West Bank and Gaza The Mounting Problem of Arrears in Electricity -- Findings and Recommendations --

FINAL DRAFT West Bank and Gaza The Mounting Problem of Arrears in Electricity -- Findings and Recommendations -- Prepared by Hernando Garzón Washington, D. C. March 4, 28

CONTENTS 1. Introduction 2. Analysis of Empirical Evidence 2.1 Magnitudes and Trend of Electricity Arrears 2.2 Distribution of Arrears across Municipalities 2.3 Relationship to Municipal Financial Situation 2.4 Relationship to Collections 2.5 Retail Sales Collection Performance 2.6 Relationship to Wholesale Payments Performance 3. Classification of Municipalities 3.1 Municipal Financial Performance 3.2 Relative Budgetary Weight of Arrears 3.3 Per Capita Arrears 3.4 Absolute Magnitude of Arrears 4. Ranking of Municipalities and Policy Recommendations 4.1 Best Performance 4.2 Good Performance 4.3 Moderate Performance 4.4 Poor Performance 4.5 Worse Performance 4.6 Consolidated Ranking and Policy Matrix 5. Policy Tools 5.1 Municipality Level 5.2 Central Government Level 6. Main Conclusions and Recommendations 2

ABBREVIATIONS AND ACRONYMS EMSRP GEDCO GOI GPGC GPP HEPCO IEC IMF JDECO LGs MDLF MOF MOLG NEDCO NIS O&M PA PCBS PEA PEC SELCO WB&G Emergency Municipal Services Rehabilitation Project Gaza Electricity Distribution Company Government of Israel Gaza Power Generation Company Gaza Power Plant Hebron Electricity Power Company Israel Electricity Corporation Ltd International Monetary Fund Jerusalem District Electricity Company Local Governments Municipal Development and Lending Fund Ministry of Finance Ministry of Local Government Northern Electric Distribution Company New Israeli Shekel Operation and Maintenance Palestinian Authority Palestinian Central Bureau of Statistics Palestinian Energy and Natural Resources Authority Palestinian Electricity Company Southern Electric Company West Bank and Gaza CURRENCY EQUIVALENTS Currency Unit = New Israeli Shekel (NIS) NIS1 = US$.24 US$1 = NIS4.2 FISCAL YEAR January 1 December 31 3

TABLES Table 1: Yearly Flow of Electricity Arrears 22-26 Table 2: Estimate of Arrears due to Collection Inefficiency 24 26 Table 3: Examples of Diversion of Revenue 26 Table 4: Table 5: Categorization of Municipalities by Financial Performance Categorization by Arrears Relative Budgetary Weight Table 6: Municipal Categorization by Per Capita Arrears - 25 Table 7: Municipal Categorization by Per Capita Arrears - 26 Table 8: Categorization by the Magnitude of the Arrears - 26 Table 9: Table 1: Table 11: Table 12: Table 13: Table 14: Table 15: Table 16: Number of Municipalities by Rank of Performance Classification by Arrears and Municipal Financial Situation Municipal Classification Criteria: Consolidation into Four Categories Policy Recommendations: Consolidation into Four Categories List of Municipalities by Four Ranks of Performance List of the Municipalities by Five Ranks of Performance Summary of Issues, Policy Recommendations and Policy Tools Relationship between Collections and Arrears Table 17: Electricity Collection Efficiency Rates for 24 to 26. Table 18: Diversion of Revenue Associated with Electricity Bills Collection 4

CHARTS Chart 1: Ranking of Municipal Per Capita Arrears 25 Chart 2: Municipal Arrears 25 Chart 3: Municipal Per Capita Arrears 26 Chart 4: Municipal Arrears 26 Chart 5: Ranking of Municipal Arrears 26 Chart 6: Deficits, Surpluses, and Arrears in 25 for 99 Municipalities Chart 7: Deficits, Surpluses, and Arrears in 25 for 39 Municipalities Chart 8: Percent Distribution of Municipal Surplus and Deficit Chart 9: Comparison of Arrears with Municipal Financial Situation Chart 1: Comparison of Financial Situation for Municipalities with Arrears Chart 11: Ranking of Municipalities by Financial Performance Chart 12: Municipal Arrears for the Sample of 99 West Bank Municipalities Chart 13: Arrear and Collection for a Sample of 29 West Bank Municipalities Chart 14: Revenue from Property tax: Distribution by Municipality Chart 15: Expenditure in Salaries: Distribution by Municipality Chart 16: Municipal Distribution of Salaries Per Capita Chart 17: Relationship between Arrears and Central Government Transfers ANNEXES Annex 1: Ranking of Municipal Per Capita Arrears 25 Annex 2: Ranking of Municipal Per Capita Arrears 26 Annex 3: Ranking of Municipal Arrears 26 Annex 4: Deficits, Surpluses, and Arrears in 25 Annex 5: Ranking of Municipalities by Financial Performance Annex 6: Ranking of Municipalities by Per Capita Arrears 22-26 Annex 7: Ranking of Municipalities by Per Capita Arrears Annex 8: Database on Electricity Collections and Payments Annex 9 Revenue from Property tax: Distribution by Municipality Annex 1 Expenditure in Salaries: Distribution by Municipality Annex 11 Distribution of Salaries Per Capita Annex 12 Fiscal Transfers and Arrears Per Capita 26 Annex 13 Brief Assessment on Municipal Revenue Sources 5

