Analysis of Low-Income Benefits in Determining Cost-effectivnes of Energy Efficiency Programs

April 14, 1999

I. INTRODUCTION

The benefits of investing in efficiency measures for low-income homes go far beyond the value of energy, capacity, transmission and distribution costs that are avoided as a result of the investments. Although some benefits are more readily quantifiable than others, it is certain that none of the benefits amount to zero. Accordingly, equity demands that all benefits be accounted for in some manner.

Some of these non-energy benefits are relatively easy to quantify. Others will be quantifiable only by approximation and with difficulty. Still others will be extremely difficult to quantify in a meaningful way. We therefore suggest that, after reasonable efforts at quantification are undertaken, the focus be turned to development of an "adder" that encompasses all that has been learned without embarking on an endless search for unnecessary precision. Non-energy, non-environmental benefits can be computed to be at least equal to energy benefits, although we propose use of an adder that is much smaller.1 The non-energy benefits that have been identified include both utility benefits and societal benefits.

This analysis follows two parameters. First, is that the benefits identified are in addition to general economic and environmental benefits which accrue to society as a whole. Using an incremental value analysis for low-income benefits results in a calculation that does not double count these broader benefits and results in values which are in addition to those values.

II. UTILITY AND ENERGY SYSTEM BENEFITS

Utility companies incur a range of costs that may be avoided through implementation of effective energy efficiency programs. Among the most quantifiable of these benefits are reductions in payment-related costs that utilities incur. In some cases, these represent transfer payments from non-low income customers to low income customers. These have value to reduce the negative effects of inter-customer transfers but generally should be excluded from a societal analysis. In addition, energy efficiency programs can serve to (1) reduce the number of emergency calls, and (2) from a utility perspective, retain customers who contribute to a utility company's fixed costs.

A. Payment-Related

This section thus examines the effect that energy efficiency programs can have on mitigating the broad range of costs that utilities and their ratepayers incur as a result of unpaid bills by low income customers. Although these costs exist in unpaid bills by non low income customers, the burden falls most heavily on low income customers where bill payment is related to ability to pay and therefore can be effected by energy cost reductions. These benefits have been excluded from calculation of general non-energy benefits to avoid duplication. These costs include arrearages and late payments,2 bad debt,3 credit and collection expenses,4 termination and reconnection costs,5 negotiation of payment plans,6 regulatory expenses.7

Literature review reveals considerable variability in the estimates of payment-related benefits stemming from utility DSM programming.8 Further, some of these costs and the mitigation benefits associated with energy efficiency programs have not been quantified. We use the information available to develop an appropriate range of adders where feasible and provide qualitative discussion in those instances where benefits have not been well quantified.

Effective programming may include not only the installation of measures to enhance efficiency and comfort, but also counseling and other informational assistance geared toward removing barriers to prompt payment. In comprehensive low-income energy programs, both installation services and information services are provided to customers to increase their ability to reduce and manage their energy-related budgets. For example, Wisconsin Public Service Co. found that adoption of a system of Customer Assistance Advisors resulted in maintenance of write-offs at around 0.25 percent of revenues, compared to an industry average of 0.51 percent. In addition, disconnections dropped to 24 per 10,000 compared to an industry average of 422 (and its own previous rate of around 120 per 10,000).9

Further, the Wisconsin study found that the majority of customers in arrears want to pay their bills but lack the resources and/or skills to successfully achieve this. In addition, researchers discovered that: (1) it is in the utility's best interest to begin discussions with customers before arrearage problems get too large, (2) individualized attention to the customer is very important, and (3) the utility can achieve positive results by assuming a "service agency" role and provide assistance to customers in bill management and payment.10

1. Reductions in arrearages and late payments

Although it is widely understood that the energy cost burden on low-income households is proportionately three to four times greater than the burden on non-low-income households,11 studies show that less than one half of utility arrearages are actually attributable to low-income customers.12 Effective energy efficiency programming targeted to low-income households in arrears can ease the financial burden and thereby increase the ability of low-income families to make utility payments.

There are two components of costs related to unpaid utility bills. One is the cost of non-payment of bills that is recovered through rates. This is a transfer payment from a societal point of view, but a benefit from the utility and ratepayer perspectives (Energy System Test). All items that might be considered transfer payments from a societal point of view are so indicated in the attached table by an asterisk. The second is the administration of these non-payment-related activities (shut-offs, payment plans, etc). These costs can be avoided through low income weatherization and therefore represent an economic benefit to the overall regulated energy system.

Despite the fact that less than one half of total arrearages have been found to be attributable to low-income customers, the arrearage reduction benefits associated with energy efficiency programs should be tabulated in the cost effectiveness analysis of low-income programs. Low-income customers are more likely to be in arrears due to lack of funds with which to pay utility bills than are non-low-income customers. Since studies show these customers want to pay their bills if they can,13 DSM measures that release funds are more likely to result in arrearage payments from low-income customers than from others. Thus, DSM programming designed to reduce consumption and thereby free up funds that may be applied toward outstanding bills is more likely to be successful when targeted to low-income households.14

Quantification of the scale of arrearage reduction and therefore the benefits associated with reducing related administrative costs has been accomplished through survey analysis of participating and non-participating households. A review of studies of arrearage reduction benefits conducted for the Boston Edison Settlement Board by the Tellus Institute indicates that energy efficiency programs generate reductions in arrearages ranging from $0 to $469 per participating household.15 In the "Progress Report of the National Weatherization Assistance Program," for example, the authors found a reduced arrearage value of $32 per weatherized low-income household relative to program costs of $1,550.16 These actual arrearage reductions represent a transfer payment when they are written off as an uncollected debt. However, the administrative costs with the collection costs may be substantial, generating a non-energy adder of 2.1 percent.

