64280 DRAFT: Not for Citation PSIA: POWER SECTOR REFORMS AND THE POOR IN VIETNAM Valerie Kozel and Nguyen Viet Cuong January 11, 2010 DRAFT: Not for Citation DRAFT: Not for Citation PSIA: POWER SECTOR REFORMS AND THE POOR IN VIETNAM EXECUTIVE SUMMARY The high level of electricity access across Vietnam means that large numbers of the poor have connections and therefore have the potential to be impacted by changes to tariffs and service levels. Two reforms to the tariff will affect the poor: • Beginning March, 2009 the incremental block tariff (IBT) structure is to be applied equally in all rural and urban areas. The first 50kWh per month of consumption is subject to a lifeline tariff. The tariff for this “lifelineâ€? block is set at 35-40% of average 2009 costs (excluding profits) and is VND 600/kWh. The 51-100 kWh block is set at VND 865/kWh, the average 2009 costs (excluding profits). Higher blocks of consumption are set above average costs (including profits) to help finance subsides provided through lower consumption blocks. The profit cost component of the average tariff is recovered through tariffs applied for residential consumption above 100 kWh as well as non residential consumption. • From 2010 further changes to the subsidy mechanism are under discussion, including (a) limiting subsidies to only those customers consuming less than 50 kWh and (b) targeting subsidies to rural customers only. These arrangements put all residential consumers on a more equal footing and aim to improve targeting of subsidies. Previously only those households which received their supply directly from EVN’s subsidiary PCs received a lifeline tariff, which was for the first block of 100kWh of consumption per month and set at VND 550/kWh. The second block of monthly consumption of 100 – 150kWh was also slightly subsidized. PCs supply all consumers in urban areas and a growing number – currently about 50 percent – of households in rural areas. The 50 percent of rural consumers not receiving supply from PCs are customers of local distribution utilities (LDUs), which are private sector or cooperative suppliers receiving supply from PCs at medium voltage and on- selling to consumers at low voltage. Until March 2009, LDUs tariffs were set by the provincial or local government. Formally these tariffs were capped by the Prime Minister at VND 700/kWh, and anecdotal evidence suggests that many LDUs charge this much or more and that few, if any, apply subsidy for low income consumers. Poverty and Social Impact of IBT Changes The poverty and social impact analysis (PSIA) analyzed expected poverty and social impacts of changes to the March 2009 IBT including both the reduction in the coverage of the lifeline tariff and higher electricity prices. The analysis suggests the poor in Vietnam remain well-protected despite the recent tariff adjustments and the narrowing of the lifeline band, and many rural consumers are likely to benefit from the introduction of a unified system of residential tariffs and expected future integration of many LDUs into the PCs. For those continuing to receive power from the LDUs, benefits are expected not only in terms of lower electricity prices but also improvements in the quality and reliability of supply, and improved standards of safety. There is a widely shared view – buttressed by experience from many countries – that aligning power tariffs with the economic cost of supply is usually regressive and almost invariably has an adverse impact on the poor. However this PSIA shows clearly that this is not so in Vietnam, which has a number of factors working in its favor: very high coverage by the system, timely bill payment and low arrears, low levels of electricity consumption (particularly for lower income households), and generally low economic costs of supply. An estimated 98 percent of households were electrified by 2008, including 99 percent in urban and 97 percent in rural areas. Access to electricity is high even among Vietnam’s poorest households: in 1998, less than half of the poorest ten percent of households (bottom decile) used electricity, compared with 88 percent in 2008. The PSIA estimated a subsistence threshold of 40-45 kWh for poor households in Vietnam, based on owning several lights, an electric fan, a rice cooker, and a color i DRAFT: Not for Citation television. In 2008, households below the poverty line consumed on average 42 kWh/month as compared with 103 kWh/month for households above the poverty line. While most of the poor in Vietnam are low consumers of electricity, so are many of the non-poor. In 2008, 65 percent of all households consumed less than 100 kWh/month, including 91 percent of poor urban households and 99 percent of poor rural households. Thus most of the electricity consumed by the poor was subsidized under the pre-2009 IBT lifeline. But so was the consumption of many of the non-poor: in 2008, only 14 percent of households in Vietnam lived below the poverty line. Moving to a lower lifeline threshold reduces leakages albeit at some cost to the poor: 30 percent of households consume less than 50 kWh/month, including 60 percent of poor urban households and 78 percent poor rural households. Despite the tariff increases, electricity remains remarkably affordable to residential consumers in Vietnam. In 2008, households in poorest 10% of the population paid on average 2.9 percent of total cash expenditures for electricity. In contrast, the wealthiest 10 percent of households paid 3.6 percent. And the share of total household spending taken by electricity has been falling in recent years despite rising consumption; the real price of electricity (adjusted for inflation) has been falling. Tariff Reform Scenarios Electricity consumption is not measured directly in the Vietnam Household Living Standards Survey (VHLSS), but spending on electricity is recorded. Electricity expenditures were converted into kWh quantities using the relevant prices (IBT structure). The VHLSS does not identify whether rural households are PC or LDU customers and so for this PSIA a unified tariff structure is assumed to apply to all rural consumers both pre and post-reform. This leads to an over-estimate of average rural consumption (because LDUs generally charge higher prices) but ensures the PSIA has made a conservative assessment of the coverage of the reduced lifeline tariff band. Electricity consumption is held fixed at 2008 levels and monetary values are adjusted, using the General Statistics Office (GSO) monthly CPIA to March, 2009. The analysis uses electricity spending as a share of cash consumption expenditures, a measure of affordability, as the criterion for comparison. The PSIA developed three tariff scenarios that are assessed against a baseline of actual patterns of consumption and electricity spending for 2008: • The first scenario simply compares the baseline against 2009 IBT adjustments. • The second scenario uses the 2009 IBT adjustments but limits those receiving the subsidized rate (VND 600/kWh) to rural consumers only. The lowest price band for urban consumers is set at the average cost of supply without profits (VND 