• DPI Home
• |
• Site Map

Advanced Metering Infrastructure Customer Impacts Study - Stage 2

Final report: 11 July 2012

Contents

Executive Summary

• 1 Introduction
  • 1.1 Context
  • 1.2 Background – AMI customer impacts study: Stage 1
  • 1.3 Limitations of the analysis
  • 1.4 Terms of reference – Stage 2
  • 1.5 Report structure
• 2 Objective and methodology
  • 2.1 Overview of approach
  • 2.2 Development of new tariff scenarios
  • 2.3 Analysis of billing data
  • 2.4 Analysis of ABS HES data
  • 2.5 Quantitative research of sensitivity to ToU pricing
  • 2.6 Qualitative market research and case study development
• 3 Impact of ToU scenarios
  • 3.1 Development of new tariff scenarios
  • 3.2 Distribution of Impacts of Flexible Pricing Scenarios
  • 3.3 Quarterly impact of ToU Pricing
• 4 Validation of vulnerable groups using HES data
  • 4.1 Prevalence of Demographic Groups
  • 4.2 Confirmation of vulnerable groups
  • 4.3 Distribution of vulnerability
  • 4.4 Further analysis of vulnerability
• 5 Focus group sessions
  • 5.1 Overview of approach
  • 5.2 Focus groups conducted
  • 5.3 Discussion guide
  • 5.4 Key insights from discussions
  • 5.5 Conclusion – focus group sessions
• 6 Price sensitivity analysis
  • 6.1 Overview of respondents
  • 6.2 Results
  • 6.3 Summary
• 7 Case studies
  • 7.1 Overview of approach
  • 7.2 Case studies developed
  • 7.3 Conclusion – case studies
• 8 Conclusions and recommendations
  • 8.1 Summary of Stage 2 analysis results
  • 8.2 Verification of Stage 1 findings
  • 8.3 Recommendations
• 9 Limitation of our work
• General use restriction

Appendix A

Appendix B

Appendix C

List of figures

• Figure 1: Distribution of residential winter vulnerability
• Figure 2: Selecting CCDs for inclusion in analysis of household type – overview
• Figure 3: Example graphical representation of analysis of Van Westendorp study*
• Figure 4: Average annual price change for customers in each group under the six ToU scenarios
• Figure 5: Annual change in electricity bill for single element customers, with elasticity effect
• Figure 6: Annual change in electricity bill for dual element meter customers, with elasticity effect
• Figure 7: Comparison of change in electricity bill across quarterly bills for Scenario F
• Figure 8: Scatterplots of income versus expenditure for the "Non-vulnerable" group
• Figure 9: Scatterplots of income versus expenditure for each demographic group
• Figure 10: Distribution of vulnerability compared between each demographic group and the “Non-vulnerable” group
• Figure 11: Percentage of each group using electricity for heating
• Figure 12: Percentage of each group using electricity for electric-boosted solar hot water
• Figure 13: Van Westendorp sensitivity scores to bill increases for each group
• Figure 14: Average rating for ‘ease of shifting heating and cooling use’ by group
• Figure 15: Average rating for ‘ease of shifting other use’ by group
• Figure 16: Average rating for ‘reducing electricity use during peak times’ by group
• Figure 17: Example Van Westendorp pricing sensitivity meter diagram with labelling of the two portions of lines used to calculate sensitivity
• Figure 18: Van Westendorp pricing sensitivity meter diagrams for each group of respondents that participated in a quantitative telephone survey on Flexible pricing