ACKNOWLEDGEMENTS This report is based on the findings from my visit to the West Bank from November 26 to December 9 of 27. This review was prepared under the auspices of Mr. Steve Karam, TTL of the Emergency Municipal Services Rehabilitation Project (EMSRP II). The preparation of this document benefited from the views and guidance provided by Mr. David Craig (Director of the West Bank and Gaza Department) and Mr. Joel Toujas-Bernate (Senior Resident Representative of the IMF in WB&G), and from several technical discussions with Mr. Karam. The study would have not been possible without the generous support given by Mr. Yousef Al-Zamer, General Director of Treasury at MOF and his staff. I am grateful to Ms. May Zeyad and Ms. Khawla Ibrahim Ali who worked on the consolidation of yearly arrears. The author is grateful to Mr. Al-Zamer for his time to discuss policy options and his insight to the main issues. Mr. Jamal Abu Ghosh, Director of the Program Monitoring Unit at PEA supplied a sample on collections and payments data, which was fundamental for the analysis of key issues. The MOLG, through Budget Director Mr. Mahmoud Zubaidi, provided information on municipal executed budgets. Dr. Iyad Rammal, General Director of MDLF, along with his staff, contributed with technical and logistical support. In particular, Mr. Mohamed Ramahi, and his staff Mr. Nizar Samhan and Miss Mateel Abuein assisted in merging databases and in managing and processing information regarding electricity arrears. The author also benefited from technical discussion with both Mr. Ibrahim Dajani and Ms. Meskerem Brhane, two World Bank staff members working in the Bank s regional offices. Two undisclosed readers familiar with the energy sector in WB&G contributed valuable comments on an earlier draft of this paper. The logistical support given by the World Bank Field Office was also vital in the timely completion of this task. Needless to say, any error of interpretation or fact is solely the responsibility of the author. 6

1. INTRODUCTION 1. Local Governments (LGs) in West Bank & Gaza (WBG) are legally responsible for, among other things, electricity and water supply, although in practice not all LGs provide these services directly. Most LGs use associations with regional public utilities to provide these services. Service provision refers mainly 1 to electricity distribution. To illustrate, none of the twenty-five municipalities in Gaza directly provides this service. It is the Gaza Electricity Distribution Company Ltda. (GEDCO), that currently distributes electricity for all the LGs in this region. 2. In the West Bank, three companies cover the central and southern areas of the West Bank. These companies are the Jerusalem District Electricity Company (JDECO) in the central area, the Hebron Electricity Power Company (HEPCO), in the southern West Bank, and the Southern Electric Company (SELCO), which covers what remains of the southern West Bank. Also, there are about 2 51 municipalities in the West Bank, out of a total of 17 that directly provide the service of electricity distribution. These municipalities are located primarily in the northern West Bank, where the creation of a new company (Northern Electric Distribution Company - NEDCO) is currently being considered in order to serve this region. 3. The electricity companies and the municipalities buy their energy wholesale from the Israeli Electricity Company (IEC) and take care of sales to customers within their jurisdictions. Currently, however, it appears that in most municipalities and electricity companies not all billing to consumers is done, not all bills are paid, and not all payments are actually allocated to cover the invoices for the corresponding IEC wholesales. Therefore, monthly arrears that occur, either in the four (4) electricity distribution companies, or the fifty-one (51) municipalities in the West Bank, are deducted by the Israeli Ministry of Finance (IMOF) from the Palestinian Clearance Revenue in order to cover any overdue payments to the IEC. Once these total monthly deductions are made, the net balance from the Clearance Revenue is transferred to the Palestinian Ministry of Finance (PMOF). The above issues regarding tariff collection and actual payments will be examined in Section 2, together with other key factors, such as: An apparent fairly weak system of accountability for municipalities, and even for electricity companies, as well as the key element of lack of affordability by some population groups across all municipalities. 4. The issue of arrears (also known as net lending from the Palestinian MOF to both municipalities and utilities) is roughly three times more severe in the electricity sector than in the water sector. In 26, deductions from the Clearance Revenue due to electricity arrears was approximately US$123.1 million, while deductions for arrears due 1 In 25 about 6% of the total supply of electricity in Gaza came from GEDCO, which buys electricity wholesale from the IEC. The remaining 4% was generated by the Gaza Power Plant. 2 Apparently, fifty-two (52) municipalities directly provide the electricity distribution service. However, this report includes only the 51 that have financial information available in MDLF municipal database. 7

