Abstract
This Selected Issues paper and Statistical Appendix addresses the question of how to interpret recent developments in the Kenyan consumer price index (CPI) properly to assess the current inflation pressure and extract signals about possible future CPI inflation trends. The paper discusses why Kenya’s exports have performed poorly over the past five years in spite of a more liberalized trade and exchange rate regime. The analysis shows that Kenya faces both price and nonprice constraints on export performance.
VI. Corruption and Kenyan Economic Performance62
A. Introduction
149. Corruption is the abuse of position or office for private gain. It exists in both public and private sectors. Corruption in Kenya is widely thought to have reached endemic proportions. Examples of corruption include bribery, extortion, nepotism, fraud, and embezzlement. In a nationwide opinion poll of Kenyans published recently63 corruption was cited more often than both poverty and unemployment as the single most important issue facing the country.
150. In recent years, the IMF has increasingly recognized the adverse impact of corruption and poor governance on macroeconomic performance and the success of economic reforms. This section surveys some of the recent research on the economic impact of corruption to illustrate the potential benefits of reducing corruption for growth and poverty reduction in Kenya. This research suggests that pervasive corruption tends to be associated with poorly enforced property rights, a weak rule of law, and low incentives for productive investment. All of this is very damaging to economic growth.
151. A considerable amount of research work in recent years has focused on the links between the level of corruption in a country and its economic growth performance. Among the different aspects of poor governance, corruption has received particular attention since the availability of corruption indices has helped to quantify its extent and allows for international comparisons. Most studies have relied on measures of corruption developed by Business International, International Country Risk Guide, and Transparency International (TI).
B. What Form Does Corruption Take?
152. The Nairobi-based Institute of Economic Affairs (1998) has suggested that corruption can be broken into three types: (i) looting, (ii) grand corruption, and (iii) petty corruption. Looting is the most extreme form of corruption; it involves the kind of scams whose financial consequences are so large that, when they are successfully implemented, they have macroeconomic implications fairly quickly, as banks collapse, inflation rises, or the currency depreciates. Looting is often politically motivated and takes place with the acquiescence of important political players. It can involve the printing of money to fund fictitious projects, or the paying of large sums to individuals who never supply goods and services. In the Kenyan context, a high-profile example of looting was the so-called Goldenberg scam of 1992 which involved making large export subsidy payments for fictitious exports of gold and cost Kenyan taxpayers hundreds of millions of dollars. The very large sums of public money involved in the Goldenberg scam have been a significant factor in focusing the attention of the international donor community on the problem of corruption in Kenya.
153. Grand corruption generally involves senior government officials and significant sums of money. Public procurement may be the area where most grand corruption occurs, and this is why an important element of the Fund- and Bank-supported program in Kenya has been to revise and strengthen procurement regulations and systems. Examples here would include collusion among bidders for government contracts, leading to higher prices; kickbacks by firms to “fix” procurement competition; or bribes to officials who regulate contractors’ behavior. This can encourage abnormally low bids that win the contract, but which are then “rectified” through the acceptance of subsequent cost overruns and project extensions that circumvent the normal competitive bidding procedures. When the government has arrears on domestic or external payments obligations, there is also the opportunity for abuse of funds as these obligations are cleared.
154. Other forms of grand corruption, for example, can involve bribing government officials to obtain licenses, bailing out failing institutions through publicly guaranteed loans or overdraft facilities, applying political pressure not to carry out normal prudential banking supervision, directed lending for political purposes, paying bribes to obtain commercial bank loans with no expectation of making repayments, or granting duty exemptions in return for payoffs. Corruption in the court system can hinder the efforts of banks, creditors and liquidators to enforce contracts, which in turn exacerbates the problem of nonperforming loans, and puts upward pressure on the spread between bank deposit rates and bank lending rates. All these factors fuel inefficiency and undermine macroeconomic performance.
155. Petty corruption, while highly prevalent, typically involves relatively junior officials and minor amounts of money, for example, paying a policeman a few shillings to ignore a traffic violation. This problem has been exacerbated by a police and court system that has in the past been largely ineffective at investigating and prosecuting serious cases of corruption. The recent Transparency International-Kenya Urban Bribery Survey published in January 200264 suggested that the general experience of the Kenyan public is that corruption may be increasing. The survey also suggests that bribery is most prevalent in law enforcement and local authorities (Table 18). Officials at the Ministry of Public Works are reported as demanding the largest bribes, while the police are by far the most frequently bribed officials in Kenya.
