
• Corruption is rated to be high by more people in
  the private sector than those in the public
  sector while more people in the public sector
  than those in the private sector think that the
  level of corruption is low.

8
MIAN FINDINGS (CONTD)

• Participation and involvement in corruption
• About one out of every four household member
  (23.6) was bribed by a Politician, a Private
  individual, Public Servant, Civil Servant and
  Others during the past twelve months while a
  similarly proportion (23.9) bribed or attempted
  to bribe a Politician, a Private individual, a
  Public Servant, and a Civil servant during the
  same period.
• Public Servants who are known to demand bribe
  most are The Police, Customs, Internal Revenue
  Service, The VAT Service, Accountant-General,
  Immigration and Pensions Officials and other
  Public Servants involved in national sports
  issues, especially in selecting national teams.
• About one out of every two respondents knows that
  officials from Education and Health demand bribe.
  Staffs from second cycle educational institutions
  also demand bribe and are also bribed more
• The proportion of accountants who are bribed by
  officers is about five times higher than that of
  auditors. Auditors are the least bribed Officials
  among Civil and Public servants.
• The region with the highest participation and
  involvement of corruption was Greater Accra
  (28.3) while Western Region had the least
  participation and involvement in corruption(1.2)
  

9
MAIN FINDINGS (CONTD.)

• Knowledge of corruption
• 27.2 of household respondents know an official
  who bribed someone, 31.2 know an official who
  was bribed by someone while 28.7 know an
  official who demanded bribe from someone
• 95.4 of household respondents know that
  officials from The Police Service demanded bribe
  from someone while 39.9 know that Civil Servants
  were bribed by someone.
• 72.4 of household respondents know that
  Judges/Magistrates and other Officials from The
  Judicial Service were bribed by someone
• 39.2 of household respondents know that
  Officials involved in education delivery
  including officials in tertiary educational
  institutions who demanded bribe from someone
  while 35.6 know that Officials involved in
  health delivery who demanded bribe from someone.
• 15.6 of workplace officials know Directors who
  bribed or attempted to bribe.
• 13.4 of Officials know officials in workplaces
  connected with land matters who bribed or
  attempted to bribe someone.
• 22.6 of typists, clerks, secretaries know the
  types of people who bribe officials in
  workplaces. 25.8 of officials in public
  institutions with 15.6 of them being in top
  management know private individuals who bribed or
  attempted to bribe officials in workplaces

10
MAIN FINDINGS(CONTD.)

• Evidence for knowledge of corrupt practices
• 44.9 of evidence available by household
  respondents on knowledge of officials who were
  bribed by someone can be substantiated while
  35.2 of evidence were media evidence.
• 39 of evidence available by household
  respondents and 46.9 of evidence available at
  the workplace on knowledge of officials who
  demanded bribe from someone cannot be
  substantiated.
• 18.4 of evidence available at the workplace on
  knowledge of officials who were bribed by someone
  were based on media evidence
• Substantive evidence constitutes 74.1 of
  evidence on officials who demanded bribe while
  47 of evidence available on officials who bribe
  as well as officers who were bribed can be
  substantiated. About 2 out of every 5 persons who
  have knowledge about corrupt officials have
  substantive evidence to prove it.

11
4.0 LESSONS LEARNED,CONCLUSIONS RECOMMENDATIONS

• From the evidence outlined in this study it is
  clear that Ghana faces a formidable task in
  addressing the corruption challenge in Ghana.
  Some key themes emerge from the study which are
  of fundamental importance in addressing the
  corruption challenge in Ghana. Many aspects of
  this challenge will require new policy options.
• Firstly, the issue of corruption is not
  well-understood by the general population and
  that corruption needs to be properly defined. It
  became evident from study that most Ghanaians who
  perceived public officials to be corrupt based
  their perception on acquisition of material
  wealth. Though this may have some relationship
  with corruption, this relationship is weak unless
  it is further related to sources of income.
•  Secondly, it became evident from the study that
  household respondents and officials working in
  media houses willingly want to talk about
  corruption than officials in working in public
  establishments. However, at the workplace level
  junior officers such as drivers, cleaners,
  messengers, secretaries were willing to talk more
  about corruption than senior officials.
•  Thirdly, bribery and corruption is high among
  officials in education and health delivery and
  this has grave implications in addressing poverty
  and also achieving our MDGs.
•  

12
LESSONS LEARNED (CONTD)

• The study also found out that Political Party
  Faithfuls and Political Party Foot Soldiers are
  not likely to be trusted and that about one out
  of every five of these types of officials are
  most likely to be bribed.
•  
• The study found out that the high level of
  corruption was mostly media allegations which
  cannot be substantiated. These media allegations
  should be investigated officially. When the media
  allegations are not investigated officially, it
  continues to keep the level of corruption to be
  high.
•  
• One other key finding from the study is that
  there is indeed some substantive evidence
  available in the public domain and that it is
  recommended that action should be taken by the
  Government to follow up on the evidence available
  in the public domain that can be substantiated.
  As the study shows it is comforting to note that
  Auditors are the types of officials who are least
  prone to corrupt practices and therefore the most
  likely officials to be trusted upon in addressing
  the corruption challenge.
•  

13
LESSONS LEARNED (CONTD)

• One of the key major methodological challenges
  that faced the Ghanaian experience in measuring
  corruption was non-response which was very high
  in workplaces than at the household levels. This
  necessitated a lot of call backs and follow-up
  actions to be taken with its implications on time
  and cost. Thus measuring corruption in Ghana is
  indeed a formidable task for Ghanaian development
  practitioners.
•  
• Again, the other major methodological challenge
  in the Ghanaian experience of measuring
  corruption is that peoples perception of
  corruption seemed to be politically motivated and
  that people dont want to talk about it with its
  implications on the error margin in measuring
  corruption.
•  
• Finally, there are various aspects of corruption
  in Ghana beyond the scope of this present study
  and that building capacity for Govt and non-Govt
  stakeholders to conduct Governance Assessments
  should be intensified to monitor national
  development plans.

14
1