List of tables

• Table 1: Glossary table for commonly used terms with their abbreviations
• Table 2: Tariff scenarios modelled in Stage 2*
• Table 3: Criteria for identifying ABS HES households by demographic group
• Table 4: Stage 1 criterion for defining vulnerability by demographic group
• Table 5: Focus groups conducted for Stage 2 qualitative market research.
• Table 6: Representative tariffs – March 2012 (including GST)
• Table 7: Summary: Tariff scenarios modelled in Stage 2 (cents, including GST)*
• Table 8: Tariff 2A – three part tariff with narrow peak times
• Table 9: Tariff 2B – three part tariff with narrow peak times and CPP
• Table 10: Tariff 2C – three part tariff with wide peak times
• Table 11: Tariff 2D – three part tariff with wide peak times and lower peak price, relative to off peak and shoulder
• Table 12: Tariff 2E – three part tariff with wide peak times, weekends at off peak rates
• Table 13: Tariff 2F – three part tariff with wide peak times
• Table 14: Estimated proportion of single element meter groups better off under each ToU Scenario (with elasticity effect)*
  
• Table 15: Estimated proportion of dual element meter groups better off under each ToU Scenario (with elasticity effect)*
• Table 16: Percentage of Victorian HES data records by demographic group
• Table 17: Percentage of vulnerable Victorian ABS HES households in each demographic
• Table 18: Focus group sessions conducted
• Table 19: Summary of attitudes by vulnerable group
• Table 20: Number of telephone survey respondents by group
• Table 21: List of case studies
• Table 22: Stage 1 key findings and comments
• Table 23: Distribution of change of electricity bill and % customers better off under each scenario and against season, vulnerability group and element type
• Table 24: Average change of quarterly electricity bill under ToU Scenario F
  

Glossary

Table 1: Glossary table for commonly used terms with their abbreviations

Term	Meaning
advanced metering infrastructure (AMI)	A specific form of smart metering capable of (among other things): recording of electrical energy imported or exported from a metering point by half hour trading interval; remote disconnection and reconnection; load control; and remote detection of loss of supply1
all residential	Signifies all residential customers/households, regardless of that customer/household being classified as elderly, low income, etc.
basic community profile (BCP)	The basic community profile is a file made available by the Australian Bureau of Statistics which provides detailed census data for small areas (the smallest of which are CCDs). They contain information that includes age, income, education and housing costs
block tariff	Pricing structures that include either an incline or decline in price per kWh beyond a given threshold of electricity use. Pricing structures which increase past a threshold of electricity use are termed ‘inclining block tariffs’ where those which decrease are termed ‘declining block tariffs’
census collection district (CCD)	A geographic identifier comprising approximately 200 to 250 households developed by the Australian Bureau of Statistics (based on the 2006 census classification system) that allows relatively detailed, yet anonymous, analysis of census-type information
controlled load	Customer appliances that are connected to a dedicated circuit and are subject to separate metering of energy use – usually water and space heating. The relevant controlled load would only be energised “ON” at restricted times of day.
critical peak incentive	An arrangement where customers are eligible for a rebate on their energy bill in return for reducing their consumption from an agreed baseline level during a critical peak period
critical peak pricing (CPP)	Under critical peak pricing, electricity prices are increased sharply for a limited duration at times when demand needs to be reduced
demand	A measure of the amount of electricity either generated or used at any given instant of time. The units of measure include kilowatts, megawatts and gigawatts.
dual element meter	A meter that separately records electricity usage at one location for 1) controlled loads; and 2) other circuits.
Department of Primary Industries (DPI)	Victorian government department that has engaged Deloitte to deliver this report on AMI Customer Impact
electricity distribution business (DB)	Electricity distribution businesses manage the distribution of electricity through the ‘poles and wires’ network. Electricity distribution businesses have been charged with implementing AMI
electricity retail business (RB)	The companies that manage consumer’s electricity accounts and who consumers’ ultimately purchase their electricity from. Retailers purchase electricity in bulk from the wholesale market and ‘rent’ distribution infrastructure to deliver this electricity to consumers
energy	A measure of the volume of electricity used or generated over a period of time. The units of measure include kilowatt hours, megawatt hours and gigawatt hours.
flat rate tariff	Electricity tariffs with a fixed price for electricity use that neither varies according to the time of use nor according the quantity used
Flexible Pricing	A term used to describe the new tariff arrangements facilitated by the Victorian rollout of smart metering, including time-of-use tariffs, critical peak pricing and incentives and direct load control.
full retail competition	A market within which all consumers, including the smallest of energy users, are able to choose their retailer
geocode	Assigning some geographic identifier (other than street address) to a data record
Household Expenditure Survey (HES)	A survey conducted by the Australian Bureau of Statistics, recording income and expenditure of Australian households
gigawatt (GW) gigawatt hour (GWh) kilowatt (kW) kilowatt hour (kWh) megawatt (MW) megawatt hour (MWh)	Electricity-specific units of measure
multivariate adaptive regression splines	A form of non-parametric regression technique that allows for the modelling of non-linearities and interactions in a regression
national metering identifier	A unique identifier of each registered metering point
non-vulnerable	Signifies residential customers/households that are not classified in a vulnerable demographic such as elderly, low income, etc.
price sensitivity	A measure of how sensitive a customer is to an increase in their bill from different bill levels, for example, whether they consider a particular bill to be too expensive (and therefore unaffordable)
single element meter	A meter that records all electricity usage at one location without distinguishing between types of load
small and medium enterprise (SME)	Defined by the Australian Bureau of Statistics as a company with headcount less than 200 employees
smart metering	A generic term to identify a meter with remote communication ability that can undertake more functions that merely interval metering – see also “advanced metering infrastructure (AMI)”
Time-of-use tariff (ToU)	Time-of-use pricing structure in which the price paid for electricity is dependent on the time at which it is used
vulnerable	In the context of this study, ‘vulnerable’ means ‘vulnerable to bill increases’. Our identification and categorisation of ‘vulnerable customers’ as being within particular groups is explained in this report.