Kenya Bribery Index 1/
Six indicators measure the dimensions and impact of bribery: Incidence, Prevalence, Severity, Frequency, Cost, and Bribe size. The first 3 indicators are percentages in the sample. The other 3, which are actual values are scaled by the highest value to achieve an index where the highest value is 100. The overall index is the (unweighted) average value of the 6 indices, with a maximum value of 100.
Survey covered 47 institutions and a further 5 aggregated groups. This table reports the highest 10 scores observed, and the lowest 2.
Kenya Bribery Index 1/
| Institution 2/ | Index | |
|---|---|---|
| 1. | Kenya Police | 68.7 |
| 2. | Ministry of Public Works | 41.0 |
| 3. | Immigration Department | 36.1 |
| 4. | Ministry of Lands | 34.8 |
| 5. | Nairobi City Council | 33.0 |
| 6. | Judiciary | 32.3 |
| 7. | Mombassa Municipal Council | 32.1 |
| 8. | Other Local Authorities | 31.5 |
| 9. | Provincial Administration | 29.5 |
| 10. | Prisons Department | 29.4 |
| 51. | Kenya Wildlife Service | 5.2 |
| 52. | Central Bank of Kenya | 0.2 |
Six indicators measure the dimensions and impact of bribery: Incidence, Prevalence, Severity, Frequency, Cost, and Bribe size. The first 3 indicators are percentages in the sample. The other 3, which are actual values are scaled by the highest value to achieve an index where the highest value is 100. The overall index is the (unweighted) average value of the 6 indices, with a maximum value of 100.
Survey covered 47 institutions and a further 5 aggregated groups. This table reports the highest 10 scores observed, and the lowest 2.
Kenya Bribery Index 1/
| Institution 2/ | Index | |
|---|---|---|
| 1. | Kenya Police | 68.7 |
| 2. | Ministry of Public Works | 41.0 |
| 3. | Immigration Department | 36.1 |
| 4. | Ministry of Lands | 34.8 |
| 5. | Nairobi City Council | 33.0 |
| 6. | Judiciary | 32.3 |
| 7. | Mombassa Municipal Council | 32.1 |
| 8. | Other Local Authorities | 31.5 |
| 9. | Provincial Administration | 29.5 |
| 10. | Prisons Department | 29.4 |
| 51. | Kenya Wildlife Service | 5.2 |
| 52. | Central Bank of Kenya | 0.2 |
Six indicators measure the dimensions and impact of bribery: Incidence, Prevalence, Severity, Frequency, Cost, and Bribe size. The first 3 indicators are percentages in the sample. The other 3, which are actual values are scaled by the highest value to achieve an index where the highest value is 100. The overall index is the (unweighted) average value of the 6 indices, with a maximum value of 100.
Survey covered 47 institutions and a further 5 aggregated groups. This table reports the highest 10 scores observed, and the lowest 2.
156. The results of this survey should be taken as indicative rather than giving precise estimates, but most bribes seem to involve relatively small sums paid very frequently. Based on their sample, TI suggest that many urban Kenyans might make as many as 16 bribe payments in a month.
C. Where Does Kenya Rank Against Other Countries?
157. It is very difficult to quantify corruption precisely and inter-country comparisons are notoriously unreliable. Table 19 shows that in the 2001 TI survey of corruption perceptions, Kenya ranked joint 84th out of the 91 countries surveyed.
Selected Countries: 2001 Corruption Perceptions Index 1/
Highest two scores and sub-Saharan Africa.