Similarly, a recent study of a Pacific Gas and Electric low-income weatherization and education program found that reduced carrying charges on arrearages range between $4 and $63 per weatherized household. Based on reported program costs of $719 per weatherized household, a non-energy benefit adder range of 0.6 percent and 8.8 percent is justified.17

2. Reductions in uncollectibles and bad debt write-offs

A number of studies demonstrate that utility DSM and customer relations programs significantly reduce the level of utility uncollectibles and bad debt write-offs. Such programs also cut costs associated with collection, termination and reconnection, negotiation and administration of payment plans, tracking hardship accounts, other administration, regulatory response, and complaint resolution.18

In Colorado, for example, write-offs dropped 18 percent at weatherized homes. Further, arrearages dropped 26 percent, emergency gas assistance calls dropped 74 percent, and bills were reduced 22 percent. Total annual benefit to the utility is estimated at $30.56 per participating household on a $2417 per household cost, not counting reductions in complaints and collection costs, increases in comfort and health, and increases in discretionary income.19 The present value of these benefits is $204.72 per participating household.20 The Colorado reduction in payment-related costs thus generated a non-energy benefit adder of 8.47 percent.

Another study found that all benefits associated with reduced uncollectibles range between $16 and $58 per weatherized household.21 With the reported total program cost of $719 per household, this benefit estimate produces a range of avoided cost adders of 2.2 percent to 8.1.

These are transfer payments from a societal point of view and benefits from a utility/ratepayer perspective.

3. Reduced Collection Costs

In a 1994 analysis, Roger Colton found that utility companies incur significant costs associated with collection activities, including telephone contacts and premise visits with customers. He further found that implementation of low-income DSM programs generates substantial utility collection-related expense savings.22 In testimony before the Pennsylvania Public Utilities Commission, Columbia Gas Company reported the following costs associated with each instance of the various collections activities:23

This table does not reflect the costs associated with collection and credit agency fees. Since these entities usually work on a commission basis, it may be assumed that the costs reflected in the above table would be higher were collection agency fees to be included.

Colton's analysis does not include estimates of program costs that would be incurred to mitigate the reported collection expenses. We therefore do not provide an estimate of an adder to be applied to the non-energy benefits identified by Colton.

1. Reduced Termination and Reconnection Costs

Another set of utility and ratepayer costs avoided through implementation of DSM programs is the processing and distribution of shutoff notices, as well as the disconnection and reconnection of customer accounts. The table below is based on the Pennsylvania PUC testimony mentioned above, and reflects the costs associated with each instance of the reported activity.24

The identified costs refer only to those incurred by the utility company, and not to the customer's costs. As noted previously, Colton does not estimate the program costs that would be incurred to mitigate the reported collection expenses. We therefore do not provide a non-energy benefits adder based on his analysis.

However, Skumatz, et al. quantified avoided utility costs that may be generated through implementation of DSM programs, including notices, customer calls, and termination and reconnection costs. The Skumatz study reported these costs to range between $2 and $12 per weatherized household.25 Under their reported total program cost of $719 per household, a range of avoided cost adders of 0.3 percent to 1.1 percent accounts for this set of non-energy benefits.26

1. Reduced Costs of Negotiation, Administration of Payment Plans, Complaint Resolution and Tracking

The Columbia Gas Company reported that it incurs significant costs in negotiating payment plans with individual customers.27 Accounting for time of customer service representatives and clerical worker along with associated overhead, Columbia Gas estimated that in 1989 it incurred a cost of $14.64 for each individual payment plan negotiation.28 To the extent effective utility DSM programs make bills more affordable, they simultaneously reduce the need for utility companies to incur costs associated with payment plan negotiation. However, we do not have sufficient data to calculate an adder for this non-energy benefit category.

2. Reduced Regulatory Costs

Utilities incur regulatory costs in dealing with payment-related problems. Such regulatory expenses in this context include the portion of rate cases devoted to issues generated by inability to pay, rulemaking attention to payment-related issues, and regulatory attention to individual complaints.29 Effective utility DSM, customer relations and customer education programs reduce these costs. Due to insufficient data, we do not attempt to quantify this benefit at this time.

3. Reduced Rate Discount Payments

Many utility companies offer reduced or negotiated payment schedules for customers with limited ability to reduce overdue balances or make full and prompt bill payments. Increasing customers' ability to pay through enhanced end-use efficiency and education regarding available social services reduces the foregone revenues associated with reduced rate discounts. Further, as a low-income household's consumption decreases as a result of implementation of DSM measures, low-income discounts paid by all of the service territory's ratepayers also decrease. The Skumatz study includes an estimated range of $42-270 per weatherized household to account for this non-energy benefit.30 Based on reported program costs of $719 per weatherized household, an adder of 5.8 percent - 37.6 percent is appropriately applied to cost-effectiveness testing.

This may be a transfer payment from a societal point of view and a benefit from a utility/ratepayer perspective.