Executive Summary

The Victorian Government has mandated the rollout of advanced metering infrastructure (AMI) to measure consumers’ time-of-day electricity use as well as provide other functionalities. One of the benefits of AMI is its ability to help deliver price signals that reflect the dynamic nature of the cost of electricity supply. Such ‘time-of-use’ (ToU) pricing encourages customers to better match the value they place on using electricity with the cost of electricity at any given time. AMI also enables valuable efficiencies in network supply, customer conservation of energy and other functionalities.

Deloitte has been appointed by the Department of Primary Industries (DPI) to determine the potential impacts on domestic and small business consumers of new pricing arrangements enabled by AMI in Victoria, with a focus on those vulnerable customers² who may be affected by changing electricity tariffs.

In Stage 1 of our study, Deloitte collected data from electricity distributors and retailers to build a state wide picture of electricity use. We combined this data with a range of demographic factors in order to analyse and identify which factors explain consumer vulnerability to changes in pricing structures. The result was a simulator which can create specific pricing designs in order to investigate the impact on particular groups.

The Stage 1 analysis was impacted by a number of factors. Firstly, data was volunteered by each of the participating organisations which varied in both definition and quality. For example, there was only a fairly small quantity of AMI meter data available so questions arose as to the applicability of identified time-of-use patterns. Deloitte engineered a number of data cleansing processes to minimise the impact but some uncertainty still existed regarding the findings of Stage 1.

Stage 2 of this study addresses these issues through a triangulated methodological approach that takes advantage of more recently acquired data, re-analysis of Stage 1 data, applicable market research techniques and the development of case studies into the impact of time-of-use pricing. The quantitative and qualitative results of Stage 2 together serve to verify and build upon our Stage 1 findings, giving confidence to our analysis of the range of potential impacts on vulnerable customers and their attitudes to, and therefore potential responses to, Flexible Pricing.

Stage 1 reported the average relative price change under different scenarios for particular vulnerable customer groups as well as for Victorian electricity customers more generally. In Stage 2 we have re-analysed the data to report the likely distribution of effect within each identified vulnerable group, using updated tariff scenarios that reflect further consultation between DPI and stakeholders, as well as the passage of time, since the Stage 1 tariffs were developed. Tariff scenarios modelled in Stage 2 are presented in the following table.

