Ghana: Poverty and Social Impact Analysis (PSIA) Electricity Tariffs, June 2010 Recently, the Public Utilities Regulatory Commission (PURC) of Ghana increased electricity tariffs. The residential tariffs rose by an average of 27%. As the majority of the poor still do not have access to electricity, the poverty implications of these new tariffs are expected to be minimal. Simulations show that overall this price increase has little impact on the poor, particularly the extremely poor. However, the most conservative estimations find that lower middle-income consumers could be negatively affected by these higher electricity prices. This is because many of these customers consume between 51-150 kwh/month and can no longer take advantage of the electricity subsidies under the lifeline tariff program, which now only applies to those consuming 0-50kWh/month. As a result, electricity prices for these users have increased by more than 42%, which is supposed to be the highest percentage increase of tariffs for residential customers. To mitigate the impact of the current prices, the lifeline tariff program should continue to subsidize users within the 51-150kWh consumption bracket. Further, a sensitivity analysis finds that tariffs, on average, could be increased even further with minimal impact on the poor if prices are increased progressively for those consuming more than 150kWh/month. Lifeline Tariff 1 Ghana has a progressive tariff structure for residential users based on levels of consumption.(see Table 1) Those consuming between 0-50kWh/month pay the lowest tariff, which is also known as the lifeline tariff. Originally, the lifeline was a flat rate applied to the first 50 kwh of consumption per customer and served as the utilities minimum monthly service charge. In 2002, the lifeline tariff became a social benefit when the Government introduced a subsidy for each user consuming only in this lowest consumption bracket. This lifeline subsidy is intended to protect low income users from tariff increases. The subsidy was increased in 2003 to ensure a flat rate of 13000 cedis/month (or 1.3 GH ) for consumers in the 0-50kWh category. After the November 2007 tariff increase, the Government of Ghana was worried that certain consumers would be unable to afford electricity at this new rate. So, the lifeline consumption threshold was increased to include customers in the 51-150kWh consumption bracket. The Government reasoned that this change was to cater for the majority of Ghanaian workers nurses, teachers and civil servants 2 whose electricity consumption generally falls in the 0-150kWh range. In the most recent PURC press release, the lifeline tariff threshold was reduced to 50kWh. Effectiveness of the Lifeline Tariff 3 Previous PSIAs concluded that the lifeline tariff is an imperfect mechanism for targeting the poor for two reasons: 1) the majority of the poor do not have access to electricity(see Table 2), and 2) because of the tariff design, a significant number of lifeline customers are non-poor. Interestingly, the households benefitting from the lifeline are not statistically different, in terms of vulnerability, from households that do not receive the lifeline. The imperfect reach of the lifeline is further complicated by the fact that many that could do not take 1 This section draws heavily from the Ghana PSIAs, 2004 and 2005. 2 Ghana Web, 2007 3 This section draws heavily from the Ghana PSIAs, 2004 and 2005 as well as the report by Wodon, 2009.