to water came to about US$ 43.4 million 3. The issue of electricity arrears seems to be more critical in Gaza than in the West Bank. Out of US$123.1 million worth of deductions from Clearance Revenue, about 55% were due to electricity arrears originated in Gaza, while the remaining 45% came from the West Bank. 5. It can be argued that several 4 factors contribute to the arrears problem. These factors range from the macroeconomic conditions in WB&G (for example, the weaknesses in current intergovernmental fiscal arrangements) to more specific problems at the local level. Previous studies 5 have argued that at the macro-level the share of central government expenditures in GDP has been increasing continuously since 1999, while revenue from external aid in times of economic crisis has gone directly to the Central Government. In contrast, local governments are left to find their own ways to compensate for declining local revenues during periods of economic contraction. The decline in local revenues, and hence on affordability, has been compounded by an apparent lack of financial and fiscal accountability in local governments, along with a spread culture of no-payment -- as reflected by free riders in the system, weak local revenue administration capacity, highly restricted local revenue structure, and fairly limited local autonomy. 6. Finally, this report recognizes and supports the current policy of the PA in the energy sector, on the transfer of electricity distribution from the local governments to regional utilities. It also acknowledges that the implementation process of this policy takes a few years to be implemented, as it has been as of now, and still needs at least two to three more years to be completed. However, given the acute crises of the municipal electricity arrears situation, this paper recommends specific strategy measures intended to serve this sectoral policy at the municipal level, in the interim of this transition. Secondly, this approach is also consistent with a broader municipal finance policy, which aims towards discipline and sound local financial management. Consequently, this report supports local institutional capacity building, which is needed to face the institutional and financial challenges for an adequate provision of municipal public goods and services to foster sustainable local socio-economic growth. 2. Analysis of Empirical Evidence 7. This analysis is based primarily on four different databases. The first database refers to the information on arrears provided by the MOF, which includes the 51 municipalities that appear as direct distributors of electricity in the West Bank. This database illustrates the flows of arrears from 22 to 26 (it does not illustrate the actual stock). These flows are analyzed across municipalities in an attempt to understand the nature of the problem. Attention should be given first to the flow, rather than the stock, in 3 This data is based on MOF s figures and IMF staff estimates. 4 This paper focuses mainly on the arrears of the municipalities and, as such, does not examine arrears of village councils in the West Bank, which may be affecting the arrears situation in some municipalities. 5 The World Bank (26), West Bank and Gaza: Intergovernmental Fiscal Relations & Municipal Finance Policy Note. Report No. 36519 WBG, FPSI/MENA, p.1. 8

an attempt of controlling the stock from getting any worse. Once the current flow is under control and practically eliminated, the next step should deal with the current stocks. This second step might be fairly critical because of the implicit moral hazard that arises if policy actions focus only on the flows -- though, it should be noted that the issue of the stock is not part of the scope of this paper. The second set of data relates to information on executed municipal budgets, provided by MDLF, which covers 99 municipalities 6, out of a total of 17 municipalities 7 in the West Bank and, randomly, includes the above 51 municipalities. These two databases were merged to help analyze the relationship between municipal financial situation and arrears. The third database, provided by the Palestinian Energy Authority (PEA), comprises information on electricity sales, collections, and payments for a sample of sixty-six local entities (including municipalities, village councils, and a few utilities). The fourth set of empirical data emerged from combining the arrears data from MOF with the collections data from PEA, which generated 29 municipalities with information on both variables. These 29 municipalities constitute a practical and random sample of the 51 municipalities providing this service in the West Bank. As noted above, Gaza municipalities do not provide an electricity distribution service. The Gaza Electricity Distribution Company (GEDCO) serves all twenty-five municipalities in that region. 2.1 Magnitudes and Trend of Electricity Arrears 8. The arrears situation became significantly worse in 26, when the actual annual flow proved much larger than the amount initially estimated for the year. Based on MOF 26 data, the total flow of electricity arrears amounted to NIS748.4 million Shekels (about US$168.5 million dollars) 8. This figure is more than one third higher (36.8%) than the estimated amount, which was roughly US$123.1 million dollars. The stability characteristic of previous years ended in 26. The yearly flow had been relatively stable during 24 and 25, fluctuating from NIS199.1 to NIS193.6 millions, respectively; then, in 26, the total arrears increased by almost three times (286.6%), jumping from NIS 193.6 millions to NIS 748.4 millions as illustrated in Table 1. A similar trend, although in a smaller proportion, can be observed for the database of 99 municipalities in the West Bank, which covers the 51 municipalities that provide this service. As stated above, while the magnitude of arrears had been broadly stable during the previous four years (from 22 to 25), in 26 municipal electricity arrears more than doubled, rising from about NIS1 million to NIS35 million (Table 1). The difference between the jumps in the total amount of arrears versus the sample of 99 municipalities is because the total includes the arrears of the electricity companies, which, based on estimated amounts, have increased approximately three times (315%) from 25 to 26 -- as illustrated in Table 1 below. 6 These are the total number of municipalities (99) for which MDLF has information available for the West Bank. The total number of municipalities in Palestine is 132, 25 of which are located in Gaza and 17 in the West Bank. In other words, this is a sample of 93% of the municipalities in the West Bank, and 75% of the municipalities in Palestine. 7 The remaining eight municipalities not covered by the sample include the following: Saltif, Ramallah, Al- Bireh, Jenin, Al-Yamoun, Aluja, Al-Newe emeh and Beta. 8 Based on June 3, 26 exchange rate of US$ 1 = 4.44 NIS. 9