4. Avoidance of Decreased Sales/Maintenance of Contributions to Fixed Costs

Niagara Mohawk's Affordable Payment and Arrearage Forgiveness Program was designed with the objectives of improving both relations with "payment-troubled" customers and company profitability. The Program's goals were to increase the regularity and total amount of payments by participating customers, increase the use of available assistance through programs such as the Low-Income Home Energy Assistance Program (LIHEAP), decrease the number of collection actions for participating customers, and eliminate arrearages of participating customers. The program reduced billing shortfalls and bad debt write-offs while retaining customers paying a portion of their monthly bills.31 Evaluators noted that as an alternative to the program, the company could terminate service to customers with payment problems. However "...from an economic perspective, as long as customers can cover variable costs, it makes economic sense to serve them."32 Fixed costs are incurred whether or not a customer consumes electricity. Maintaining a customer who pay enough to cover all allocated variable costs plus makes some contribution to fixed costs contributes more to net income than does termination of service to that customer. This is similar to the rationale for economic development rates for industrial customers.

Thus, as long as negotiated customer payments combined with payment assistance exceed variable costs, there is a utility and ratepayer benefit generated from incremental contributions to fixed costs. However, we currently lack sufficient data to quantify this additional benefit, which is a transfer payment from a societal point of view but a benefit from a utility/ratepayer perspective.

B. Reduced Emergency Calls

1. Gas

Several analysts have assessed the health and safety impacts of low-income weatherization and DSM programs. Many low-income households have old and poorly maintained space and water heating systems that present health and safety risks.

For example, a 1997 evaluation of Louisville Gas and Electric Company's Energy Partners Program33 conducted by Proctor Engineering Group included a tabulation of the frequencies of various health and safety problems encountered by program personnel.34 Among participating households, 23 percent had gas leaks, 26 percent had inadequate draft for space or water heaters, nine percent had high carbon monoxide levels (100-400 ppm), seven percent had very high carbon monoxide levels (over 400 ppm), and one percent had very high carbon monoxide levels and inadequate draft on the same appliance. The report characterized the combination of very high carbon monoxide levels and poor draft as a "potentially lethal safety problem," and noted that the program provided significant health and safety benefits to participants by reducing the risk of illness or death from high levels of noxious combustion gasses, fires, and explosions resulting from gas leaks.35

In addition to mitigating costs directly associated with illness or death as noted above, effective DSM programming that includes maintenance and repair of space and water heating appliances reduces the need for gas utility companies to incur costs associated with making emergency calls to deal with potentially hazardous problems. Public Service Colorado estimated this savings to average nearly $16 per weatherized household during the first year after delivery of weatherization services.36 Skumatz estimated the value over time of utility benefits associated with fewer emergency gas calls to range between $84 and $170, resulting in an adder range of 11.6 percent to 23.6 percent.

2. Electric

While gas service calls have the most significant impact on energy system costs which can be avoided as the result of DSM programs, the electric distribution utilities have some added costs in responding to customer requests which result from poor quality or malfunctioning electric systems. Unsafe and hazardous conditions as well as disputes in payment between owner and renters often results in service calls which can be avoided as the result of a well planned and executed energy efficiency program. There has been little hard analysis on the exact costs of these services but they are certainly a component in the overall cost of service.

II. SOCIETAL BENEFITS

In addition to generating benefits associated with direct energy savings and savings to utility ratepayers and shareholders, utility low-income DSM programs generate a range of benefits to society at large. However, societal benefits are often very difficult to quantify. Some societal benefits may be considered "transfer payments" among sectors within society, thus raising questions about the application of certain quantified benefits in cost-effectiveness testing while others clearly represent costs avoided by some sector of society (either public or charitable).37

Among societal benefits are the following: (1) incremental economic development associated specifically with delivery of DSM services to low-income households; (2) maintenance of the real estate tax base and reduction of public expenditures associated with delivery of medical, firefighting and fire prevention, and social services; operation of homeless and housing programs; and unemployment payments; (3) increased equity; and (4) benefits to low-income people, including housing, reduced moving and homelessness, maintenance of utility services, improved property values, and improved health. We deal with each of these benefit categories below.

A. Incremental Economic Development

Investments in energy efficiency lower consumer energy expenditures, thereby allowing increased spending in other sectors of the economy. While specific changes in total employment generated by energy efficiency expenditures depend on the structure of a local or regional economy, research has generally demonstrated that increased non-energy expenditures produce net employment gains as well as other contributions to economic well being.38 These have been incorporated into the general economic benefit calculations made to support a general economic adder.

Energy expenditures typically represent cash outflows from a regional economy. Efficiency-based expenditure reductions are generally redirected in a manner that, particularly after accounting for multiplier effects, produce significant net employment and income gains. Further, the electric and gas utility industries, as well as the oil and gas mining industries, are among the most capital-intensive in the economy. Redirecting expenditures away from these industries and toward more labor-intensive sectors, such as retail trade or services, results in total employment and income gains.39

There is an incremental economic development benefit associated with energy efficiency investment in low-income households. Accepted macroeconomic theory holds that, as income declines there is an increasing propensity to spend and a proportionate decreasing propensity to save.40 Therefore, in the case of low-income households, all savings stemming from energy efficiency improvements are likely to be immediately redirected into the local economy. Higher income households are more likely to save a portion of the savings, thus reducing the economic "ripple effect" that re-spending would create. To our knowledge, this incremental economic development benefit has not been quantified, so we treat it in a qualitative manner for the purposes of this analysis.