Selected Countries: 2001 Corruption Perceptions Index 1/
| Country | Country | 2001 | Surveys | Standard | High- Low |
|---|---|---|---|---|---|
| Rank | Score | Used | Deviation | Range | |
| 1 | Finland | 9.9 | 7 | 0.6 | 9.2-10.0 |
| 2 | Denmark | 9.5 | 7 | 0.7 | 8.8-10.0 |
| 26 | Botswana | 6.0 | 3 | 0.5 | 5.6 - 6.6 |
| 30 | Namibia | 5.4 | 3 | 1.4 | 3.8-6.7 |
| 38 | South Africa | 4.8 | 10 | 0.7 | 3.8-5.6 |
| 40 | Mauritius | 4.5 | 5 | 0.7 | 3.9-5.6 |
| 59 | Ghana | 3.4 | 3 | 0.5 | 2.9-3.8 |
| 61 | Malawi | 3.2 | 3 | 1.0 | 2.0-3.9 |
| 65 | Senegal | 2.9 | 3 | 0.8 | 22 - 3.8 |
| 65 | Zimbabwe | 2.9 | 6 | 1.1 | 1.6-4.7 |
| 75 | Zambia | 2.6 | 3 | 0.5 | 2.0-3.0 |
| 77 | Côte d’Ivoire | 2.4 | 3 | 1.0 | 1.5-3.6 |
| 82 | Tanzania | 2.2 | 3 | 0.6 | 1.6-2.9 |
| 84 | Cameroon | 2.0 | 3 | 0.8 | 1.2-2.9 |
| 84 | Kenya | 2.0 | 4 | 0.7 | 0.9 - 2.6 |
| 88 | Uganda | 1.9 | 3 | 0.6 | 1.3-2.4 |
| 90 | Nigeria | 1.0 | 4 | 0.9 | 0.0-2.0 |
Highest two scores and sub-Saharan Africa.
Selected Countries: 2001 Corruption Perceptions Index 1/
| Country | Country | 2001 | Surveys | Standard | High- Low |
|---|---|---|---|---|---|
| Rank | Score | Used | Deviation | Range | |
| 1 | Finland | 9.9 | 7 | 0.6 | 9.2-10.0 |
| 2 | Denmark | 9.5 | 7 | 0.7 | 8.8-10.0 |
| 26 | Botswana | 6.0 | 3 | 0.5 | 5.6 - 6.6 |
| 30 | Namibia | 5.4 | 3 | 1.4 | 3.8-6.7 |
| 38 | South Africa | 4.8 | 10 | 0.7 | 3.8-5.6 |
| 40 | Mauritius | 4.5 | 5 | 0.7 | 3.9-5.6 |
| 59 | Ghana | 3.4 | 3 | 0.5 | 2.9-3.8 |
| 61 | Malawi | 3.2 | 3 | 1.0 | 2.0-3.9 |
| 65 | Senegal | 2.9 | 3 | 0.8 | 22 - 3.8 |
| 65 | Zimbabwe | 2.9 | 6 | 1.1 | 1.6-4.7 |
| 75 | Zambia | 2.6 | 3 | 0.5 | 2.0-3.0 |
| 77 | Côte d’Ivoire | 2.4 | 3 | 1.0 | 1.5-3.6 |
| 82 | Tanzania | 2.2 | 3 | 0.6 | 1.6-2.9 |
| 84 | Cameroon | 2.0 | 3 | 0.8 | 1.2-2.9 |
| 84 | Kenya | 2.0 | 4 | 0.7 | 0.9 - 2.6 |
| 88 | Uganda | 1.9 | 3 | 0.6 | 1.3-2.4 |
| 90 | Nigeria | 1.0 | 4 | 0.9 | 0.0-2.0 |
Highest two scores and sub-Saharan Africa.
158. Since the TI survey was first published in 1996, Kenya’s highest rating was 73rd out of the 85 countries surveyed in 1998.65 In the years since 1998 it has been in the bottom decile of the surveyed countries. TI’s corruption perceptions index score reflects the degree of corruption as observed by business people, academics, and risk analysts, and ranges from 10 (highly clean) to 0 (highly corrupt). In 2001, some of the richest countries in the world—Finland, Denmark, New Zealand, Iceland, Singapore and Sweden scored 9 or higher out of a possible clean score of 10, indicating very low levels of perceived corruption in government and public administration. The eight countries with a score of 2 or less were Azerbaijan, Bolivia, Cameroon, Kenya, Indonesia, Uganda, Nigeria and Bangladesh. Among the sub-Saharan Africa countries covered, Kenya was ranked above Uganda and Nigeria, at the same level as Cameroon, but below the eleven other countries surveyed. While the degree of precision of these surveys would not be sufficient to argue that corruption perceptions in Kenya are significantly different from those in Tanzania, Uganda or Cameroon, there is more of a consensus that corruption is becoming more prevalent in Kenya.