Table 2: Tariff scenarios modelled in Stage 2*

Tariff scenario
2A – Three part tariff with narrow peak times
2B – Three part tariff with narrow peak times and CPP
2C – Three part tariff with wide peak times
2D – Three part tariff with wide peak times and lower peak price, relative to off peak and shoulder
2E – Three part tariff with wide peak times, weekends at off peak rate
2F – Three part tariff with wide peak times which starts and ends one hour later than Scenario 2C

*Note: Both the level of the tariff (c/kWh) and the time in which peak/off peak/shoulder tariffs apply will affect the overall revenue recovered by retailers and also affect customer bills.

We have found that although the average customers in a vulnerable group will only be slightly better or worse off under different time-of-use scenarios (which confirms our Stage 1 findings), there is significant variation around the average. Some individual customers may therefore experience large annual bill changes. For example, 50% of residential customers could individually experience either a reduction in their annual bills of greater than 24.3%, or an increase in their annual bills of more than 12.0%. Our analysis has primarily been carried out on the basis of annualised bills, however, we have undertaken some quarterly bill analysis which shows that summer bills are, on average, likely to be the most negatively affected by Flexible Pricing (in that the bill change is either a greater increase or smaller decrease in summer as compared to other quarters). The adoption of monthly billing (which can now be carried out more efficiently using smart meters) should reduce cash flow difficulties associated with large seasonal bills.

Since the completion of Stage 1, the Australian Bureau of Statistics has released a new Household Expenditure Survey (HES). We accessed this survey data to cross-validate findings in relation to vulnerability reported in Stage 1. We were able to closely match similar groups of HES households with the vulnerable groups identified in Stage 1. We also confirmed our Stage 1 findings that HES households in these vulnerable groups are amongst those Victorians spending the greatest proportion of their income on electricity. Further statistical modelling and correlation analysis of the HES data did not highlight any additional groups of vulnerable Victorian households that needed to be considered alongside the existing vulnerable groups being analysed. That is, the factors that indicate vulnerability in the HES data are accounted for in the existing categories of vulnerable customers in this study.

Stage 2 included a quantitative survey of consumers to determine their sensitivity to price change and compare the affordability each vulnerable group of consumers indicates against households that are not in any vulnerable group. All vulnerable groups expressed much higher sensitivity to bill increases than residential households that were not in a vulnerable group, except Health Care Card holders that did not differ significantly in their response as ‘Non-vulnerable’ households. The highest levels of sensitivity to price increase were particularly noted amongst Regional and Single Parent households. All participants of the survey indicated it would be difficult to shift their heating and cooling from peak times. However, in comparison, all participants generally felt that they are slightly more likely to reduce consumption rather than shift it to non-peak times, but indicated that any change would be difficult.

Six focus group sessions were held with representative consumers of each vulnerable group, gathering their views on the AMI program, on Flexible Pricing and energy affordability issues. The focus groups enabled an opportunity to observe attitudes to electricity pricing and use and test out the potential behavioural responses to Flexible Pricing in a qualitative manner. All participants of the focus group sessions expressed the view that more information was needed on household energy use, with many expressing their current frustration in understanding the connection between their own appliance use and their electricity bills, as well as a lack of basic knowledge of what is driving their bills. Some participants expressed a distrust of energy companies and a belief that government should assist households by providing more information on what drives their energy use, particularly to help them deal with the new pricing structures.

Once the various concepts, reasons and issues were explained to each focus group, participants seemed to view Flexible Pricing as ‘fair’ but in some cases ‘difficult’.

Questions relating to the elements of Flexible Pricing other than three rate ToU tariffs (such as Critical Peak Pricing) were also asked in the focus group sessions. In general, Critical Peak Incentives or Rebates were viewed more favourably than Critical Peak Pricing, while Direct Load Control of air conditioning was viewed as problematic and unpopular. In Home Displays were considered useful, with around half the participants indicating that they would be willing to pay a one off fee of between $50 and $100 to receive such a device.

Finally in Stage 2 the four strands of research have been drawn together to produce a series of case studies that draw on real life experiences to illustrate the range of consumer lifestyle and economic impacts of time-of-use pricing, attitudes towards shifting demand and capacity to change behaviour. The case studies present the range of views expressed during focus group sessions and the other research components of Stage 2.