Table 1 Yearly Flow of Electricity Arrears In millions of NIS Years 22 23 24 25 26 Total Arrears 122.8 99.6 199.1 193.6 748.4 Yearly changes (%) n.a. - 18.9% 99.8% - 2.77% 286.6% Arrears in a sample 9 94.3 83.2 8.6 98.5 353.7 of 99 Municipalities Yearly changes (%) n.a. - 11.8% - 3.13% 22.2% 259.1% Utility Arrears 1 28.5 16.4 118.5 95.1 394.7 Yearly changes n.a. -42.5 622.6-19.7 315% Source: Ministry of Finance (MOF) and calculations of the author. 9. As noted above, it has been argued in previous studies 11 that the deterioration in the WB&G economic financial situation in 26 may have contributed to the increasing level of arrears in electricity distribution companies. In 25, the collection rates among utilities fluctuated between 66% at GEDCO to over 1% at JDECO. The cases in between were HEPCO, with a 98% collection rate, and SELCO, with an 82% collection rate. In contrast, during the first 1 months of 26, it was reported that GEDCO only managed to collect about 29% of the total amount billed to customers. The SELCO collection rate was even lower, at about 2% 12. It is clear that the financial viability of some of the electricity distribution companies has grown dire. However, apart from the macroeconomic factors, which may have contributed to the arrears issue in general, (i.e., for both municipalities and utilities), other determinants will be examined below, with a focus on the municipalities in order to identify specific short-term policy options. 2.2 Distribution of Arrears across Municipalities 1. In 25, the flow of municipal arrears on electricity involved about 39 municipalities out of the 51 that have been reported as direct providers of this service (Annex 3 and Chart 3). However, in 26 when, among other factors, the Palestinian economic situation deteriorated, twelve (12) municipalities (Annex 5) that had no flow of arrears in 25 joined this group and five (5) municipalities that had a flow of arrears for 25 were reported with zero flow in 26. This yields a net increase of seven (7) municipalities from 25 to 26, bringing its total from 39 to 46. This total of 46 plus the five (5) with no arrears makes up for the total of 51 municipalities that act as service providers. It should be noted that Hebron is one of the municipalities with arrears in 26 and, as such, it appears as a direct service provider, despite the fact that Hebron Electricity Power Company (HEPCO) is the utility that serves this area. With the exception of Hebron, arrears for utilities are usually accounted under each regional company, and not under the municipality. In any case, the analysis focuses on these 51 municipalities, which are 9 As explained above, the arrears in this sample come from any of the 51 municipalities that provide the service. 1 Utility arrears are estimated as the difference between the total arrears minus the municipal arrears. 11 Ibid. p. 1 12 The World Bank (May 27). West Bank and Gaza: Energy Sector Review. Table 4, p. 23. 1

located mainly in the northern region of the West Bank where, as already mentioned, the creation of a new company (Northern Electric Distribution Company - NEDCO) is currently being considered to serve this region. Not until them, the issue of arrears needs to be addressed in the immediate term. 11. In terms of policy implications for this group of 51 municipalities, it must be noted that the transfer of electricity retail from municipalities to a regional utility is viewed as settled policy and is currently being implemented. It is, however, important to recognize that there have been delays in the transfers of service, including not only NEDCO, but also SELCO and HEPCO. There is clearly a real need to address the issue of municipal arrears in the interim term, as is recommended in the sections of this paper regarding policy options. This is also particularly important because the expectation of the transfer of the service may exacerbate municipal payment discipline -- assuming that the trend between 25 and 26 is not already due in part to these expectations. The implicit moral hazard of this transfer needs to be eliminated with concrete actions at the municipal level to rule out any expectations of receiving the potential benefits of a free rider in the system. Each municipality should be made fully liable for collection and payment up until the date at which these services are transferred to a regional level. 2.3 Relationship to Municipal Financial Situation 12. Data for 26 suggest a direct relationship between the magnitude of the municipal operating deficits 13 and the magnitude of the arrears. This means that, in general, the larger the operating deficits are, the larger the arrears are (Charts 3 and 4). This is the case, for instance, in Nablus, which has an operating budget deficit practically equal to its arrears 14. A Nablus deficit of about 135% (equivalent to NIS33.5 millions Annex 1) has arrears of about NIS34 millions (Annex 3). The municipalities of Qabatya and Seelet-Ad- Daher also have higher deficits and larger arrears (Annexes 3 and 4). Nevertheless, one should know that there are a few municipalities that have surpluses in their operating budgets and still have relatively large arrears. Among them are: the Ya bad municipality, which has arrears of NIS6 millions along with a surplus of about NIS2.8 millions; Tulkarm, with a surplus of NIS24 millions in 26 and arrears for NIS37 millions in the same year; and Dura, where the surplus was NIS1.3 millions and its arrears were worth NIS7.7 millions. So, although arrears and deficits seem directly correlated, it cannot be argued that deficits are the main cause of the arrears. Further analysis of other potential factors causing the arrears can be found below. 13. It should be highlighted that those municipalities with both surpluses and arrears are in fact choosing not to pay their bills i.e., these municipalities could simply use their surpluses to make payments that will lower their balances of their electricity bills. This situation seems to suggest that, among other related causes, there is a fairly weak system of local financial accountability. As it appears, the problem of affordability of 13 Computed as cash basis deficits, rather than as accrual basis deficits. 14 It should be noted that the arrears data does not come from cash basis executed budgets, but from accrual basis accounting and, as such, arrears are not included in cash basis executed expenditure budgets, or in the computation of cash basis deficits. 11