advantage of the subsidy. In general, there is a lack of understanding of how the lifeline works or of how to conserve energy to meet the lifeline KWh limit, particularly in areas with low literacy levels. Wodon (2009) presents a study on the targeting of social programs in Ghana that also concludes that the electricity consumption subsidies mostly benefit the non-poor. He estimates that only 8% of the subsidies reach the poor. A suggested alternative to this lifeline scheme is to implement connection subsidies. This subsidy scheme would provide free connections to the electricity network for households that live in areas that have access to electricity, but are not connects to the network. Such a program is estimated to increase the share of outlays benefiting the poor to 29%. Even though the lifeline is deficient, there is little incentive to reform the tariff structure in the short-term. According to the 2004 PSIA, the cost of a better targeting mechanism outweighed the savings that it would generate. 4 While the costs of these lifeline subsidies are substantially higher at present day, there is little evidence to support a complete overhaul of the current targeting system. Further, the reach of this program will increase as access to electricity expands to new customers in rural areas that are likely to consume less than 50KWh. More importantly, the effectiveness of the lifeline can be significantly improved through consumer education campaigns, particularly programs that cater to less literate consumers. The cost and administrative burden of increasing consumer understanding of the current lifeline tariff program will be substantially lower than implementing a lifeline that better targets the poor. In addition, there are concerns that the poor do not receive the lifeline subsidy because they live in compound houses with shared meters. In such a living situation, there is little motivation to conserve electricity. Unless the total consumption of all the households in the compound is less than 50KWh, there is no way to benefit from the lifeline. The PSIAs dispel these concerns as they conclude that compound house residents are more likely to be in a high income bracket and that these households would not necessarily be better off with individual meters. Recent Tariff Increase As mentioned earlier, the tariff increases have no effect on the extremely poor. As shown in Table 3, there is no change in the poverty levels of users in the 10 th and 20 th income percentiles, or the extremely poor. Conversely, under conservative simulations, consumers whose income is just slightly above the poverty line, or those at the lower middle-income level, might be adversely affected.(see Table 3 Scenario 1) These calculations do not factor in income growth or substitution to other energy sources and, therefore, present a worst-case scenario. In this case, poverty levels for those in the 30 th and 40 th income percentiles increase by 0.7 and 2.7 percentage points respectively and the overall poverty rate only increases by 0.3 percentage points. These income deciles are the most affected by the new tariffs because the majority of the electricity users in these income brackets are in the 51-150kWh consumption bracket. Because the lifeline threshold drops to 50kWh in the new tariff regulation, 51-150kWh consumers face a steep price increase. 5 (See Table 1) The tariff 4 The PSIA estimated the lifeline subsidy cost to the Government of Ghana was less than 1% of the utility s revenues in 2004. 5 The June 2010 tariff increase is yet to be fully implemented. At present, only prepayment customers have seen a change in their electricity rates. The Government of Ghana will likely to subsidize consumers in the 51-150kWh consumption
for this group rose from 9.50 Ghp to 17.00 Ghp, which translates into a 79% increase. 6 This price increase is substantially more than 42%, which according to the PURC is the highest percentage increase for residential users. Once income growth is considered (Table 3 Scenario 2) 7, the increase in tariffs has no impact on the poor at all. Rather, poverty rates drop in the face of higher tariffs. Overall, one can conclude that the new tariffs have little impact on the poor. Tariff Sensitivity Analysis Despite the recent tariff hike in June, the utility companies will still face a financial gap of GH 49 million. 