865/kWh). • The third scenario uses the 2009 IBT adjustments, but assumes subsidies are “telescopedâ€? and only are provided to households which consume less than 50 kWh/month. In other words, a consumer using 51kWh/month pays the unit tariff of VND 865/kWh for all 51kWh consumed. The effects of the different scenarios on different groups are shown in the table below. Across all scenarios, and the key groups of concern, namely poor and rural consumers, impacts on affordability are small. This is particularly the case for the first scenario, which models the actual IBTs now in effect in Vietnam. The increase in tariffs was for the most part offset by inflation, and the narrowing of the lifeline tariff from 100 kWh to 50 kWh only increased the share of the poor’s spending on electricity by 0.2 percentage points. The other scenarios also had only modest impacts on affordability. ii DRAFT: Not for Citation Scenar ios for Affor dability Following Tar iff Refor m Baseline: Reform Scenarios (% spending on electricity) 2008 VHLSS Scenario 1: Scenario 2: Scenario 3: 2009 (% spending 2009 IBT 2009 IBT, Rural IBT, Telescoped on electricity) Only subsidy 50 kWh Poverty Status Poor 2.6 2.8 3.1 2.8 Non-Poor 3.2 3.2 3.7 3.3 Location Urban 4.0 4.0 4.4 4.2 Rural 2.8 2.9 3.3 2.9 All Vietnam 3.2 3.2 3.3 3.6 Source: 2008 VHLSS, Staff Estimates Coverage of the Urban Poor – Migrant Workers and Temporary Residents While Vietnam’s tariff adjustments do not on average appear to have adverse distributional consequences, specific groups still may be losing out. A study was conducted in one low income, high-migrant neighborhood in each of Ho Chi Minh City, Hai Phong, and Hanoi to assess service quality and affordability for migrant workers and other temporary residents who live in apartment buildings or rooming houses in densely settled urban areas. Interviews were conducted with residents in the buildings, landlords, and local officials from the wards and districts. In nearly all cases, tenants pay electricity charges based on metered consumption; electricity sub-meters were installed by the landlords in rented rooms and apartments and connected to regulated (EVN) meters in the building. Electricity contracts are organized between the owners of the buildings and EVN, and not between EVN and individual tenants/groups of tenants. In each of the neighborhoods studied, the vast majority of respondents are on low incomes and consume no more than 10-20 kWh/month. In Hanoi, landlord-imposed electricity charges range between VND 1,000-2,500/kWh, with VND 1,500-2,000/kWh being the most common charge. Rates are lower in Hai Phong and substantially higher in Ho Chi Minh City, where landlord-imposed electricity rates range from VND 2,500-3,500/kWh. In addition to metered charges, tenants also pay for electricity used for pumping water. This class of consumer would therefore appear to be relatively unaffected by tariff changes resulting from reforms implemented in 2009, since they are charged rates by landlords which are weakly linked to the official rates and they do not receive subsidies through the lifeline tariff. iii DRAFT: Not for Citation DRAFT: Not for Citation 1. INTRODUCTION 1. Vietnam is implementing a long run reform agenda for the power sector, with the aim of restructuring the sector to improve internal operations, efficiency, and the quality of services. The World Bank has an ongoing engagement in the sector which started in 1995. The design and implementation of the power sector reform strategy has been supported through technical assistance and lending operations, as well as preparation of a proposed series of Power Sector Reform Development Policy Loans (DPLs). 2. A number of Poverty and Social Impact Assessment (PSIA) activities will be carried out in support of Vietnam’s power sector reforms and as part of the World Bank’s due diligence work for the power sector DPLs. PSIA has done ex ante analysis to identify potential adverse distributional impacts of specific reform measures. It will (a) support ex post monitoring and special studies (led by Electricity Regulatory Authority of Vietnam – ERAV) to identify adverse consequences for poor and marginal households that may emerge over time; and (b) identify specific measures to mitigate anticipated or actual adverse impacts. 3. This is the first in a series of PSIA notes for power sector reforms. It is based on early access to a preliminary version of the Vietnam Household Living Standards Survey (VHLSS) 2008; all analyses based on these data are preliminary World Bank estimates. This PSIA underpins reforms supported under the First Power Sector Reform DPL. It builds on a number of earlier studies on welfare impacts of rural electrification1,2 and on power tariffs, including subsidy and fund mechanisms 3,4,5. 4. This PSIA focuses on assessing the distributional impacts of recent tariff reforms introduced in March, 2009 which changed Vietnam’s incremental block tariff (IBT) structure for residential consumers. Over time, tariff subsidies will gradually be phased out and tariffs will ultimately cover the full economic cost of supply. Complementary explicit subsidies targeted at low income households are under consideration. Recent reforms both narrowed the band of consumption which receive a preferential rate (known as the “lifeline tariffâ€?) from 100 KWh to 50 kWh per month and increased the retail price of electricity to all users. Of particular relevance for PSIA, retail prices were increased for users falling within the new lifeline band (first 50 kWh) as well as users in the upper half of the old lifeline band (51-100 kWh). In addition to changes in lifeline tariffs and coverage, all on-grid rural consumers were brought under a unified tariff structure. 6 Questions Addressed 5. This first PSIA addresses three questions: • What are the distributional impacts of the new power tariffs introduced in March, 2009? In particular: − How are poor and near-poor households likely to be affected by the reduction in lifeline tariff band from 100kWh to 50 kWh and accompanying increase in the retail price of electricity? Does the lifeline band provide sufficient coverage of the poor? 1 Institute of Sociology of the Vietnam Academy of Social Sciences: Impacts of Rural Electrification in Vietnam, 2009 2 World Bank: Welfare Impacts of Rural Electrification: Evidence from Vietnam (Khandker, Barnes, Samad, Huu Minh) September 2009 3 Economic Consulting Associates: Bulk Power, Distribution Margin, Retail Consumer Tariff Design and Development of an Independent Creditors’ Model, July 2005. 