electricity tariffs (inability to pay) by some residents (the poor), as well as the apparent weakness in local institutional capacity regarding tariff collection, seems to be compounded by an unwillingness to pay, not only by some residents (i.e., free-riders ), but also by some LGs. This is because of weak or nonexistent -- enforcement at the local level, and weak or nonexistent -- accountability to the Central Government/MOF. In terms of policy implications, strengthening financial accountability will be considered among the policy options. 2.4 Relationship to collections 14. For policy purposes, one basic question that has been examined is the actual relationship between arrears and collections. Based on empirical evidence, it can be argued that across municipalities there is a direct relationship between these two variables. Hence, for each individual municipality it could be expected that the greater are the collections, the smaller might be the arrears (r = ). However, from a cross section analysis the greater are the collections, the greater are the arrears, since their absolute magnitudes are generally related to the potential for either collections or arrears. So that, it might be argued that the arrears problem, among other factors, may be due to high collection inefficiencies, or, if collections were relatively adequate, the arrears could be due to potential diversion of revenue from electricity payments. The arrears problem could also be a combination of both the potential diversion of revenue and high collection inefficiencies. 15. An empirical analysis was carried out based on a sample 15 of 29 municipalities, for which data was available on both of these variables electricity 16 bills collection and arrears. A correlation analysis was run for this sample which confirmed a direct relationship between bills collection and arrears in the cross sectional analysis. The results are included in Table 18 below. The correlation coefficient for these two variables indicates a positive relationship between them (r =.94). The absolute magnitude of this coefficient suggests a close, positive relationship between arrears and collections. It can generally be argued that the greater the collections, the larger the arrears across municipalities, or, equivalently, the higher the arrears, the larger the collections across municipalities (Chart 13). This result clearly suggests the need for policy makers to make it a priority to address this issue in the larger municipalities where the arrears problem is more concentrated, where the impact of improvements in collections could be much greater. The sections below examine collection efficiency rates and possible diversion of electricity revenue, which could be aggravating the arrears problem. 15 The Palestinian Energy Authority supplied a sample of 66 observations including municipalities, village councils, and some utilities. However, as noted in the text, this data must be considered as preliminary, since it is still subject to verification. Six observations were eliminated. 16 The Palestinian Energy Authority supplied the data on bills collection, which must be considered as preliminary, since it is still subject to verification, while the Ministry of Financed (MOF) provided the data on arrears based on IEC records. 12

2.5 Retail Sales Collection Performance 16. There is a negative trend in collection efficiency 17. Payments of monthly bills were examined for a three-year period (from 24 to 26) and the results show that the average collection rate is dropping. Based on a net 18 sample of 6 local governments (including municipalities and Village Councils) it was found that while the average collection efficiency rate was 76.82% in 24, it gradually dropped to 61.72% in 26. This finding clearly indicates that one of the causes of the arrears problem is major inefficiency in monthly bills collection, which has been deteriorating over the most recent years. In principle, inefficiencies in collections may be due to inability to pay, inefficiencies in collection administration, and other factors, such as low collection effort for political reasons. Hence, it could be argued that arrears due to different collection problems have increased from 23.18% to about 38% during this three-year period (Table 12 below). 17. These results suggest not only the obvious importance of strengthening the municipal bills collection capacity, but also the need, among some population groups, of cross and direct subsidies to be able to afford the monthly payment of the electricity bill. The apparent lack of ability to pay, combined in some cases with weaknesses in the collection system, has increased, as stated above, from 23.18% to 38% between 24 and 26. As such, the importance of specific policy action to strengthen the collection system, the re-design of some of the tariff structures to allow for internal cross subsidies, and the implementation of a program of direct subsidies to the poor should be considered. 18. Lastly, other factors beyond a weak local institutional capacity and the inability to pay may be associated with the arrears issue, like, for example, the lack of political will of the administration to enforce collections, as well as potential diversion of revenues from tariff collections. The latter will be examined below. Specific policy measures are recommended in the policy section to deal with these factors. 17 Collection efficiency is computed as the ratio of actual collection with respect to receivables from sales (billing). 18 Three electricity companies were exclude from the larger sample of 66, as well as two observations characterized as outliers, and one LG, for which there was no complete data. 13

Table 2 Average Collection Efficiency Rate Based on a sample of 6 LGs 24 25 26 Average Collection 76.82% 69.24% 61.72% Efficiency Rate Arrears due to inability to pay, collection inefficiency, 23.18 3.76 38.28 and other factors. Total 1 1 1 Source: Table 17 and Annex 3. 2.6 Relationship to Wholesale Payments Performance 19. It has been argued above that part of the arrears problem may not only be due to low collection rates, but also because of revenue diversion from electricity collection for the financing of municipal deficits, including capital expenditure programs. In order to determine whether there is revenue diversion, this section compares the amounts of collections from retail sales of electricity to the actual payments made to the IEC. This analysis has been done using the same sample of 66 observations. Clearly, if the difference of collections minus payments were positive, this would indicate that the corresponding difference has been diverted. The empirical results confirm that part of the arrears problem is due to diversion of revenue from electricity collections. Almost onefourth of the municipalities (about 23.8% of them) have diverted revenue from electricity in the three-year period between 24 and 26. The difference has been positive for 15 out of a total of 63 LGs. In monetary terms, about 26% of the revenue that was collected from electricity payments was diverted, which clearly is not a small proportion. In the sample, this diversion amounted to NIS556.6 millions (equivalent to US$125.3 millions). In 26, for instance, Jenin and Nablus diverted NIS18.5 millions and NIS6 millions respectively, while some utilities, such as JEDCO and HEPCO, diverted NIS65.7 millions and NIS87.6 millions respectively. These results are included in Annex 8. As it appears, it can be argued that, based on empirical evidence, policy measures also need to address the issue of revenue diversion among both municipalities and utilities, as this diversion is clearly contributing to the current arrears situation. 2. Some mayors have apparently been reluctant to join regional electricity distribution companies because the opportunity cost of the revenue forgone, if they decided to join. As such, policy options need to consider the specific actions to address both the apparently large diversion of revenue and the deficiencies in tariff collection. The community of donors, through MDLF and in coordination with MOF, should assist in addressing the problem of revenue diversion, particularly when these resources are being used for local development programs. Short-term assistance, in the form of matching grants for local development, should be considered and made conditional to the reallocation, in full, of the electricity revenue, on a permanent basis, to the exclusive 14