B. Public Funds

Utility DSM programs reduce the need for a wide variety of public expenditures. Below, we review the positive effects that DSM programs have on public expenditures in the areas of healthcare, public safety, housing, housing values, unemployment insurance, and social services.

1. Health

The elderly poor are particularly susceptible to weather-induced health problems.41 Indeed, hypothermia and hyperthermia are examples of potentially fatal health conditions that are most common among elderly people with limited ability to pay for adequate levels of energy service.42

Energy efficiency programs targeted to low-income mitigate a variety of health effects and the costs associated with treatment. An obvious example is that weatherization combats hypothermia and the use of carbon-monoxide-producing appliances. One study estimates that the value of reduced illnesses and increased health is $1,300 per weatherized household.43 Under the reported program cost of $719 per weatherized household, an adder of up to 181 percent reflects this value.

2. Fire

Many low-income households have old and poorly maintained space and water heating systems that present safety risks to occupants.44 Gas leaks in space heating and water heating equipment pose the threat of a house fire. Further, high utility bills and service disconnections lead to use of fire-hazardous alternative heating sources, such as electric space heaters or gas grills.45

Much as effective weatherization programming reduces the need for emergency gas service calls, it too reduces public expenditures for fire fighting and prevention. The Oak Ridge National Laboratory's "Progress Report of the National Weatherization Assistance Program" concludes that the value of reduced incidence of fire attributable to weatherization activities is $3 per weatherized household.46 Therefore, an adder of less than one percent reflects this benefit.

3. Building Inspection

Low-income Weatherization Assistance and utility DSM programs include components that improve a building's heating system and envelope.47 To the extent that these programs are successful in bringing substandard buildings up to building and health codes, they reduce the need for building inspections.

4. Homeless Shelters

As noted in Section III.D.3, below, DSM programs contribute to the prevention of homelessness and housing abandonment by enhancing energy service affordability and by reducing the number of service terminations that lead to loss of residency. Reductions in homelessness have the added public funds benefit of reducing the financial strain on homelessness shelters.

5. Maintenance of Real Estate Tax Base

As noted, weatherization aids in the prevention of housing abandonment.48 In addition, home energy efficiency investments increase housing values. For example, one study found that home values increase by 20.7 times the annual reduction in fuel use.49 Because real estate tax rates are directly tied to property valuations, increased housing value caused by home energy efficiency investments are directly tied to maintenance or enlargement of the real estate tax base.

6. Housing Programs

Low-income DSM and Weatherization programs include components that reduce the need for publicly funded housing agencies to expend funds. For example, agencies that deliver low-income weatherization services often work with local housing authorities to improve building shell and heating system conditions. The absence of these services would require expenditures by the housing agency to either improve sub-standard conditions or achieve optimal energy efficiency levels.50

7. Reduced Unemployment Insurance Payments

As noted above, there is an incremental economic development benefit associated with low-income DSM programs. This benefit, which may be stated in terms of increased employment and personal income, carries an added public funds benefit of reduced unemployment insurance payments. Evaluation of the National Weatherization Assistance Program referenced a non-energy benefit of avoided cost of unemployment benefits to be associated with the program. This benefit was estimated to be $82 per weatherized household, justifying the use of an adder of 5.29 percent.

This may be considered a transfer payment from a societal point of view but a benefit from a utility/ratepayer perspective.

8. Cost Reduction from Efficiencies in Social Service Delivery

Weatherization Assistance and utility DSM programs can incorporate components geared toward referring participants to other available social services, thus reducing the costs associated with outreach and administration of these other programs. For example, many households that receive weatherization assistance also receive a referral to separate assistance programs such as the Low Income Home Energy Assistance Program. The per-client outreach costs of the referred program are thus reduced.51

C. Equity

The energy cost burden52 of a low-income household is three to four times higher than that of a median income household.53 For example, expenditures for electricity by low-income households represent, on average, 7.7 percent of their total income; the very poor, living at less than 50 percent of the federally-determined poverty level spend 23 percent. In contrast, the average residential consumer spends only 2.4 percent of income on electricity.54

Clearly, households qualifying for the federal Weatherization Assistance Program cannot reconcile monthly income with expenses for basic necessities. Paying energy and utility bills requires that other necessities must be foregone. This energy budget dilemma is faced uniquely by the poor.55

The equity benefits of energy efficiency programs within the low-income community, represented by the reduced societal disparity in proportionate household energy expenditures, are thus equal to the energy cost savings benefits of the programs.56 This is true because each energy or utility dollar a low-income household is not required to spend reduces the energy burden of that household, thereby reducing the energy burden "gap" between the poor and non-poor. The societal benefit stemming from the reduction of this gap is reflected not only by increased fairness, but also by the reduced requirement of low-income households to forego other necessities. The public, in recognition of the benefit associated with reduction of the energy burden gap, has shown strong support for taking care of the energy needs of low-income households.57 Assuming that the energy savings benefit over time of a given DSM program is at least 75 percent of total program costs, it is appropriate to apply an avoided cost adder of 75 percent to this non-energy benefit.