4 Pedro Antmann: Review of International Experiences on Retail Tariff Setting Methodologies and Recommendations for Vietnam, July 2007 5 Vietnam Electricity: Power Tariffs, 2007 6 New tariffs are described in PM Decision 21/2009/QD-TTg of February 12, 2009 “Electricity Retail Tariff of 2009 and years 2010 – 2012 Following the Market Mechanismâ€? and MOIT Circular No. 05/2009/TT-BCT of February 26, 2009 “Prescribing 2009 Electricity Sale Prices and Guiding the Application Thereofâ€?. -1- DRAFT: Not for Citation − How are households likely to be affected by the move to a unified tariff structure in rural areas? Before the 2009 tariff reforms, some 6.2 million rural customers were supplied by Local Distribution Utilities (LDUs) that purchased electricity at wholesale rates from EVN and sold it to rural consumers at a higher official ceiling price (700 VND/kWh) than the retail prices paid by customers supplied directly by EVN’s power distribution companies (known as Power Companies, or PCs). • Are some groups of poor and vulnerable households not well covered under the residential tariff lifeline band? In particular, are Vietnam’s migrant workers who live temporarily in urban and peri-urban areas supplied from the formal system and how much do they pay for electricity? Are additional measures needed to ensure that (poor) migrants receive subsidy entitlements? • A high proportion of poor and vulnerable households currently have access to a reliable supply of electricity at a reasonable price. This situation may change in the future as reforms progress. Hence what additional information should be collected by ERAV as reforms are implemented to identify any emerging adverse distribution impacts and help in developing effective policy responses? What special studies are needed? 2. VIETNAM’S POWER SECTOR: THE IBT STRUCTURE 6. Residential and rural wholesale consumption accounted for about 41 percent of EVN PC’s total sales in 2009. Vietnam put in place an incremental residential block structure beginning in 1999, which initially had five blocks. An additional block was added at the top end in 2007, and a division in the lowest 100 kWh block (0-50, 51-100 kWh) added in March, 2009 (Table 1). Table 1: IBTs for Vietnam’s Power Sector : 1999 – Pr esent (VND/kWh) 10/1999 – 11/2002 – 2/2005 – 1/2007 2/2007 – 3/2009 – date 10/2002 1/2005 3/2009 PM Decision PM Decision PM Decision 215/2004/ PM Decision PM Decision kWh 193/1999/QÄ?- 124/2002/QÄ?- QÄ?-TTg dated 29/12/2004 276/2006/QÄ?- 21/2009/QÄ?- TTg dated TTg dated <300 >300 TTg dated TTg dated 22/09/1999 20/09/2002 kWh/month kWh/month 04/12/2006 12/02/2009 0-50 600 500 550 550 1100 550 51-100 845 101-150 704 900 900 1100 1110 1135 151-200 957 1210 1210 1100 1470 1495 201-300 1166 1340 1340 1340 1600 1620 301-400 1400 1720 1740 1397 1400 1400 >400 1500 1780 1790 Lifeline Tariffs 7. IBTs are often used by electricity retailers to provide cross subsidies to users who consume less than a subsistence threshold considered adequate for meeting basic needs; hence the use of the term “lifelineâ€? rate. The benefits of tariff subsidies for low-income consumers are well understood and visible, although their costs often are not. Lifeline subsidies entail low administrative costs, and they often enjoy widespread political support, particularly when the subsistence threshold is set sufficiently high that subsidies also benefit the less poor. 8. One of the major drawbacks of lifeline tariffs is that they may not be well-targeted – a household’s level of electricity consumption may not be a good indicator of poverty – and leakages to the non-poor are often high. In lieu of lifelines, many countries compensate poorer households for high electricity tariffs through income transfer and related social protection measures. No compensation method clearly dominates another: a number of practical considerations and design features determine which works best in a specific country context, including a judicious combination of lifelines and income transfers. -2- DRAFT: Not for Citation 9. The lowest 100 kWh block was designated as Vietnam’s lifeline band and priced at less than half the economic cost of supply in 2005. Despite inflation and rising costs, the lifeline tariff was kept constant in nominal terms until the adjustments approved by the Prime Minister in February, 2009, which meant the level of subsidy in the lifeline band increased. There is substantial leakage to the non-poor: according to a study in 2005 (ECA, 2005), 46 percent of all EVN retail customers consumed less than 100 kWh each month representing 51 percent of sales. Vietnam’s official poverty rate in 2004 was 19.5 percent: rural households consumed on average 53 kWh of electricity each month, while households below the poverty line consumed on average 29 kWh of electricity each month 7. Moreover, all residential consumers received the subsidized rate for their first 100 kWh each month, which further spread benefits to higher income consumers. 10. The Prime Minister’s Decision of February 2009 reduced the first block from 100 kWh to 50 kWh and set the new lifeline tariff (tariff for first block) at around 40 percent of the economic cost of supply (without profits). The next block (51-100 kWh) is priced at the economic cost of supply, also without profits for the power companies. Profits are covered by residential tariffs in higher blocks as well as cross subsidies from other tariff categories, mainly industrial and commercial users. With the aim of ensuring fair treatment of all households in Vietnam, the PM’s Decision also put all residential consumers under a single, unified tariff structure, regardless of whether they were supplied by LDUs or EVN. The new tariffs went into effect in March, 2009 for all EVN customers. The LDUs were grandfathered under the old tariff structure through September, 2009, with the aim of giving them time to transition to the new cost structure. 11. The February Decision increased tariffs for consumption of up to 100 kWh for the first time since 2002. Tariffs for the 0-50 kWh block were increased from VND 550 to 600/kWh, and tariffs for the 51-100 kWh were increased from VND 550 to 865/kWh. Although an increase in nominal terms, after adjustments for inflation the first block tariff is still lower in real terms than tariffs set in 2002. The government does not provide any financing for subsidies so the system is entirely dependent on charges from other classes of consumer making up any shortfall from the lifeline revenues. Not increasing the lifeline tariff would have resulted in an erosion of the revenues of the PCs. Making larger adjustments to the tariffs paid by better off consumers or non-residential customers in order to increase cross-subsidizes to low income households would have introduced additional distortions and reduced overall efficiency. Unified Tariffs in Rural Areas 12. More than 5,600 LDUs were operating in rural communes as of June, 2008. Many LDUs are very small and operate on a narrow profit margin. Studies 8 suggest that several LDUs struggle with issues of capacity, quality of supply, and safety: an estimated 200 electricity-related deaths were reported in one province last year. The government requires LDUs to operate according to a new set of performance criteria in the future, including obtaining an electricity distribution and retail supply license, developing a reliable and transparent system of accounts, entering into a supply contract with each customer, issuing monthly bills and ensuring all customers have a certified power meter. They must also adhere to the unified tariff. EVN estimates that less than 700 existing LDUs will be able to meet the standards and operate profitably under the unified IBT system. The assets of those that cannot will be transferred to the PCs: by June, 2009, 8.8 million rural households were supplied by EVN (6,364 communes out of around 9,310 in total). An estimated 3,300 LDUs were transferred to the EVN PCs between June, 2008 and June, 2009. 