payment of electricity bills. The diversion of funds creates a soft budgetary constraint for some municipalities, thus compromising fiscal discipline. 21. In contrast to the above diversion of revenue, it must be highlighted that empirical evidence has also shown that most LGs (about 69.8% of those in the sample) have diverted revenue in the opposite direction. This means that most local governments have actually used other revenue sources to contribute to the payment of the wholesale electricity bill. The difference between collections minus payments is negative 19 for 44 LGs out of a net total of 63 of them. These contributions have not been large enough to cover the total value of the bill; instead, they have been fairly small, equivalent to 5% of the total payments to the IEC (NIS83 millions or, approximately, US$18 millions). A small proportion of LGs (about 6.3%) are clearly the exception to the rule, having allocated the entire amount of electricity collections to the payment of the electricity bill. Needless to say, the difference between the two variables is zero for these local governments, which includes only four of them. This happens in cases in which the collections possibilities were so limited that local governments actually prefer to use other revenues to contribute to the payment of the wholesale electricity bill. If this were the case, it would indicate again, to policy-makers, that there is a need to use cross and direct subsidies to help consumers with the monthly payments. On this regard, it should be noted that some central government officials estimate that about 2% of the municipal electricity bills may need to be subsidized by MOF. Table 3 Examples of Diversion of Revenue 26 Municipality Collection Payment Difference NIS NIS NIS 1. Nablus 79,115,233 73,84,191 6,31,42 2. Hebron/HEPCO 65,699,73 65,699,73 3. Qalqilia 12,686,948 5, 12,186,948 Tulkarem 17,863,764 3,129,85-12,266,41 Source: Table 18 included at the end of this Report. 22. It is worth highlighting the situation of four particular municipalities that owe the largest amounts of arrears, and which represented more than half of the total arrears in 26. In three of these four cases (Nablus, Hebron, and Qalqilia) it appears that there was a diversion of revenues. This is in strong contrast to the fourth municipality, Tulkarem, which has made a payment larger than its own collection, though not large enough, despite its operating surplus, to be able to pay its wholesale bill in 26. These findings suggest that in the transitional period of the service to a utility, different policies may be needed to apply to each type of LG, depending on, among other things, their actual financial situation and their behavior in regards to the handling of their wholesale electricity bill. 19 The opposite transfers (i.e., the negative figures) are indicated in parenthesis in Annex 8. 15

23. The case of Hebron and HEPCO needs special attention and more in-depth examination going far beyond the scope of this review. Based on MOF/IEC s data, the Hebron Municipality owes one of the largest amounts of electricity arrears, even though, under the current institutional arrangements, the Hebron Electricity Power Company (HEPCO) is the only provider of electricity distribution in this municipality. Therefore, it is not clear why the Hebron municipality is reported by MOF/IEC as having electricity arrears in 26. 24. Additionally, based on the database on collections, the figures indicate that HEPCO did not make a single payment for the purchase of wholesale electricity from the IEC in 26. The reported amount paid by HEPCO in the PEA s survey is zero. However, in the same database it is reported that HEPCO made collections worth NIS65.6 million (Table 3 above). It seems as though HEPCO handed over all its collections to Hebron municipality, which has been reported to owe arrears to MOF/IEC worth NIS 19.2 millions. These findings seem to suggest a mixed budgetary situation between these two entities. This lack of clear-cut responsibilities divided between Hebron municipality and HEPCO does not appear to be common practice, but rather the exception to the rule among other electricity utilities and the municipalities in the West Bank. Though this issue is not reflected in the database as having an effect on other institutions, it should be further investigated in order to determine specific sectoral policy recommendations for such cases. 25. Summarizing, empirical evidence supports both a low collection efficiency rate, including its negative trend, and a high diversion of electricity revenue, as two contributors to the arrears issue. The policy implications of these findings are clear: there is a need to enforce payment compliance to bust collections. However, this policy needs to be complemented with cross and direct subsidies to be fully effective. Also, there is a need to address diversion of electricity revenue, which should be approached through some form of municipal fiscal restructuring program (MFRP), since this diversion of revenues, as it appears, is already ingrained in the local fiscal system -- in both the operating and development budget. These policy options will be considered in more detail in the policy section. 3. Classification of Municipalities 26. This section examines both the fiscal situations and the distribution of arrears across municipalities. The goal of this examination is to create a classification system for these variations. This would facilitate the design and application of specific policy measures tailored to each category of municipality. The criteria that would be used include: (i) indicators of municipal financial performance, (ii) the relative weight of arrears with respect to the municipal budget, (iii) the level of municipal per capita arrears, and (iv) the absolute amount of municipal arrears. Each of these classifications is further analyzed below. 16