D. Low-Income Benefits

1. Housing Development

Residential energy efficiency programs may appropriately be viewed as housing programs. The broad concept of housing includes far more than merely the physical structure of a dwelling. Rather, housing may be understood as a disparate but inseparable bundle of goods, services and qualities, including shelter, location, investment, privacy, proximity to amenities, and accessibility to utilities and other tangible services, which together comprise a household's living arrangement. When viewed in this light, it is obvious that action of most any kind directed at any one aspect of the bundle's attributes will have repercussions on other aspects or attributes.58

Accordingly, energy efficiency programs that improve housing development economics, reduce customer payments to utilities, reduce homelessness, improve housing maintenance, maintain or improve property values, reduce housing abandonments, reduce moving costs and reduce adverse health effects must be viewed as generating non-energy benefits in addition to the energy savings benefits.

Unfortunately, while many of these benefits may lend themselves to quantification through careful analysis, little work has been in this area to date. We review below some of the existing literature related to housing and community development aspects of weatherization and utility DSM programs, and provide quantitative estimates of such benefits where feasible.

The economics of low-income housing development are usually characterized by a precarious balance between rigid development costs, limited prospective income streams, and a developer's ability to obtain attractive financing and subsidies. DSM programming targeted toward new residential construction can enhance these economics in a number of ways. First, DSM programming can have a mitigating effect on the marginal cost of meeting or exceeding the energy requirements of the Massachusetts Building Code. Overall project economics would then be improved by the resulting reduction in this development cost. In addition, DSM measures that improve the efficiency of energy usage in the new facility reduce the level of future energy expenditures and therefore the operating cost associated with the facility. Reduced strain on operating budgets of residents and/or property managers may induce some financial institutions to offer relatively attractive financing terms.59 These benefits will tend to be project-specific.

2. Reduced Mobility

An analysis conducted by the Upjohn Institute of the determinants of the decision of low-income renters to move out of their dwellings reveals that low-income renters are willing to pay sizable portions of their annual incomes to not move.60 The study further reveals that there is a very high psychological and financial cost of mobility among low-income renters, particularly those who are elderly or whose households include children. The study found average moving costs for "typical" low-income households to be between ten percent and 20 percent of annual income.61

One important implication of the results of the Upjohn Institute Study is that, given the high cost of mobility to low-income households, there is particular value to policies and programs that reduce the need of these households to move. Low-income DSM and weatherization programs reduce mobility in at least three ways. First, energy efficiency improvements reduce the level of energy/utility expenditure required to attain a minimal living standard thus freeing up funds to pay rent or other required housing costs. Second, weatherization improvements ameliorate dangerous or substandard conditions in heating equipment or building shell that might otherwise force a household to relocate.62 Finally, some utility programs include arrearage forgiveness or payment plan components that result in fewer service terminations and, therefore, reduced mobility.63

Researchers estimate the value of reduced mobility among program participants was as much as $840 per weatherized household.64 Under the study's reported program cost of $719 per weatherized household, an adder of up to 117 percent is justified as a quantification of this non-energy benefit to low-income participants.

3. Reduced Homelessness, Housing Loss, Housing Abandonment

Studies have demonstrated the clear link between homelessness and utility terminations. As indicated above, energy efficiency programming and customer relations help to ameliorate late payment problems.65 It follows that terminations may also be reduced as a result of such programming.

According to surveys conducted by the Energy Coordinating Agency of Philadelphia and Institute for Public Policy Studies of Temple University, there was an average of over 60,000 gas electric and water service terminations each year in the city during the years of 1984 through 1989. The study further found that, of homes where utility service was terminated, 32 percent of electric and 24 percent of gas cases led to abandonment within one year of the utility termination. Through a name match between Philadelphia Electric Company's list of termination notices and lists of homeless adults served by the City of Philadelphia, the study found a discernable relationship between utility termination and homelessness. In surveys of individuals living in emergency shelters, 7.9 percent of respondents cited utility terminations as the reason for their homelessness. (Higher percentages cited related causes, such as "eviction for non-payment" and lack of housing in the income range as the causal factors.) The study noted that of the many factors contributing to homelessness, mitigation of high energy costs is among those "most susceptible to remedy."66 Similarly, a study of homelessness in Northern Kentucky indicates that utility shutoffs were among the primary causes of homelessness in that region.67

The studies cited above do not provide the information necessary to quantify or project the benefit of homelessness prevention associated with energy efficiency programming. Similarly, they do not project the costs associated with such programming.

4. Reduced Loss of Service Due to Termination

By enhancing energy affordability and arranging payment plans with customers in arrears, DSM programs reduce the number of service terminations suffered by customers. Researchers estimate that the value of fewer service terminations to customers is as high as $425 per participating household.68 Under the reported program cost of $719 per weatherized household, an adder of up to 59.1 percent would be applied to reflect this value.

5. Improved Maintenance, Maintenance of Property Values

Literature review demonstrates a link between residential housing values, energy costs and energy efficiency characteristics.69 In a review of seven studies conducted between 1970 and 1985, a study published in The Appraisal Journal found that (1) value of energy efficient homes with low structural heat loss was $3,248 higher than comparable inefficient homes, and (2) home value increased by about $20.73 for every $1 decrease in annual fuel bills.70 In addition, many DSM measures contribute to the maintenance of property by improving thermal integrity and heating system efficiency.

Evaluation of the National Weatherization Assistance Program computed a non-energy benefit of increased property value associated with the program. This benefit was estimated to be $126 per weatherized household,71 justifying an adder of 8.13 percent.