13. Rural customers are expected to benefit in several ways from the unified tariffs and accompanying expansion of services by PCs: poor households will pay lower prices for their first 50 kWh of consumption. In addition, there should be substantial improvements in the quality and reliability of supply, as well as improved safety standards over the medium term. 7 Based on WB analysis of the 2004 VHLSS 8 Reference studies from ERAV -3- DRAFT: Not for Citation 3. TARIFF ADJUSTMENTS AND THE POOR 14. There is a widely shared view – buttressed by experience from many countries – that aligning power tariffs with the economic cost of supply is often regressive and will invariably have an adverse impact on the poor. However impacts depend critically on country specific conditions prevailing at the time reforms are introduced. Vietnam has three factors working in its favor: very high coverage by the electricity system, low levels of electricity consumption (particularly for lower income households), and fairly low economic costs of supply (in comparison to a number of other countries). 15. The remainder of this Section 3 describes the current coverage of the power sector, supply costs, levels of demand and minimum needs consumption for the poor, and assesses the impact of the March 2009 tariff adjustments, including further narrowing of the lifeline tariff band and limited subsidies to rural consumers. Section 4 reports key findings of the ongoing study of the impact of power sector reforms on migrant workers. A list of core monitoring indicators and additional studies/analyses are included in the fifth section. Defining the Poor 16. The PSIA draws heavily on recent rounds of the Vietnam Household Living Standards Surveys (VHLSS), a series of nationally representative household surveys conducted by the Government Statistics Office (GSO) for the first time in 1993, then in 1997/98, 2002, 2004, 2006, and 2008. The World Bank was given advance permission to use the newest VHLSS (2008) for the PSIA, which covered more than 9,000 household located in all provinces of the country. The VHLSS includes detailed information on household incomes and expenditures, poverty status, ownership of durables including electrical appliances, and access to electricity including monthly electricity payments. 17. The “poorâ€? are defined as all persons living in households whose per-capita consumption falls below some agreed cut-off level or poverty line. Two poverty lines are frequently used for Vietnam: The first (referred to as the WB/GSO poverty line) is produced by the Government Statistics Office (GSO) and defined as the cost of a food consumption basket allowing a daily intake of 2,100 calories per person per day, plus the cost of a related non-food consumption that allows for a healthy life. The WB/GSO poverty line has been held roughly constant in real purchasing power since the early 1990s, to facilitate comparisons over time. According to this line, poverty levels have fallen from 58 percent of the population in 1993 to 14 percent in 2008 (Figure 1). 18. Over the years, the Ministry of Labor, Invalids, and Social Assistance (MOLISA) has produced several different poverty lines, initially based on the income required to purchase an adequate amount of rice for daily consumption. The most recent MOLISA poverty line is based on a similar 2,100 calorie benchmark used in the WB/GSO poverty line, applied to per-capita incomes rather than per-capita expenditures. Using the MOLISA approach in 2006, the Government of Vietnam set its official poverty lines at VND 260,000 per person per month for urban areas and VND 200,000 for rural areas. Joint work is underway (MOLISA with support from GSO) to revise the official lines to reflect better current living conditions in Vietnam. The official lines are used to draw up MOLISA’s poor household list at the commune level. The two lines currently yield very similar estimates of the national poverty headcount, although there are sometimes differences at provincial and regional levels. -4- DRAFT: Not for Citation Figur e 1: Pover ty Rate in Vietnam (% ) 1993 – 2008 Source: VHLSS 1993, 1997, 2002, 2004, 2006, 2008 VHLSS; 2008 poverty rates are preliminary World Bank estimates, all others from GSO 19. The PSIA uses the WB/GSO poverty line applied to household per-capita consumption measured through various rounds of the VHLSS. Key findings were cross-checked using an alternative definition based on the MOLISA poor list. Results were in all cases broadly consistent. System Coverage 20. System coverage is high in Vietnam compared with other countries at similar levels of development. As of June, 2009, EVN estimated that 96 percent of households were connected to the grid, including 94.7 percent of households in rural areas. Only 224 communes (in mountainous or other difficult terrain) were not electrified. 21. The VHLSS surveys show similar high rates of coverage. By 2008, 97.7 percent of households were electrified 9, including households on the grid as well as a small number using other sources such as pico-hydro sets (Table 2). Table 2: Household Access To Electr icity By Location and Income Gr oup (Per centage of Households using Electr icity) 1998 2002 2004 2006 2008 Non-poor 86.3 93.0 96.3 97.6 98.8 Poor 62.2 72.9 82.3 87.6 89.7 Urban 95.2 97.3 99.8 Rural 72.3 84.5 81.5 89.5 96.9 Expenditure Quintile Poorest 10% 49.6 62.2 78.4 85.3 88.4 Decile 2 63.1 77.5 88.9 93.2 96.1 Decile 3 70.8 84.4 91.4 95.0 97.9 Decile 4 74.0 88.7 95.3 96.4 98.5 Decile 5 78.1 90.3 95.0 97.4 98.8 Decile 6 83.2 91.0 96.7 98.3 99.0 Decile 7 83.3 93.4 96.9 98.8 99.7 Decile 8 88.6 94.8 98.5 99.0 99.5 Decile 9 95.8 97.3 98.6 99.2 99.6 Wealthiest 10% 98.5 99.2 99.4 99.6 99.9 All Vietnam 78.5 87.9 93.9 96.2 97.7 Source: VHLSS 1997, 2002, 2004, 2006, 2008, staff calculations (preliminary 2008 VHLSS estimates) 9 Survey results are based on a response to the question “what is your main source of lighting?