8 The PURC believes that the funds needed to cover all the utilities expenses cannot be raised purely through increased consumer prices. Nevertheless, one wonders if this deficit can be mitigated somewhat by higher tariffs. So the question is: can residential customers afford to pay even more for electricity? What is the impact of tariff increases that are higher than the June 2010 increase of 27%? The sensitivity analysis seeks to answer these questions by simulating the poverty implications of average tariff increases of 35%, 40%, 45% and 50% instead of the June 2010 increases. The results of this analysis are included in Tables 4. In these tables, three different shocks are applied: 1) Shock 1 estimates the poverty implications of an average tariff applied uniformly across all consumption levels; 2) Shock 2 shows the impact of an average tariff increase applied progressively to all consumption categories except the lifeline - i.e. no changes to the lifeline and increasingly larger tariff increases for consumers of higher levels of electricity; 3) Shock 3 uses the same mechanism as Shock 2, but the tariff increase is only applied to consumption categories larger than 150kWh i.e. no changes to tariffs for those consuming up to 150kWh. Under conservative assumptions (no income growth or energy substitution), the most affected group the lower middle income are better off with a tariff scheme that only increases prices for those consuming more than 150kWh. That is, the changes in poverty rates under Shock 3 are the smallest. Further, changes in poverty rates using average tariff increases of 35% or higher are smaller than changes in poverty rates calculated using the June 2010 tariff increases, an average increase of 27%. These findings provide evidence that consumers can tolerate even higher tariff increases particularly if those in the 0-150kWh consumption bracket are protected. bracket again, but until then the prepayment customers in the 51-150kWh consumption bracket will be significantly impacted. (Per Sunil Mathrani Email 6/9/2010) 6 The prices quoted here follow PURC published tariffs. However, in reality, prices for this consumer group were much lower prior to the June 2010 tariff increase. Using these prices, the tariff increase for these customers is 115%. (Per Sunil Mathrani Email 6/9/2010) 7 The alternatives to electricity for a household s energy needs in Ghana include: kerosene, candles, firewood, charcoal, gas and generators. As many of these options are not a good substitute for electricity or they are a very expensive alternative (i.e. generators), it is assumed that the level of substitution between these forms of energy is very low. 8 PURC Press Release, 2010
ANNEX 1: Tables and Methodology Table 1: Residential Tariff Structure In Cedis Kwh/Month Feb 98 * Sep 98 * May 01 * Aug 02 * 2/ Mar 03 Oct 03 3/ 0-50 (Lifeline) 1/ 87 174 339 391 565 565 51-150 50 120-150 242-304 400 550 610 151-300 50 120-150 242-304 400 550 610 301-600 75-180 220-350 570 960 960 1065 Over 600 75-180 220-350 570 960 960 1065 Lifeline (Cedis/month) 2000 4000 7800 9000 13000 13000 In GHp Percent Increase KWh/Month 2007 4/ Nov 07 * Jun 10 (Oct 03-2007) (2007 - Nov 07) (Nov 07 - Jun 10) 0-50 (Lifeline) 1/ 7.00 9.50 9.50 24% 36% 0% 51-150 5/ 7.00 9.50 17.00 15% 36% 79% 151-300 7.00 12.00 17.00 15% 71% 42% 301-600 12.00 16.00 21.00 13% 33% 31% Over 600 14.00 19.00 23.00 31% 36% 21% * Denotes Tariff adjustment due to a major tariff review Source: PURC website, PURC Press Release, 2007, PURC Press Release, 2010 Notes: 1/ This is a flat rate. Assumes average consumption levels of 23KWh/month. 2/ Government subsidy of GH 0.5 (5,000 cedis) starts. 3/ Government subsidy increased to GH 0.608 (6,080 cedis) to keep lifeline constant. 4/ Rates at the time of the Nov. 2007 tariff decision. (See PURC Press Release, 2007) 5/ Lifeline threshold increased to 150kWh. According to the June 2010 Press Release, the lifeline threshold is now 50kWh.
Table 2: Basic Statistics on Poverty, Access to Electricity and Consumption, 2005 Rate of access to Decile Poverty Rate 1/ electricity Total Annual Consumption per adult equivalent 2/ Annual Electricity Expenditure per adult eq. 2/ (In GH - real terms) 1 100.0% 9.1% 153 0.34 2 100.0% 24.5% 262 1.06 3 85.3% 29.4% 343 1.83 4 0.0% 38.9% 423 3.21 5 0.0% 41.6% 510 4.09 6 0.0% 47.4% 612 5.51 7 0.0% 55.3% 729 7.68 8 0.0% 64.6% 903 11.18 9 0.0% 71.6% 1,168 16.15 10 0.0% 83.1% 2,267 35.03 All 28.5% 46.5% 737 8.49 Source: Ghana Living Standards Survey 2005/6 (GLSS5) Note: 1/ Poverty line is the upper line of 370.89 Gh (3,708,900 cedis) per adult per year. This is the 900,000 cedis poverty line used in the 2000 Poverty Profile by the Ghana Statistical Service, inflated to 2006 price levels. 2/ Consumption and Electricity Expenditure are only for households with access to electricity.