6. Health

Older people living in poverty are more likely than their non-poor counterparts to experience rapidly declining health and to develop difficulties performing routine daily activities as they age. Thus, low-income individuals are at a much higher risk of requiring nursing home care as they age.72 Further, among those most likely to develop hypothermia are the poor who can not afford to pay for adequate home heating.73 In addition, low-income households are at increased risk of fire and exposure to hazardous fumes due to use of unsafe heating sources because of utility terminations.74 Finally, high energy burdens cause low-income households to forego expenditures on preventive health measures and nutritional food items.75 DSM programs thus improve participants' health by preventing such dangers as hypothermia, carbon monoxide poisoning, and fires. We have not quantified this benefit.

III. CONCLUSION

Based on the foregoing presentation of benefits to society, individuals, utilities, and ratepayers from delivery of comprehensive low-income energy efficiency programs, a benefit adder of between 17 percent and more than 300 percent could reasonably be incorporated to represent the incremental value of a low-income focus beyond the general societal, economic, and environmental benefits of efficiency programs.

Furthermore, this omits a benefits that have not been quantified in the literature but that can be reasonably quantified as follows:76

• Reductions in electricity emergency calls can be extrapolated from experience with gas. Assuming the rate of electricity emergency calls is 20 percent that of gas, a reasonable estimate of the avoidable cost is 2 percent - 4.5 percent of avoided costs.
• The incremental economic multiplier effect from low-income energy efficiency programs due to increased level of spending can be reasonably estimated at 1 percent-2 percent of avoided cost.
• A reasonable lower bound for increased property values can be calculated as a $4000 improvement in value of an $80,000 home, or 4 percent - 5 percent of avoided cost.
• The value of increasing the ability of the housing market to sustain affordable housing for low-income families is difficult to precisely measure. However, it can be estimated by placing a reasonable value of 50 percent of the value established for maintenance and improvement in housing values. This would translate to an adder of 2 percent - 4 percent of avoided cost.
• The value of efficiencies in social service delivery can be estimated by assuming each household provided with comprehensive energy efficiency services also receives one referral at a (conservative) average cost savings of $20. This would translate to an adder of 3 percent.

At the lower bound, these estimates add 12 percent of avoided cost to the computation of a reasonable adder, without counting other nonquantified cost savings. When added to the lower estimates appearing in the literature and described in this paper, the total approaches the consensus value of a 50 percent adder. Thus 50 percent represents a value that is close to the bottom of a reasonable range of benefits from low-income energy efficiency programs.

As the table that follows displays, an adder of more than triple avoided costs can be justified for all low-income DSM programs - more for technology-specific measures such as gas-related, space-heating measures, and new construction.77 There is a wide range in estimates and in the precision with which they are computed; for many factors, no quantification has been done to date. Nevertheless, the lowest estimate of an adder to avoided cost is 17 percent without counting unquantified benefits such as avoided utility collection costs, avoided administrative costs, avoided regulatory costs, incremental economic development benefits, public fund savings (e.g., Fire and Building Departments), and reduced homelessness.

The mid-point of the range of estimates - still not counting the unquantified benefits - is 172 percent of avoided costs. Even taking only a third of the range (going one-third of the distance from low to high) yields an adder of 103 percent for all low-income programs.

Indeed, removing the estimates for equity and all the items that may be considered to be transfer payments from the societal perspective - still not counting the unquantified benefits - and taking only a third of the range instead of the midpoint - yields an adder of 59 percent.

In the spirit of compromise, at least as a starting point for later review as benefit quantification and future low-income DSM program measures are refined, we propose a cost-effectiven4ess adder of 50 percent of avoided cost for all low-income DSM programs with additional adders as appropriate that are technology-specific, program-specific, or site-specific. We submit that anyone taking the table that follows as a menu, selecting only the benefits and values in which they believe, will be able to justify a 50 percent adder. Analysts have thus reached their justification of 50 percent in different ways but are able to reach a consensus that 50 percent of avoided cost is a reasonable and appropriate adder for low-income DSM programs.

Table 1 Low-Income Benefits as a Percentage of Avoided Cost

__________________________________________________

1 A Non-Energy Benefit Avoided Cost Adder reflects the ratio of the estimated present value in dollar terms of the benefit to total program costs (which equal avoided energy benefits where the benefit cost ratio is 1.0). This ratio is then added to a utility's avoided cost during a cost-effectiveness test of a specific DSM measure or program. This method of calculating and applying an adder is appropriate and justifiable because it allows for a consistent accounting of non-energy benefits irrespective of a utility's particular avoided cost.

The adder is intended to provide a framework to be used in the development and evaluation of utility DSM programs, and for the accounting of benefits beyond those directly related to energy savings. The adder is not intended to provide a precise dollar value of non-energy benefits. Our calculation and recommended application of the adder is based on the following assumptions: (1) the energy benefit to total cost ratio of the programs reviewed in the development of the adders specified in this paper is 1.0; and (2) for a given DSM program, there is a positive correlation between the level of energy benefits and the level of the non-energy benefits identified in this paper that are generated by the DSM program.

2 Money that is owed utilities from previous consumption and one or more late payments. Biewald, et.al., "Non-Price Factors of Boston Edison's Demand-Side Management Programs: A Review of the Societal Benefits of Energy Efficiency," (1995), p. 14-1.

3 Debt that is incurred if collection efforts are unsuccessful. Id.

4 Expenses associated with credit and collection of unpaid bills, including issuance of shutoff notices, personal contact with customers, disconnection of service, reconnection of service, and payments to collection agencies. Id.