â€?. -5- DRAFT: Not for Citation 22. Rates are highest for better-off households (98.9 percent) but still respectable (89.7 percent) for the 14 percent of households below the poverty line in 2008. There has been substantial progress in providing access to electricity for the poorest households in Vietnam: in 1998, less than half the households in the poorest decile had access to electricity. By 2008, over 88 percent of the poorest 10 percent of households used electricity. Regional coverage also improved dramatically: in 1998, electricity coverage hovered at 50-60 percent in North East, North West, Central Highlands and the Mekong Delta. By 2008 coverage had increased to over 95 percent in all regions except the North West, where coverage still lags at 84 percent (Table 3). Table 3: Household Access to Electr icity by Region: 1998 – 2008 Percent of Households with Electricity 1998 2002 2004 2006 2008 Region Red River Delta 97.9 99.3 99.8 99.7 99.9 North East 59.8 80.3 91.5 93.7 94.5 North West 90.5 50.9 73.3 78.2 84.1 North Central Coast 82.4 95.2 96.9 97.9 98.9 South Central Coast 81.8 92.9 98.4 98.6 99.1 Central Highlands 50.1 73.1 87.4 93.2 96.1 North East South 89.8 93.0 96.5 97.9 98.8 Mekong River Delta 54.5 76.1 86.8 93.4 97.1 All Vietnam 78.5 87.9 93.9 96.2 97.7 Source: VHLSS 1997, 2002, 2004, 2006, 2008, staff calculations (preliminary 2008 VHLSS estimates) Electricity Demand Measuring Household Consumption 23. There are different sources of information on how much electricity is consumed by households in Vietnam. EVN has good measures for the customers it serves directly; however, half of rural consumers are still served by LDUs and not included in EVN’s calculations. VHLSSs surveys are used for PSIA estimates because they have broader coverage. 24. Unfortunately, electricity consumption is not measured directly in the VHLSS (experience suggests that most respondents would not be able to answer accurately if asked to gauge their consumption). Instead, households are asked for information on the amount paid for electricity each month, consistent with the objective of VHLSS which is to assess household expenditures. For purposes of the PSIA, electricity expenditures were converted into kWh quantities using the relevant prices in place at the time each household was interviewed and as reported in Table 1. The VHLSS does not identify whether rural households are EVN PC customers or purchase from an LDU (at the time of the 2008 VHLSS survey about 75 percent of rural consumers were served by LDUs). For lack of better information, the PSIA uses the IBT tariff structure to estimate consumption for all rural households. This will lead to an over-estimate of average rural consumption because LDUs generally charge higher prices than PCs but ensure the PSIA makes a conservative assessment of the coverage of the reduced lifeline tariff band.10 Durables and Rising Demand 25. The demand for electricity is a derived demand, determined by demand for the services provided by electrical appliances and durables. Low income households own fewer appliances (Table 4) and use them less intensively than better off households. According to the panel study of rural 10 Information on the relative share of EVN and LDU customers in each province has been requested from EVN; if available, this will improve VHLSS estimates of electricity consumption and more importantly, provide a better basis for assessing the welfare impacts of moving to a unified tariff system in rural areas. -6- DRAFT: Not for Citation energy use (Institute of Sociology, 2009), electricity consumption increases sharply for the first six years after a household is connected and starts to purchase electrical appliances and then begins to level off. All other things held equal, at the second year of having electricity a rural household consumes about 36 kWh a month, and by the ninth year consumption increases to around 60 kWh a month. Table 4: Pr opor tion Of Households Owning Electr ical Appliances, by Pover ty Status 2002 – 2008 (Pover ty by WB/GSO Pover ty Line) TV Stereo Computer Fan Fridge Electric Water Washer Pump Cooker Heater 2002 VHLSS Non-poor 78.3 8.0 3.4 75.0 16.6 47.2 4.6 5.3 32.7 Poor 37.5 0.9 0.0 48.4 0.3 7.9 0.3 0.1 11.8 Areas Urban 87.5 15.3 8.9 82.0 37.9 69.5 12.0 14.4 29.0 Rural 61.6 3.4 0.5 64.0 4.4 27.0 0.8 0.6 27.0 2004 VHLSS Non-poor 85.2 11.8 7.4 84.6 21.6 60.5 6.4 7.9 39.6 Poor 45.1 0.8 0.0 54.3 0.2 13.0 0.0 0.1 13.8 Areas Urban 91.6 20.2 18.7 89.2 47.8 79.8 17.1 21.6 35.3 Rural 73.6 6.2 1.6 75.9 7.1 42.5 1.1 1.2 35.2 2006 VHLSS Non-poor 89.3 14.9 9.4 85.5 27.1 67.8 8.3 10.7 40.2 Poor 53.2 1.6 0.0 57.4 0.3 17.8 0.1 0.1 13.3 Areas Urban 94.1 20.5 22.0 88.6 54.1 83.1 21.4 28.0 34.1 Rural 80.8 10.2 2.9 79.0 11.9 52.6 1.7 2.1 37.4 2008 VHLSS Non-poor 93.1 17.7 13.1 88.7 37.3 77.0 11.4 15.2 47.0 Poor 65.4 3.7 0.0 61.8 1.2 31.3 0.1 0.2 18.6 Areas Urban 95.2 21.5 28.5 89.9 64.3 85.5 25.3 36.9 36.5 Rural 87.6 14.0 4.9 83.9 20.9 66.1 4.2 4.4 46.4 Source: VHLSS 2002, 2004, 2006, 2008, staff calculations (preliminary 2008 VHLSS estimates) 26. Durable ownership has increased dramatically over time in Vietnam, not only among better off and urban households but also among poor households. The non-poor own a wide range of electrical appliances – by 2008, 93 percent of households reported owning a television, 89 percent owned fans, 77 percent used electric rice cookers, and 47 percent used electric water pumps, and a smaller but substantial number of households owned stereos, refrigerators, and household computers. Even among the poor, two-thirds of households owned a television by 2008, 61 percent had at least one fan, and 31 percent owned an electric rice cooker. Ownership of heavier appliances has risen sharply in urban areas: in 2002 only 38 percent of households owned a refrigerator and 9 percent owned computers. Six years later, in 2008 the proportion of families using refrigerators had nearly doubled to 64 percent) and the proportion with computers had more than tripled to 28 percent. The lifestyles and consumption patterns of Vietnamese families are changing rapidly. 27. Electricity consumption has risen in parallel with durable ownership, for rural and urban households and across all income groups (Figures 2 and 3). The rate of increase has been slightly faster in urban areas and among better off households, reflecting a more rapid rise in relative incomes. In line with changing life styles and rising ownership of durables, the poor also are consuming more -7- DRAFT: Not for Citation electricity: average consumption for poor households rose from 28.6 kWh per month in 2002 to 42.4 kWh in 2008. Figur e 2: Aver age Electr icity Consumption for Residential Consumer s, by Pover ty Status (kWh per month): 2002-2008 Electricity Consumption (kWh) 180 160 140 120 100 80 60 40 20 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Non-poor Poor Source: VHLSS 2002, 2004, 2006, 2008 and staff calculations Figur e 3: Aver age Electr icity Consumption for Residential Consumer s, by Ur ban and Rur al Location (kWh per month): 2002-2008 180 Electricity Consumption (kWh) 160 140 120 100 80 60 40 20 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 Year Urban Rural Source: VHLSS 2002, 2004, 2006, 2008 and staff calculations Implications of Narrowing the Lifeline Band 28. At 50kWh/month, the size of the lifeline block is low by comparison with other countries where it generally ranges from 25 to 300 kWh per month, with an average value of 90 kWh per month. The subsistence consumption threshold in a particular country typically depends on patterns in appliance ownership and use: for example, in the mid-1990s, the subsistence threshold for rural households in Latin America was deemed to be around 40 kWh a month – sufficient to support a few light bulbs and a radio. The equivalent subsistence threshold for urban households was 120 kWh a month – which supported a few lights, a small refrigerator, and a television (Antmann, 2007). Thresholds are higher on