D. The Impact of Corruption on Economic Performance
159. Recent empirical research by Fund staff has highlighted the impact of corruption on growth, public finances, poverty, income inequality, and the provision of social services. Table 20 summarizes the results. A number of studies have explicitly identified the negative impact of corruption on economic growth. Mauro (1996) finds that increasing corruption by one unit (on a scale from 1 to 10) would lower real per capita GDP growth by 0.3 to 1.8 percentage points, while Leite and Weidmann (1999) and Abed and Davoodi (2000) have reported a somewhat narrower range also centered on about 1 percent.
Impact of Increasing Corruption by One Unit 1/
Corruption is measured on a scale of 0 (highly clean) to 10 (highly corrupt).
Impact of Increasing Corruption by One Unit 1/
| Author(s) | Impact on | Finding |
| Mauro (1996) | Real per capita GDP growth | −0.3 to -1.8 percentage points |
| Leite and Weidmann (1999) | Real per capita GDP growth | −0.7 to -1.2 percentage points |
| Tanzi and Davoodi (2000) | Real per capita GDP growth | −0.6 percentage points |
| Abed and Davoodi (2000) | Real per capita GDP growth | −1 to -1.3 percentage points |
| Mauro (1996) | Ratio of investment to GDP | −1 to -2.8 percentage points |
| Mauro (1998) | Ratio of public education | −0.7 to -0.9 percentage points |
| spending to GDP | ||
| Mauro (1998) | Ratio of public health | −0.6 to -1.7 percentage points |
| spending to GDP | ||
| Gupta, Davoodi and Alonso-Terme (1998) | Income inequality (Gini coefficient) | +0.9 to+2.1 Gini points |
| Gupta, Davoodi and Alonso-Terme (1998) | Income growth of the poor | −2 to -10 percentage points |
| Ghura(1998) | Ratio of tax revenues to GDP | −1 to-2.9 percentage points |
| Tanzi and Davoodi (2000) | Measures of government revenues to GDP ratio | −0.1 to -4.5 percentage points |
| Gupta, de Mello and Sharan (2000) | Ratio of military spending | + 1 percentage point |
| Gupta, Davoodi and Tiongson (2000) | Child mortality rate | + 1.1 to 2.7 deaths per 1000 live births |
| Gupta, Davoodi and Tiongson (2000) | Primary student dropout rate | +1.4 to 4.8 percentage points |
| Ratio of public investment to | ||
| Tanzi and Davoodi (1997) | GDP | +0.5 percentage points |
| Tanzi and Davoodi(1997) | Percent of paved roads in good condition | −2.2 to -3.9 percentage points |
Corruption is measured on a scale of 0 (highly clean) to 10 (highly corrupt).
Impact of Increasing Corruption by One Unit 1/
| Author(s) | Impact on | Finding |
| Mauro (1996) | Real per capita GDP growth | −0.3 to -1.8 percentage points |
| Leite and Weidmann (1999) | Real per capita GDP growth | −0.7 to -1.2 percentage points |
| Tanzi and Davoodi (2000) | Real per capita GDP growth | −0.6 percentage points |
| Abed and Davoodi (2000) | Real per capita GDP growth | −1 to -1.3 percentage points |
| Mauro (1996) | Ratio of investment to GDP | −1 to -2.8 percentage points |
| Mauro (1998) | Ratio of public education | −0.7 to -0.9 percentage points |
| spending to GDP | ||
| Mauro (1998) | Ratio of public health | −0.6 to -1.7 percentage points |
| spending to GDP | ||
| Gupta, Davoodi and Alonso-Terme (1998) | Income inequality (Gini coefficient) | +0.9 to+2.1 Gini points |
| Gupta, Davoodi and Alonso-Terme (1998) | Income growth of the poor | −2 to -10 percentage points |
| Ghura(1998) | Ratio of tax revenues to GDP | −1 to-2.9 percentage points |
| Tanzi and Davoodi (2000) | Measures of government revenues to GDP ratio | −0.1 to -4.5 percentage points |
| Gupta, de Mello and Sharan (2000) | Ratio of military spending | + 1 percentage point |
| Gupta, Davoodi and Tiongson (2000) | Child mortality rate | + 1.1 to 2.7 deaths per 1000 live births |
| Gupta, Davoodi and Tiongson (2000) | Primary student dropout rate | +1.4 to 4.8 percentage points |
| Ratio of public investment to | ||
| Tanzi and Davoodi (1997) | GDP | +0.5 percentage points |
| Tanzi and Davoodi(1997) | Percent of paved roads in good condition | −2.2 to -3.9 percentage points |
Corruption is measured on a scale of 0 (highly clean) to 10 (highly corrupt).