5 Costs associated with disconnecting and reconnecting service in the event of non-payment.

6 Utility personnel costs of negotiating payment plans with customers. Id.

7 Ratecases and other regulatory proceedings that are, in part or in total, devoted to issues related to treatment of customers with outstanding bills or customer complaints against utility companies. Id.

8 The variability of study results is due to both program distinctions and the studies' methodological distinctions.

9 R. Grosse, "Win-Win Alternatives tot Credit & Collections" (Wisconsin Public Service Co. 1997). Information regarding the utility expenditures required to generate these reductions, as well as the dollar value of the reductions, was not provided.

10 Id.

11 J. Oppenheim, "The Utilities," Access to Utility Service, National Consumer Law Center, 1998 Supplement, pp. 30-31 from U.S. Department of Energy/Energy Information Administration, "Electric Sales and Revenue, 1996," Table 14 (1997); U.S. Census, March 1998; "Current Population Survey," Table H-8; U.S. Census, 1990 summary tape, file 3A, Tables H3, P3, P80, P121. See § III.C, Equity Benefits, below.

12 Quaid, M., and Pigg, S., "Measuring the Effects of Low-Income Energy Services on Utility Customer Payment Behavior," Proceedings of the 1991 Fifth International Energy Program Evaluation Conference, 1991.

13 See § II.A, above.

14 On the other hand, a review of Niagara Mohawk's low-income DSM program showed that customers with the least ability to pay bills paid the smallest dollar amount against arrearages after DSM. Customers with ability to pay reduced their arrearages by an average of $469 after delivery of the DSM program, no doubt at least partially because they had larger arrearages since level of usage is correlated with income. No program costs were reported. Alliance to Save Energy, "Evaluating the Benefits of Comprehensive Energy Management for Low-Income, Payment-troubled Customers," 1992.

15 Biewald, et al., at 14-2 - 14-5. The authors issue numerous caveats regarding the comparison of results from different studies. For example, they cite differences in the measures installed and information provided through different programs, other administrative and programmatic distinctions, and variations of benefit measurement methodologies.

16 Linda G. Berry, et al., "Progress Report of the National Weatherization Assistance Program," at 38, 45 (Oak Ridge National Laboratory, 1997).

17 Lisa A. Skumatz, Chris Ann Dickerson, "Extra! Extra! Non-Energy Benefits Swamp Load Impacts for PG&E Program!" 1998 Summer Study on Energy Efficiency in Buildings Proceeding, pp. 8.301-8.307 (American Council for and Energy Efficient Economy, 1998). In the study, present values were calculated based on a ten year lifetime, discounted at four percent annually. Items in bold are summarized in the attached table.

18 Payment assistance programs yield similar types of benefits. For example, the Clark County (Washington) Public Utility District capped low-income families' gas service payments at nine percent of income starting in 1988, at a cost to it of $450,000 per year. The program also provided positive assistance to customers in upgrading bill management and payment skills. Over two thousand customers participated in the program, which yielded the following results: (1) Write-offs dropped 36 percent, saving $300,000 per year; (2) disconnections dropped 65 percent; (3) delinquent balances fell from 67 percent of the eligible population to 13 percent. Collection costs saved were about $100,000 per year at the rate of $108 per customer in arrears per year. Per customer payments increased from $29 per month to $52, as customers felt more hopeful about handling the more affordable payment. Thus the direct annual utility benefits of about $500,000 exceeded total program costs by 11 percent, generating a one-year benefit cost ratio of 1.11:1. Weiss, "Low-Income Assistance Pays for Itself," Northwest Energy Coalition (1998).

19 J.K. Magouirk, "Evaluation of Non-energy benefits from the Energy $avings Partners Program," 1995 Energy Program Evaluation Conference, Chicago, pp. 155-175 (1995).

20 The Colorado study reported savings only for the first year after weatherization. We calculated the present value of savings over measure lifetimes based on a ten-year lifetime, discounted at 4 percent annually.

21 Skumatz at 8.307.

22 Colton, "Identifying Savings Arising from Low-Income Programs," National Consumer Law Center, 1994, p. 16.

23 Id. at 3.

24 Id.

25 Skumatz at 8.307. The variance in costs reported by Colton and Skumatz is due largely to the fact that Colton reports the utility cost of each instance of the particular collection activity, while Skumatz reports an average over all participating households.

26 This is reported in the attached table under "Reconnection and Termination" since the other elements are not otherwise quantified.

27 Colton at 7.

28 Id.

29 Id. at 6-7.

30 Skumatz at 8.307.

31 Response Analysis Corporation, "Niagara Mohawk Power Corporation Affordable Payment and Arrearage Forgiveness Program," p. 1-3 (1992).

32 Id. at p. 3-9.

33 Louisville Gas and Electric Company's Energy Partners Program is designed to reduce the energy consumption of low-income and "payment-troubled" customers. Program goals included saving 15%-20% of participants' energy usage; reducing bills and therefore disconnections, arrearage levels and collection actions; and improving the health and safety of participants. Blasnik, "Impact Evaluation of Louisville Gas and Electric Company's Energy Partners Program: Final Report," (1997), p. 1.

34 Id. at 39.

35 Id. at 40.

36 Skumatz at p. 8.305.

37 Benefits that may be considered transfer payments are noted in the attached table with an asterisk.

38 See, e.g., Skip Laitner, et al., "Energy Efficiency as an Investment in Ohio's Future," p. 30 (American Council for an Energy Efficient Economy, 1994).