average in Europe and Central Asia (reflecting higher incomes as well as -8- DRAFT: Not for Citation higher energy demands because winters are cold) and less in some lower income countries in East Asia. Subsistence Threshold for Poor Households 29. Based on current patterns of durable ownership and use, a reasonable subsistence threshold for Vietnam would be 40-45 kWh a month – still within the 50 kWh threshold of the new lifeline block. The PSIA subsistence threshold is estimated based on a daily average of: • 10 hours use of a light bulb • 2 hours use of an electric fan (higher in warm months, lower in cool months) • 2 hours use of color television • 1 hour use of a rice cooker Assessing Targeting and Leakages 30. Providing power subsidies through an IBT lifeline can be attractive in terms of administrative costs and ease of delivery; but they may not be a good way to target them to poor households. The extent to which they serve the needs for the poor depends on the characteristics of the country and the design of the system. In the case of Vietnam, the IBT lifeline provides relatively good coverage of the poor, albeit at the cost of moderate to high leakages to the non-poor. 31. The cumulative distribution of monthly electricity consumption for poor and non-poor households living in urban areas is shown in Figure 4 and for rural areas in Figure 5, based on the 2008 VHLSS. They provide a frame of reference for examining lifeline coverage and leakages linked to different lifeline cut-off levels. 32. Before March 2009, the IBT lifeline applied to the first 100 kWh of monthly electricity consumption. In 2008, 65 percent of households in Vietnam consumed less than 100 kWh per month and thus all their consumption was subsidized. Although coverage was high – over 95 percent of the poor consumed less than 100 kWh – leakages were also high: only 14 percent of households fell below the poverty line in 2008 but 65 percent received all their electricity at subsidized rates. 33. Reducing the IBT lifeline from 100 kWh to 50 kWh will have different impacts on urban and rural households. According to Figure 4, 91 percent of the poor in urban areas consumed less than 100 kWh/month as compared to 30 percent of the urban non-poor. However the poverty rate in urban areas is only 3 percent, which implies very high leakages to 97 percent of the non-poor urban population. And all households, regardless of total electricity consumed, receive subsidies on the first 100 kWh, which further contributes to leakages. Reducing the IBT lifeline to 50 kWh will reduce coverage of the urban poor (60 percent of the poor consume less than 50 kWh, Figure 4) but also sharply reduces leakages (only 8 percent of the non-poor consume less than 50 kWh). 34. Reducing the IBT lifeline also reduces leakages in rural areas. 98 percent of poor rural households consumed less than 100 kWh in 2008 and those who purchased electricity from the PCs were fully covered by price subsidies. Reducing the IBT lifeline to 50 kWh reduced full coverage to 78 percent of rural households. This moderate reduction in coverage was accompanied by a more substantial reduction in leakages; full coverage for the non-poor fell from 74 percent (100 kWh) to 30 percent (50 kWh). Although poverty levels are higher in rural areas (18.1 percent poor in 2008) compared to urban areas, leakages to the rural non-poor under the new IBT lifeline are still substantial. Although difficult to quantify given existing information, integration of LDUs into the PCs and full implementation of the unified tariff structure is likely to further improve targeting of subsidies to rural poor households. -9- DRAFT: Not for Citation Figur e 4: Distr ibution of Electr icity Consumption for Ur ban Households, by Pover ty Status: 2008 100.00 90.00 80.00 70.00 60.00 Percent 50.00 40.00 30.00 20.00 10.00 0.00 Monthly Electricity Consumption Non-poor Poor Source: VHLSS 2008, staff calculations Figur e 5: Distr ibution of Electr icity Consumption for Rur al Households, by Pover ty Status: 2008 100.00 90.00 80.00 70.00 60.00 Percent 50.00 40.00 30.00 20.00 10.00 0.00 Monthly Electricity Consumption Non-poor Poor Source: VHLSS 2008, staff calculations 35. Another way to assess how well the IBT tariff structure targets poorer households is to look at how households are distributed within the lifeline band (Figure 6). Although a high percentage of Vietnam’s poorest households (bottom two per-capita expenditure deciles) consume less than 50 kWh, many better off households do as well, particularly in rural areas. For example, 30 percent of rural households in the 5th decile consume less than 50 kWh as do 27 percent of households in the 6th decile. Thus despite the reduction in the lifeline band, many households (poor, near-poor, and not so poor) still benefit from tariff subsidies. - 10 - DRAFT: Not for Citation Figur e 6: Assessing Leakages: Per cent of Households Consuming less than 50 kWh per Month, by Consumption Decile and Location: 80.00 70.00 Percent 50 kWh or less 60.00 50.00 40.00 30.00 20.00 10.00 0.00 Welfare Decile Urban HHs Rural HHs Source: VHLSS 2008, staff calculations Affordability 36. Households in Vietnam, both poor and better off, spend a relatively small and decreasing share of household (cash) expenditures on electricity, from of 3.6 percent of expenditures in 2004 to only 3.2 percent in 2008 (Table 5). Urban households spend a higher proportion than rural households and poorer households spend slightly less (as a share of total spending) than the better off. By 2008 electricity shares were remarkably stable across the income distribution (ranging from 2.9 percent in the poorest 10 percent to 3.6 percent for the wealthiest 10 percent of the population). Power is not only widely available in Vietnam but it is also affordable. Table 5: Electr icity Expenditur es as a Per cent of Total Cash Expenditur es: 2004-2008 2004 2006 2008 Location - Urban 4.6 4.3 4.0 - Rural 3.2 3.1 2.8 Poverty Status - Poor 3.1 3.0 2.9 - Non-poor 3.7 3.5 3.2 Expenditure Decile - Poorest 10% 3.2 3.2 2.9 - Decile 2 3.3 3.1 3.1 - Decile 3 3.3 3.3 3.2 - Decile 4 3.3 3.3 3.1 - Decile 5 3.5 3.3 3.2 - Decile 6 3.4 3.3 3.1 - Decile 7 3.5 3.4 2.9 - Decile 8 3.6 3.3 3.2 - Decile 9 4.0 3.8 3.3 - Richest 10% 4.6 4.0 3.6 All Vietnam 3.6 3.4 3.2 Source: various rounds of VHLSS, staff calculations - 11 - DRAFT: Not for Citation Tariff Reform Scenarios 37. Three scenarios were developed to assess the impact of the new IBT structure on affordability for different groups of households (poor/non-poor and urban/rural). These were assessed against a baseline scenario defined as the actual patterns of consumption and related tariffs in place in 2008, but without discriminating which households are served by LDUs and which by PCs. For lack of better information, a unified tariff structure is assumed to apply to all rural consumers both pre and post- reform 11. For all scenarios, electricity consumption is held fixed at 2008 levels 12 and monetary values are adjusted (using GSO’s official CPIA) to March, 2009. 