3.1 Municipal Financial Performance 27. As discussed above, one of the main problems associated with arrears is the current fiscal situation of each municipality. Further examination 2 of current surpluses and/or deficits in the operating budget (for the sample of 99 municipalities, including the 51 municipalities currently providing the service) has been carried out to determine each municipality s financial performance; this data will help determine the relevant policy measures for each group according to their particular circumstances. These results were subsequently merged with the database on arrears for the 51 municipalities. 28. In 26, results indicate that 11.8% of the municipalities are running an operating surplus larger than 1%. This group of municipalities has been classified as Category A -- in Table 4. The next group, Category B, includes 5.9% of the municipalities, which are all characterized by having operational budgets that are more or less balanced. This group includes three municipalities with surpluses or deficits that are less than 1% (and therefore considered relatively small). In the next group, Category C, seven municipalities (i.e., 13.7% of them) have deficits greater than 1% t and no larger than 5% of their actual operating revenue budget. Most of the municipalities, about 68 percent of them, are in Category D and E characterized by deficits greater than fifty and one hundred percent, respectively. In some of the municipalities, deficits are fairly large, sometimes reaching magnitudes greater than twice their actual operating revenue (i.e., deficits greater than 2%). These municipalities, which are in a very critical financial situation, have been classified as Category D, in Table 4 below. Table 4 Categorization by Municipal Financial Performance: Deficit or Surplus In Operating Municipal Budgets 26 Municipal Categories RANKING Number of Providers A Surplus greater than 1% 6 11.8 B 21-1% < balance < 1% 3 5.9 C 1% < deficit < 5% 7 13.7 D 5% < deficit < 1% 16 31.4 E 1% < Deficit 19 37.2 Total 51 1. Source: Based on current municipal executed budgets Annex 2 29. Several policy implications emerge from these results. It is logical that those municipalities having both arrears and a surplus in their budget should use that surplus to make payments toward any outstanding balance. They should also implement a set of % 2 The analysis of the financial/fiscal situation is based on executed municipal budgets. 21 The limit of 1% could be modified to 15% to make these categories broader, which would not compromise the proposed policy measures for each of these categories. 17

standard measures, such as those collection tools described in Section 5.. In contrast, those municipalities with very large deficits, and presumably with very large arrears, should enter into a program of municipal fiscal restructuring (MFRP). This will make it possible for them to substantially reduce their current level of arrears. Category B (characterized by having a relatively balanced budget) may only need to set targets over a specific time period so that they can gradually eliminate the flow of arrears. This category should also adopt standard measures, including, at the very least, the following: (i) Upgrading of collection and enforcement systems, (ii) redesign of the tariff structure to include cross subsidies, (iii) gradual elimination of any diversion of revenue, and (iv) implementation of a system of direct subsidies to the very poor. The fourth measure can be completed with the financial support of the Central Government. The third and fourth groups (C and D)), characterized by deficits less critical than those in the worst group of municipalities (E), may need to implement standard policy measures and to conduct some fiscal restructuring. The analysis of arrears, in the following section, refers to the 51 municipalities currently providing this service in the West Bank. These 51 municipalities form part of the total sample of ninety-nine (99) municipalities. 3.2 Relative Budgetary Weight of Arrears 3. In 25, most of the flow of arrears was owed by municipalities in Category C (Table 5 below) characterized by having a relatively moderate weight in proportion to their total revenue budget. About 74% of this flow was in the hands of 27.5% of the municipalities (14 out of 51) having a relative budgetary weight ranging from 2% to 5% of their total budgets. This means that the largest proportion of arrears is affecting municipalities at levels that appear to be financially manageable. The classification shows that a small proportion of arrears are affecting municipalities at levels that will be more difficult to manage, such as those in categories D and E, with relative weights as high or higher that 1% of the total municipal budget. In fact, only 1% of the flow of arrears falls in Category E, affecting three (3) municipalities out of the 51. 31. The fact that different municipalities face different levels of difficulty when it comes to handling the arrears problem justifies the need to classify them in varying levels. Each category of municipalities, depending on their level of difficulty, will need to take a different approach toward paying off their arrears. For example, Category B encompasses those municipalities with smaller arrears, with a relative budgetary weight of less than 2%. Though Category B holds only 4% of the yearly arrears, it includes the largest number of municipalities, which is equivalent to 29.4% of them. The municipalities with no arrears have been classified under Category A. Twelve 22 of them are included in this category since they had zero arrears. Categories C and E were described above, while category D represents poor performers, which is a considerably difficult group, given the large magnitude of their arrears in proportion to their budgets. 32. Considering that these 14 municipalities make up 74% of the total municipal arrears, this group should be, in principle, be the main target of a program to eliminate the flow of arrears. These municipalities will need to implement changes to their current 22 Twelve municipalities that showed yearly arrears in 26 had zero yearly arrears in 25. 18