39 Id. at 31.

40 See, e.g., Paul A. Samuelson, Economics, pp. 210 - 215 (McGraw-Hill, 1976).

41 See § III.D.7, below.

42 Spade, et al., "The Energy Affordability Crisis of Older Americans: An Examination of the Hazards to Health and Well-being Posed by the Growing Incidence of Unmet Home Energy Needs," p. 28 (National Consumer Law Center, 1995).

43 Skumatz at 8.307.

44 See § II.B, above.

45 Spade at 36.

46 Berry, et al., at 38, 39. The report does not distinguish between public and private costs of fire avoided by the Weatherization Assistance Program.

47 Id. at 7.

48 See § III.D.3, below.

49 Nevin, et al., "Evidence of Rational Market Valuations for Home Energy Efficiency," The Appraisal Journal, p. 403 (Appraisal Institute, 1998).

50 Interview with Art Wilcox, South Middlesex Opportunity Council (1999).

51 Id.

52 The energy burden refers to percentage of household income devoted to home energy costs.

53 Tannenbaum, et. al. "Low-Income Energy Services in a Competitive Environment," Energy Center of Wisconsin. 1998. Also, Argonne National Laboratory, "Residential Energy Consumption Survey" reported in Rabago, et. al. "An Alternative Framework for Low-Income Electric Ratepayer Service." 1992. p. 2.

54 Computed by J. Oppenheim, "The Utilities," Access to Utility Service, National Consumer Law Center, 1998 Supplement, pp. 30-31. from U.S. Department of Energy/Energy Information Administration, "Electric Sales and Revenue, 1996," Table 14 (1997); U.S. Census, March 1998; "Current Population Survey," Table H-8; U.S. Census, 1990 summary tape, file 3A, Tables H3, P3, P80, P121.

55 The concept of "Shelter Poverty" was developed initially by Michael E. Stone in the mid-1970s and more recently in Shelter Poverty: New Ideas on Housing Affordability, (Temple University Press, 1993) Shelter Poverty is a framework used to demonstrate that non-shelter necessities must compete for left-over dollars after shelter (housing and utility) costs are taken "off the table" to avoid homelessness.

56 In a 1996 national study, the U.S. Department of Energy found that weatherization saved 33.5 percent of the gas space heating cost of an average low-income household. Berry, et al., p. 22.

57 There is overwhelming public support for programs to ensure that all households have their basic energy needs met. For example, a national survey found that 89 percent of those with an opinion favor federal low-income energy payment assistance and 79 percent of those with an opinion favor an increase in such funding. Behavior Research Center, "Public Opinion National Survey on Low-Income Home Energy Assistance Program," p. 2 (1998). In a 1997 survey conducted by El Paso Electric Company, respondents in aggregate rated the factor of meeting everyone's basic energy needs as highly important. This factor received an aggregate rating of 8.9 on a scale of 0 (not at all important) to 10 (extremely important). Guild, et al., "Southwest Town Meeting on Electricity Issues" (El Paso Electric Company, 1997). In addition, results of a 1987 residential survey of Connecticut residents demonstrate strong public support for energy cost assistance to low-income and elderly persons. Further, the study identified strong public support for the notion that access to energy for residential use is a right in our society. John M. Kennedy, "Public Support for Residential Energy Assistance," 71 Sociology and Social Research 308 (1987).

58 Montgomery et al., Housing in America: Problems and Perspectives, 1979. pp. 3-6.

59 This is a similar concept to that which was behind the development of the "Home Energy Rating System" that has been implemented in a number of different states (e.g., Vermont). Simply put, participating financial institutions offer reduced mortgage rates for purchase of dwellings that exceed certain energy efficiency standards.

60 Bartik, et al., "Maximum Score Estimates of the Determinants of Mobility: Implications for the Value of Residential Attachment and Neighborhood Amenities," Upjohn Institute Staff Working Paper 90-01, pp. 1, 10-11(1990).

61 Id. at 10-11. A "typical" low-income household, based on overall means of the sample population, consisted of a non-minority household, with no spouse present, two children, and a head age 44, which had been at its current residence for 48 months.

62 See § II.B and III.B.2, above.

63 See § II.A, above and III.D.3, below.

64 Skumatz at 8.307.

65 See § II.A, above.

66 Liz Robinson, "An Examination of the Relationship between Utility Terminations, Housing Abandonments and Homelessness," pp. 1, 2 (Energy Coordinating Agency of Philadelphia, 1991).

67 William K. Woods, et al., "Homelessness and Low-Cost Housing in Northern Kentucky," p. 2 (Northern Kentucky Coalition for the Homeless and Applied Information Resources, 1990).

68 Skumatz, p. 8.307.

69 Nevin, et al., at 403.

70 Id.

71 Berry, et al., at 38.

72 Interview with Raymond Coward, Dean of the School of Health and Human Services, University of New Hampshire from "USA Today Magazine," April 1998, v 126 n2635 p. 5.

73 Bonnie Guiton, "Special Report on Cold Stress and Heat Stress," p. 1 (U.S. Office of Consumer Affairs).

74 Colton, 1993.

75 Cambridge Systematics, Inc., "Hard to Quantify Benefits and costs Scoping Study," prepared for the New York Low-Income Evaluation Task Force. 1994

76 Interview with Stephen Cowell, Conservation Services Group. Documentation to be provided.

77 Sources for the table are indicated in the text and footnotes of the paper.