38. Scenario 1 looks at the impact of moving to the March 2009 IBT structure. As anticipated and despite the reduction in the lifeline, the overall impact of tariff adjustments are small – the increase in tariffs were to a large extent offset by inflation (particularly for the lower bands), distributional impacts were small, and the poor paid a still reasonable share of total expenditures for electricity, increasing from 2.8 percent to 2.9 percent of cash expenditures. 39. Scenarios 2 and 3 simulate additional adjustments to IBT lifeline subsidies, looking towards future tariff adjustments. Using the current IBT structure, Scenario 2 simulates a policy to provide tariff subsidies only to rural consumers. This further restriction would have only a small impact on affordability for the poor because a significant majority of the poor currently reside in rural areas. 40. Scenario 3 assesses a telescoped variant of the current lifeline tariff band: only households who consume 50 kWh or less per month are given the subsidized VND 600/kWh lifeline tariff. Households consuming more than 50 kWh per month pay VND 865/kWh – the ‘at cost’ tariff for all of their first 100 kWh electricity consumption. As might be expected, spending on electricity as a share of household expenditures would increase under Scenario 3 – moderately for the poor and households living in rural areas, and more significantly for the non-poor and urban residents. But urban households would still only spend an average of 4.4 percent of household expenditures for electricity, which is modest in comparison to many other countries. Table 6: Baseline and Tar iff Adjustment Scenar ios BASELINE (2008) SCENARIO 1: SCENARIO 2: REFORM SCENARIO 2009, MARCH IBT 2009, MARCH IBT 3: TELESCOPED 2009, MARCH IBT RURAL ONLY kWh % Exp on kWh % Exp on kWh % Exp on kWh % Exp on Cons Electricity Consum Electricity Consum Electricit Consume Electricity umed ed ed y d Location Urban 152.5 4.0 152.5 4.0 152.5 4.4 152.5 4.2 Rural 75.1 2.8 75.1 2.9 75.1 3.3 75.1 2.9 Poverty Status Poor 42.4 2.6 42.4 2.8 42.4 3.1 42.4 2.8 Non-poor 103.9 3.2 103.9 3.2 103.9 3.7 103.9 3.3 Quintile Poorest 49.8 3.0 49.8 3.0 49.8 3.4 49.8 3.1 Quintile 2 70.8 3.1 70.8 3.2 70.8 3.7 70.8 3.3 Quintile 3 88.0 3.2 88.0 3.3 88.0 3.8 88.0 3.4 Quintile 4 106.9 3.1 106.9 3.1 106.9 3.5 106.9 3.2 Wealthiest 170.6 3.5 170.6 3.4 170.6 3.7 170.6 3.4 Vietnam 97.2 3.2 97.2 3.2 97.2 3.6 97.2 3.3 11 This is a strong assumption. At present we do not have sufficient information to identify which households are EVN PC customers and which purchase from LDUs. In any case, the move towards integration and a unified system of rural tariffs will only improve supply and affordability for rural households. 12 Price effects, i.e. reduced consumption due to higher electricity prices, were ignored. - 12 - DRAFT: Not for Citation 4. REACHING THE MARGINAL CONSUMER: MIGRANTS AND OTHER SHORT-TERM RESIDENTS IN POOR NEIGHBORHOODS 13 41. While Vietnam’s tariff adjustments are not likely to have major adverse distributional impacts, there is always a possibility that specific groups will suffer unforeseen losses a result of reform measures. A qualitative study was launched to assess issues related to electricity access, quality of services and affordability for migrant workers and other temporary urban residents who typically live in apartment buildings or rooming houses in densely settled urban neighborhoods close to sources of employment. Only a small proportion of migrant workers and other temporary residents are captured in GSO’s standard household surveys, including the VHLSS. 42. The study was conducted in HCMC (Go Vap and District 7), Hai Phong (Kien An and An Lao) and near the Dang Anh Industrial Zone in Hanoi 14. The research team visited 15-20 residences in each city and interviewed landlords and a range of tenants living in the buildings. Most landlords in these neighborhoods kept detailed records of tenants and related charges (e.g. for electricity and water as well as the room itself). Electricity charges were based on metered consumption; most landlords installed electricity sub-meters in each rental room or apartment, and the sub-meters were connected to main (EVN) meters in the building. With only one exception, all contracts for power supply were between the landlords and EVN, and not with individual tenants or groups of tenants. The study team also met with local officials and other key informants. Key Findings 43. Although the vast majority of urban respondents were indeed low income and consumed no more than 15-20 kWh of electricity (per room), none of them were covered by lifeline tariffs. In Hanoi, landlord-imposed electricity charges ranged between VND 1,000-2,500/kWh, with VND 1,500-2,000 /kWh being the most common charge. Rates were a little lower in Hai Phong and substantially higher in HCMC, where landlord-imposed electricity rates ranged from VND 2,500- 3,500 VND/kWh. In addition to metered charges, tenants also paid for electricity used for pumping water. 44. Although per kWh prices are high, most respondents did not consume enough electricity for high prices to matter: 10-20 kWh per person per month was the norm. Most migrants work all day and use their few electrical appliances in the evenings: a few lights, a fan, and possibly (although rarely) a television. 45. Few landlords had registered their buildings to qualify for lower tariff bands which they said entail complicated procedures and are difficult with the rapidly turnover of tenant contracts. For example, the Go Vap District Electricity Division requires a “green bookletâ€? (the equivalent of KT3 registration) to register 4 tenants as one household and formally to apply the IBT structure. Given the short duration of many rental arrangements, landlords were not willing to file the necessary paperwork and act as specific guarantors. Moreover landlords and tenants alike often said they preferred to keep their arrangements private and informal, rather than involving officials and dealing with many regulations that may not be designed with their circumstances and needs in mind – that is, to ensure affordable rental accommodations with flexible contract arrangements. 46. When asked about recent tariff adjustments and their concerns, few tenants had heard of recent changes to electricity tariffs and none of the respondents noted an increase in electricity charges since March, 2009. Nor were they informed about new supporting policies, such as complaint and grievance mechanisms, although district and ward officials told the team that related 13 Nguyen Tam Giang, A Quick Assessment of the Impacts of the Electricity Reforms on Migrant Groups in the Cities, January 2010. 14 In HCMC interviews were carried out in Go Vap (where many day laborers, students and factory workers reside) and District 7 (accommodating workers from nearby industrial parks); in Hanoi interviews were carried out in Kien An (day laborers, students, factory workers) and An Lao (near an industrial park); in Hai Phong interviews were carried in Dang Anh (near an industrial park). - 13 - DRAFT: Not for Citation information had been widely promulgated in the ward bulletin, mass media, and neighborhood meetings. Tenants who signed rental contracts before March 2009 had not experienced an increase in electricity prices. Respondents expressed general concerns that prices of many goods and services had risen in 2008, were still creeping up, and wages have not kept pace with the rising cost of living in urban areas (similar comments were reported in recent rapid assessment work on impacts of the global economic crisis). 