160. In these studies corruption was shown to lower growth by reducing private investment, attracting talented individuals into unproductive activities, and encouraging poor management of natural resources. Abed and Davoodi argue that structural reforms aimed at rationalizing the role of the state, increasing reliance on market-based pricing and creating a sound regulatory environment should contribute to growth directly and indirectly by lowering the incidence of corruption, and provide supporting evidence from the former Soviet Union and Eastern Europe.
161. A number of other studies have also found that corruption distorts the composition of public expenditures in favor of sectors where bribes are easier to collect. Corruption typically shifts spending away from routine maintenance and repair, education and health to excessive and inefficient public investments and higher military spending. Gupta, Davoodi and Tiongson (2000) find that higher corruption has adverse consequences for social indicators such as child mortality rates, and student dropout rates. Ghura (1998) suggests that corruption in the form of abuse of public funds and resulting in weak social indicators also weakens revenues because it contributes to tax evasion, improper tax exemptions, and weak tax administration.
162. Real growth rates in Kenya have been relatively weak and declining for the past decade, and this in large part is a reflection of weak and deteriorating governance. Growth of real GDP per capita has on average been minus 0.3 percent per year from 1992–2001, leading to a declining level of real incomes and increased poverty. Based on the lower bound of the estimates outlined in Table 20 below, one might suggest that raising the corruption perceptions index from 2 to 4, that is, to the level of corruption thought to exist in Ghana or Mauritius, might help to raise GDP per capita growth by at least 1 percent and perhaps significantly more. Improving the perceived level of corruption to the level of South Africa might increase GDP per capita growth by 1 and a half percent or more. Figure 14 illustrates the potential implications for income.
Kenya: Real GDP per Capita, 1991–2001
(U.S. dollars)
Citation: IMF Staff Country Reports 2002, 084; 10.5089/9781451821062.002.A006
Kenya: Real GDP per Capita, 1991–2001
(U.S. dollars)
Citation: IMF Staff Country Reports 2002, 084; 10.5089/9781451821062.002.A006
Kenya: Real GDP per Capita, 1991–2001
(U.S. dollars)
Citation: IMF Staff Country Reports 2002, 084; 10.5089/9781451821062.002.A006
163. Kenyan GDP per capita in 1991 is estimated to have been about US$333. The lower line in Figure 14 illustrates the level of GDP per capita implied by the actual growth rates of GDP per capita achieved over the past decade and is thus an estimate of GDP per capita in 1991 prices. The trend has been down, implying falling income and increased poverty levels. By 2001, the estimate of real GDP per capita in 1991 prices had fallen to US$320. The middle line illustrates what would have been the path real income, if, ceteris paribus, corruption was one unit lower during the last decade, and therefore the growth of real income per capita been 1 percent higher. It would have implied a slight increase in real per capita income reaching about US$350. In order to achieve significantly higher incomes, however, the chart suggests that growth rates at least 2 percent higher would have been needed. Clearly, this requires a significant improvement in the actual and perceived level of corruption in Kenya. Raising the growth rate of GDP per capita from, for example, 1 percent to 3 percent would reduce the number of years required to double per capita income from about 70 to about 25.
E. AntiCorruption Strategy
164. The Kenyan government has recognized that a vigorous anti-corruption campaign is required to generate the conditions for stronger growth of per capita income in Kenya. Successful anti-corruption strategies have typically been predicated on there being a real and effective deterrent in place to curb the individual’s instinct to abuse his or her public office for personal gain. It is widely accepted that, to be effective, public officials need to believe that if they abuse their office, there is a substantial prospect that they will be caught, convicted, and punished. Moreover, Klitgaard, Maclean-Abaroa, and Parris (2000) point out that experience indicates that prosecuting serious high-profile corrupt actors is “an essential element of an anti-corruption strategy so that a cynical citizenry believes that an anti-corruption drive is more than just words.” This needs to be supplemented with an ongoing education program that informs the public about the negative implications of corruption, and with a system that allows the public to report acts of corruption without fear of retaliation. It is also critically important to liberalize and reform institutions and practices to reduce the opportunities for rent seeking and corruption, and to significantly strengthen public audit functions.