Table 3: Poverty Implications of June 2010 Tariff Increase 9 Poverty Rates 1/ Scenario 1 - No income Growth Scenario 2 - Growth Change in Change in Decile Nov 2007 Jun 2010 Rate Nov 2007 Jun 2010 Rate 1 100.0% 100.0% 0.0 100.0% 100.0% 0.0 2 100.0% 100.0% 0.0 100.0% 92.1% -7.9 3 87.1% 87.7% 0.7 52.1% 1.6% -50.5 4 0.0% 2.7% 2.7 0.0% 0.0% 0.0 5 0.0% 0.0% 0.0 0.0% 0.0% 0.0 6 0.0% 0.0% 0.0 0.0% 0.0% 0.0 7 0.0% 0.0% 0.0 0.0% 0.0% 0.0 8 0.0% 0.0% 0.0 0.0% 0.0% 0.0 9 0.0% 0.0% 0.0 0.0% 0.0% 0.0 10 0.0% 0.0% 0.0 0.0% 0.0% 0.0 All 28.7% 29.1% 0.3 25.2% 19.4% -5.8 Source: Ghana Living Standards Survey 2005/6 (GLSS5) Note: 1/ Poverty line is the upper line of 370.89 Gh (3,708,900 cedis) per adult per year. This is the 900,000 cedis poverty line used in the 2000 Poverty Profile by the Ghana Statistical Service, inflated to 2006 price levels. 9 See Methodology for details on calculations
Table 4: Sensitivity Analysis - Poverty Implications of Various Tariff Increases 10 Poverty Rates 1/ Scenario 1: No income growth Shock 1 Change in Poverty Rate - Shock 1 2 100.0% 100.0% 100.0% 100.0% 100.0% 0.0 0.0 0.0 0.0 3 87.1% 88.4% 88.4% 88.6% 88.8% 1.4 1.4 1.6 1.7 4 0.0% 0.2% 1.0% 1.2% 1.4% 0.2 1.0 1.2 1.4 All 28.7% 28.9% 29.0% 29.0% 29.0% 0.2 0.2 0.3 0.3 Shock 2 Change in Poverty Rate - Shock 2 2 100.0% 100.0% 100.0% 100.0% 100.0% 0.0 0.0 0.0 0.0 3 87.1% 87.6% 87.6% 87.7% 87.7% 0.5 0.5 0.7 0.7 4 0.0% 0.7% 1.1% 1.3% 1.4% 0.7 1.1 1.3 1.4 All 28.7% 28.8% 28.9% 28.9% 28.9% 0.1 0.2 0.2 0.2 Shock 3 Change in Poverty Rate - Shock 3 2 100.0% 100.0% 100.0% 100.0% 100.0% 0.0 0.0 0.0 0.0 3 87.1% 87.1% 87.1% 87.1% 87.1% 0.0 0.0 0.0 0.0 4 0.0% 0.8% 0.8% 0.9% 0.9% 0.8 0.8 0.9 0.9 All 28.7% 28.8% 28.8% 28.8% 28.8% 0.1 0.1 0.1 0.1 Source: Ghana Living Standards Survey 2005/6 (GLSS5) Note: 1/ Poverty line is the upper line of 370.89 Gh (3,708,900 cedis) per adult per year. This is the 900,000 cedis poverty line used in the 2000 Poverty Profile by the Ghana Statistical Service, inflated to 2006 price levels. 10 See Methodology for details on calculations
Table 4: Sensitivity Analysis - Poverty Implications of Various Tariff Increases 11 Poverty Rates 1/ Scenario 2: growth Shock 1 Change in Poverty Rate - Shock 1 2 100.0% 91.9% 92.0% 92.0% 92.0% -8.1-8.0-8.0-8.0 3 52.1% 1.4% 1.4% 1.6% 1.6% -50.7-50.7-50.5-50.5 4 0.0% 0.0% 0.0% 0.0% 0.0% 0.0 0.0 0.0 0.0 All 25.2% 19.4% 19.4% 19.4% 19.4% -5.9-5.8-5.8-5.8 Shock 2 Change in Poverty Rate - Shock 2 2 100.0% 92.0% 92.0% 92.0% 92.0% -8.0-8.0-8.0-8.0 3 52.1% 1.1% 1.3% 1.3% 1.4% -51.0-50.8-50.8-50.8 4 0.0% 0.0% 0.0% 0.0% 0.0% 0.0 0.0 0.0 0.0 All 25.2% 19.4% 19.4% 19.4% 19.4% -5.9-5.9-5.9-5.9 Shock 3 Change in Poverty Rate - Shock 3 2 100.0% 91.9% 91.9% 91.9% 91.9% -8.1-8.1-8.1-8.1 3 52.1% 0.6% 0.6% 0.6% 0.6% -51.5-51.5-51.5-51.5 4 0.0% 0.0% 0.0% 0.0% 0.0% 0.0 0.0 0.0 0.0 All 25.2% 19.3% 19.3% 19.3% 19.3% -5.9-5.9-5.9-5.9 Source: Ghana Living Standards Survey 2005/6 (GLSS5) Note: 1/ Poverty line is the upper line of 370.89 Gh (3,708,900 cedis) per adult per year. This is the 900,000 cedis poverty line used in the 2000 Poverty Profile by the Ghana Statistical Service, inflated to 2006 price levels. 11 See Methodology for details on calculations