47. Many landlords also claimed ignorance about March tariff adjustments, which is surprising. As a result of IBT reforms, landlords should have witnessed a reduction in electricity bills from EVN PCs because rooming houses and rental properties were moved from the industrial tariff to the residential schedule. 48. Landlords and tenants alike raised concerns about the quality of electricity supply, particularly in the dense Hanoi neighborhoods included in the study. Many cited damage to appliances caused by voltage spikes, and others said the flow of electricity was so weak during peak hours that it took 2 hours to cook rice. EVN PC recently increased the number of transformers in several of the sites visited, but supply still fails to meet the ever growing local demand, including from high numbers of migrants and temporary residents. 5. NEXT STEPS: MONITORING AND SPECIAL STUDIES 49. The PSIA suggests that poor households in Vietnam continue to have good access to affordable electricity despite recent tariff adjustments and the narrowing of the lifeline band from 100 kWh to 50 kWh. Due to the introduction of unified tariffs and integration of LDUs into rural PCs, rural electricity customers are expected to benefit not only from lower prices but also improvements in the reliability and quality of the electricity supply. 50. However, this is an assessment of the situation today – based on the best available information and ex ante analysis. What happens on the ground may be different from ex ante predictions and unexpected impacts, both adverse and beneficial, may emerge as power sector reforms are further designed and implemented. In addition, the situation in Vietnam is changing: incomes will continue to rise, lifestyles will change and Vietnamese households will purchase more appliances, use them more intensively, and consume more electricity. 51. Looking forward, there is a need to: • Monitor the overall implementation of reforms, including quality of service and key performance indicators, with particular focus on vulnerable and poor groups. • Undertake special studies as needed to ensure all households benefit equally from future reforms (e.g. to IBT subsidies), also to help in designing mitigation measures if needed. 52. ERAV, in its role as regulator, has primary responsibility for monitoring performance and system quality, as well as the application of retail tariffs. EVN and other PCs are the regulated entities. ERAV also is responsible for monitoring distributional impacts and undertaking special studies to ensure the poor and other at-risk households are not adversely impacted by ongoing reforms. Monitoring Quality of Service 53. The Distribution Code and Grid Code are a part of the new regulatory framework for the power sector; they are available in draft and awaiting final approval. Once approved, the Distribution Code will establish the quality of service and performance obligations of the PCs. The best way to monitor impacts on specific groups of households (e.g. poor, at-risk) is to ensure the system for overall performance monitoring is well-designed and has good spatial and socio-economic coverage. Also that ERAV develops a timely system of reporting from and feedback to the PCs so they respond in a timely manner to performance problems, including problems in delivering reliable and high quality services to more isolated regions. In additional to standard reporting, additional formats should be designed to identify and highlight performance problems specific to low income areas, both in rural areas and urban neighborhoods. - 14 - DRAFT: Not for Citation Need for Further Special Studies 54. ERAV has requested TA support from the World Bank to undertake a series of special studies to ensure adequate protection for the poor, including the design of mitigation measures and more effective mechanisms to ensure electricity remains accessible and affordable to all households. The current system of IBT lifeline subsidies does a reasonable job of protecting the poor, albeit at the cost of substantial leakages to the non-poor. Over the longer term, ERAV is interested to move away from tariff based subsidies and instead deliver subsidies directly to poor households. This will require a careful assessment and piloting of good practice mechanisms (e.g. targeted social assistance) that are consistent with Vietnam’s poverty reduction and social protection strategies and acceptable to the government. 55. The World Bank is keen to support ERAV’s work in this area. Based on this first round PSIA, we recommend special studies on topics such as: • International practices in targeting and delivering power subsidies through IBT tariffs versus broader SP mechanisms, with particular focus on costs and targeting efficiency. The findings will support more informed policy discussions in Vietnam on the design and implementation of future subsidies. • In-depth study of progress in implementing unified tariffs for rural electricity consumers, including an assessment of progress in LDU improvements and integration. ERAV currently does not have a mandate to monitoring quality of service delivered by LDUs, as they are formally outside the power market. Information on LDU performance is limited. If LDUs continue to supply power to a substantial proportion of rural consumers (particularly lower income consumers) it is important to develop mechanisms to monitor and assess their performance, also ensure compliance with broader reform initiatives. The rural poor should not continue to pay higher prices for low service quality. • More comprehensive data collection and analysis of service delivery and affordability in low income urban neighborhoods, with particular focus on migrant workers and other temporary dwellers. The small study done for the PSIA indicates that migrants and other temporary residents currently consume very little electricity but pay high per-unit prices. They are poor but are not covered under the IBT lifeline. Extending coverage through the IBT lifeline would be difficult; however, migrants and others living in temporary dwelling units should be a part of the target group when considering alternative delivery mechanisms (non-tariff). • Assessment of relative costs and affordability in the context of rising incomes and increasing demand for electricity in Vietnam. Electricity consumption is currently very low, particularly for poor and near-poor households. But demand is increasing across the income distribution and it is important that gains in access and affordability are maintained in the future. Policy makers in Vietnam are particularly concerned about rising inequality, in incomes as well as access to basic goods and services. - 15 -