165. Since the demise of the Kenya Anti-Corruption Authority (KACA), the Kenyan authorities have developed an new anti-corruption strategy (see staff report Box 3). The authorities’ plan involves the introduction and enforcement of codes of ethics for all public officials, including the civil service, the judiciary, and the legislature. Public officers would be required to file annual declarations of all their assets and liabilities. The legal status of the Anti-Corruption Police Unit would be strengthened to ensure that all the elements of the structure of the former KACA remain active and operationally autonomous and to ensure that there are appropriate safeguards against political interference. The Attorney General’s office would be strengthened to ensure timely action on all pending corruption cases, and measures will be taken to ensure that the Attorney General’s actions and decisions relating to the prosecution of corruption cases are held publicly accountable. Special courts are to be established to deal with corruption cases on a fast-track basis, and reputable judges will be appointed specifically to these courts, and the authorities plan to suspend all public officials whose cases are brought to trial, irrespective of their seniority or position.
166. These actions will have to be reinforced by substantially strengthening other institutions whose operations have a significant bearing on corruption. These include, in particular, the office of the Controller and Auditor General, the internal audit, the procurement system, and the judiciary (see staff report para.20). The anti-corruption measures need to be complemented by other structural reforms, including privatization, liberalization of the coffee sector, and civil service reforms that make abuse of public funds and other rent-seeking activity more difficult. If implemented vigorously, and if major cases of corruption are seen by the public to be effectively prosecuted this governance plan has the potential to start changing attitudes towards corruption in Kenya. The evidence suggests that this could in turn have significantly beneficial effects on economic growth and the level of poverty.
References
Abed, George, and Hamid Davoodi, 2000, “Corruption, Structural Reforms, and Economic Performance in the Transition Economies,” IMF Working Paper 00/132 (Washington: International Monetary Fund).
Ghura, Daneshwar, 1998, “Tax Revenue in Sub-Saharan Africa: Effects of Economic Policies and Corruption,” IMF Working Paper 98/135 (Washington: International Monetary Fund).
Gupta, Sanjeev, Hamid Davoodi, and Rosa Alonso-Terme, 1998, “Does Corruption Affect Income Inequality and Poverty?” IMF Working Paper 98/76 (Washington: International Monetary Fund).
Gupta, Sanjeev, Luiz de Mello, and Raju Sharan, 2000, “Corruption and Military Spending,” IMF Working Paper 00/23 (Washington: International Monetary Fund).
Gupta, Sanjeev, Hamid Davoodi, and Erwin Tiongson, 2000, “Corruption and the Provision of Health Care and Education Services,” IMF Working Paper 00/116 (Washington: International Monetary Fund).
Institute of Economic Affairs, 1998, Our Problems, Our Solutions: An Economic and Public Policy Agenda for Kenya, (Nairobi, Kenya: Institute of Economic Affairs).
Klitgaard, Robert, Ronald Maclean-Abaroa, and Lindsey Parris, 2000, Corrupt Cities, A Practical Guide to Cure and Prevention (Oakland, California: ICS Press).
Leite, Carlos, and Jens Weidmann, 1999, “Does Mother Nature Corrupt? Natural Resources, Corruption, and Economic Growth,” IMF Working Paper 99/85 (Washington: International Monetary Fund).
Mauro, Paolo, 1996, “The Effects of Corruption on Growth, Investment, and Government,” IMF Working paper 96/98 (Washington: International Monetary Fund).
Mauro, Paolo, “Corruption and Composition of Government Expenditure,” Journal of Public Economics, vol. 69 (August), pp. 263 –79.
Tanzi, Vito, and Hamid Davoodi, 1997, “Corruption, Public Investment, and Growth,” IMF Working Paper 97/139 (Washington: International Monetary Fund).
_____, 2000, “Corruption, Growth, and Public Finances,” IMF Working Paper 00/182 (Washington: International Monetary Fund).
Transparency International, 2001, Global Corruption Report (Berlin: Transparency International).
Transparency International-Kenya, 2002, The Kenya Urban Bribery Index (Nairobi: Transparency International - Kenya).
Prepared by Robert Powell.
International Republican Institute, October 2001.
Survey of 1,164 individuals conducted in March and April 2001. The survey was a pilot study and the sample was not sufficiently representative to allow for firm conclusions about the urban population as a whole.
Kenya was not surveyed in 1997.