Methodology: Table 3: Poverty Rates Scenario 1: Poverty Rate calculations only simulate effects of a tariff increase. The Nov 2007 poverty rate is calculated using 2005 consumption (per adult equivalent) that is reduced by the difference between the 2005 electricity expenditure and the Nov. 2007 electricity expenditure (2005 electricity consumption inflated to 2007 levels using the Nov. 2007 percentage increase in tariffs). The June 2010 poverty rate is calculated using 2005 consumption (per adult equivalent) that is reduced by the difference between the Nov 2007 electricity expenditure and the June 2010 electricity expenditure (2005 electricity consumption inflated to 2010 levels using the June 2010 percentage increase in tariffs). All consumption and expenditure values are converted into real terms using a consumer price index provided in the GLSS5 dataset. Table 4: Scenario 1 Poverty Rates: Scenario 1: Poverty Rate calculations only simulate effects of a tariff increase. The Nov 2007 poverty rate is calculated using 2005 consumption (per adult equivalent) that is reduced by the difference between the 2005 electricity expenditure and the Nov. 2007 electricity expenditure (2005 electricity consumption inflated to 2007 levels using the Nov. 2007 percentage increase in tariffs). The other poverty rates are calculated using 2005 consumption (per adult equivalent) that is reduced by the difference between the Nov 2007 electricity expenditure and new 2010 electricity expenditure (2005 electricity consumption inflated to to 2010 levels using Shocks 1,2 or 3). All consumption and expenditure values are converted into real terms using a consumer price index provided in the GLSS5 dataset. Shock 1: Applies average tariff increases of 35%, 40%, 45% and 50% uniformly across all tariff categories. Shock 2: Applies average tariff increases of 35%, 40%, 45% and 50% as follows: 0% to the lifeline group and the rest progressively applied to higher consumption categories. Shock 3: Applies average tariff increases of 35%, 40%, 45% and 50% as follows: 0% to the lifeline group, 0% to the 51-150Kwh group and the rest progressively applied to higher consumption categories. Table 4: Scenario 2 Poverty Rates Scenario 2: Poverty Rate calculations simulate the effects of a tariff increase as well as income growth. The Nov 2007 poverty rate is calculated using 2005 consumption (per adult equivalent) inflated by GDP growth from 2005 to 2007 and is reduced by the difference between the 2005 electricity expenditure and the Nov. 2007 electricity expenditure (2005 electricity consumption inflated to 2007 levels using the Nov. 2007 percentage increase in tariffs). The other poverty rates are calculated using 2005 consumption (per adult equivalent) inflated by GDP growth from 2005 to 2007 then 2007 to 2010 and is reduced by the difference between the Nov. 2007 electricity expenditure and new 2010 electricity expenditure (2005 electricity consumption inflated to 2010 levels using Shocks 1,2 or 3). All consumption and expenditure values are converted into real terms using a consumer price index provided in the GLSS5 dataset. Shock 1: Applies average tariff increases of 35%, 40%, 45% and 50% uniformly across all tariff categories. Shock 2: Applies average tariff increases of 35%, 40%, 45% and 50% as follows: 0% to the lifeline group and the rest progressively applied to higher consumption categories. Shock 3: Applies average tariff increases of 35%, 40%, 45% and 50% as follows: 0% to the lifeline group, 0% to the 51-150Kwh group and the rest progressively applied to higher consumption categories.
Bibliography Energy Sector Management Assistance Programme (ESMAP), Ghana Poverty and Social Impact Analysis of Electricity Tariffs, ESMAP Technical Paper, World Bank, December 2005 (Ghana PSIA, 2005) Ghana Public Utilities Regulatory Commission (PURC), New Electricity Tariffs, Press Release, October 2007 (PURC Press Release, 2007) Ghana Public Utilities Regulatory Commission (PURC), Approve Electricity Tariffs Effective 1 st June 2010, Press Release (PURC Press Release, 2010) Ghana Web, Government Intervenes in Electricity Tariffs Increment, December 2007. Available at http://www.ghanaweb.com/ghanahomepage/health/artikel.php?id=135567 (Ghana Web, 2007) Government of Ghana, Ministry of Energy, PSIA Steering Committee, Ghana Poverty and Social Impact Analysis Electricity Tariffs: Phase 1, World Bank, June 2004 (Ghana PSIA, 2004) Wodon, Q., Improving the Targeting of Social Programs in Ghana, Draft for Internal Discussion, World Bank, October 2009 (Wodon, 2009)