revenue and expenditure structures in order to address this problem. Municipalities in this category should be required to submit a financial plan within a given time frame to eliminate the flow. These plans should specify the targets to be met over a period no longer that 12 months and TA from MDLF should be provided as a condition to benefit from MDLF funds. For municipalities in Category D, a Municipal Fiscal Restructuring Program (MFRP) should be mandatory. For the three municipalities in Category E the removal of the service to a separate authority should be required. This is the case in the municipalities of Beit Ula, Qubatya and Seelet Ad Daher, in which the arrears are larger than their total budgets, and as such will be the most difficult to manage. Performance Category Table 5 Relative Budgetary Weight of Municipal Arrears and their Distribution by Category 25 RANKING As % of Total Revenue Budget Number of Munis. % Yearly Arrears NIS A No arrears 12 23.5 - B Arrears < 2 % 15 29.4 3,823,143 3.9 C 2% < Arrears < 5% 14 27.5 72,845,812 73.9 D 5% < Arrears < 1% 7 13.7 11,947,866 12.1 E 1% < Arrears < 12% 3 5.9 9,9, 1. Total 51 1. 98,516,821 1. Source: Municipal executed budgets and MOF data on arrears in Annex 4. 33. Though municipalities in Categories D constitute a second priority to those in Category C, a broad municipal fiscal restructuring plan (MFRP) would be needed to address the severity of their arrears issue. A detailed fiscal and financial analysis of each of these seven local governments would need to be carried out directly on the ground in order to ensure an adequate and viable MFRP design (e.g., specific components, targets, gradual adjustments in their expenditure 23 and revenue 24 structures, together with the support of intergovernmental agreements with MOLG/MDLF and MOF, among several other key elements). The MOLG, in coordination with the MOF and with the support of the MDLF, would be the key institution when it comes to moving the MFRPs forward. % 23 For instance, it is interesting to note that similarly to arrears, a few municipalities have extremely large per capita expenditures in salaries; such as: Nablus, Tulkarm and Hebron. The latter is not among the three largest, but its per capita expenditures in salaries are still greater than twice as much the national average. These three municipalities, in fact, happen to have the largest per capita arrears (Annex 9, and Chart 9). These findings clearly suggest the need of some structural adjustments in the expenditure structure of the relevant municipalities to be able to resolve the arrears problem. 24 For instance, the per capita revenue from the property tax for many municipalities can be practically zero, (less than one Shekel); while for others it can be as high as three times the national average or even higher. The highest per capita property tax is about 7 Shekels. For Nablus and Tulkarm, for instance, the per capita property taxes are 4 and 28 Shekels, respectively (Annex 11 and Chart 11). These findings, as a whole, also suggest the need, and the potential, for significant structural adjustments in the revenue structure of the relevant municipalities, to be able to curb the yearly arrears. 19

3.3 Per Capita Arrears 34. Based on 25 data, it can be argued that there is a very wide range of variation in per capita arrears in different municipalities. These arrears can be anywhere from 1 shekels to 39 each year, such as in NbaniN eim and Tulkarm municipalities, respectively (Annex 5 and Chart 5). In addition, it should be pointed out that most of the arrears were concentrated in very few municipalities. In 25, the three municipalities with the highest per capita arrears were liable for 3 percent of all the municipal arrears as shown in Table 6 below. These three municipalities are: Qabatya, Seelet Ad Daher and Tulkarm. 35. This concentration of arrears in only a few municipalities can be illustrated as follows: About 25% of them (13 municipalities) owe 8% of the arrears. They are classified in Categories D and E in Table 6 below, both of which correspond to those municipalities with the highest arrears per capita. These two categories include the three municipalities with the highest arrears in absolute terms, which are: Nablus, Tulkarm, and Qabatya (as shown in Chart 6). In 25, Nablus represented 34.9% of the absolute arrears, while in Tulkarm, the arrears were equivalent to 21.8%. In other words, two municipalities concentrated more than half (56.7%) of all the municipal arrears. If Qabatya were added to these two, the arrears of these three municipalities amounts to 63% of the municipal arrears problem. 36. The main conclusions and policy recommendations that emerge from the above data may be summarized as follows: first, strategic plans to tackle the arrears problem must be tailored to different types of municipalities. The degree of variation from one municipality to another may be extremely large, not only in per capita terms and their absolute amounts, but also in terms of the municipal fiscal and financial capacities available to deal with this problem, as discussed above. Secondly, attention should be focused primarily on the thirteen municipalities responsible for 8% of the municipal arrears. Third, out of this group of thirteen, those with the highest absolute arrears within the group should, strategically, be the first priority for the design and implementation of a Municipal Fiscal Restructuring Program (MFRP), particularly for the three municipalities that owe 63 percent of the arrears. The rest of the municipalities in categories D and E (about ten of them) should be the next priority. Fourth, the main criteria to decide which municipality should be a priority within any category should be the absolute amount of their arrears, assuming that the objective is, as it should be, the eradication of the flow of arrears in a timely and efficient manner. In Category C, for instance, the highest absolute amount is owed by Qalqilia, which represent 45% of the arrears of this category and, as such, should be a priority. However, as pointed out above, other factors may need to be taken into account, such as the financial situation of the municipality and the relative weight of the arrears in comparison with the size of the municipal budget. 2