The analysis and results outlined in this report serve to highlight and reinforce a consistent underlying theme of Stages 1 and 2 of the Customer Impact Study, being that the effects of Flexible Pricing on vulnerable customers are likely to be highly variable both among and within different groups.

However, our analysis suggests that, on average, if vulnerable people elect to take up Flexible Pricing, they will be better off than they are currently (in that their total electricity bills over a year will be lower). Following this, if they also respond to the price incentives created by shifting or lowering their peak consumption, our analysis concludes that on average they will face even lower electricity bills over the year.³ This result suggests that it is not necessarily the case that all customers who identify as vulnerable will find Flexible Pricing results in higher electricity bills; indeed, the majority of customers are likely to be better off or to face very little changes to their electricity bills. This is not to say that some vulnerable customers won’t face increases in bills. While on average there will be reductions, a proportion of customers will face increases if they elect to move to Flexible Pricing.

Based on our findings, we have developed three core recommendations for government in relation to the introduction of Flexible Pricing:

• The introduction of ToU pricing, including CPP and other incentives, should be on a voluntary or ‘opt in’ basis; this is consistent with the government’s position on ToU pricing. Our analysis shows that Flexible Pricing will have a wide range of impacts on vulnerable customers’ electricity bills. Maximum contract periods or ‘cooling off’ periods for Flexible Pricing could minimise the risk of significant hardship if an increase in bills does occur.
• The government should provide information resources for consumers so they can determine if they are likely to be better off under Flexible Pricing, as well as presenting ways to save electricity. In our view, the provision of consistent, clear information to assist customers to understand smart meters and Flexible Pricing is critical to its successful introduction.
• If Flexible Pricing is utilised, it can be done under the broad electricity concessions framework currently operating in Victoria, provided the ToU tariffs are offered on a voluntary or ‘opt in’ basis, and that consistent information is provided to vulnerable customers on how Flexible Pricing would affect them⁴.

1 Introduction

1.1 Context

In February 2006, the Victorian Government mandated the accelerated rollout of advanced metering infrastructure (AMI) to 2.6 million households and small businesses. As at May 2012, more than half of these new meters have been installed by the five Victorian electricity distribution businesses (DBs).

AMI includes functionality that will improve the efficiency of electricity network operations and present opportunities for customers to better understand and engage in their electricity use. AMI has the capability to help deliver price signals that reflect the dynamic nature of the cost of electricity supply. Time-of-use (ToU) or, more broadly, Flexible Pricing⁵ structures can be assigned or marketed to customers to encourage them to better match the value they place on using electricity with the cost of supplying electricity at any given time. Although the benefits of Flexible Pricing are only a subset of the many benefits prospectively available from the introduction of AMI⁶, this aspect of AMI is the focus of this study.

In December 2011, the Victorian Government extended the moratorium on the automatic reassignment of customers onto Flexible Pricing by Victorian electricity DBs (although voluntary ToU pricing has remained available to applicable customers on request) through to the end of 2012. The moratorium was introduced in response to community concern about the potential distributional impacts of Flexible Pricing. Its extension provided an opportunity for joint assessment by government, industry and consumer groups of the potential impact of the new pricing, and to ensure that the transition to a new pricing structure is managed carefully and sensibly.

One potential concern was that AMI-enabled Flexible Pricing might penalise the very households that government customer protections and concessions are designed to support — that is, households and small businesses with limited capacity or discretion to respond to new consumption data and price signals to limit their energy consumption.⁷ Some of these households can be identified through their eligibility for concessions (such as disability, aged and veterans pension recipients and unemployment and sole parent payments) or very low consumption patterns (that can indicate significant energy affordability issues). However, more sophisticated data and analysis is required to identify home-based and other small businesses and households who are under energy stress, including the working poor, self-funded retirees and other vulnerable consumers ineligible for Commonwealth-determined concession entitlements. Further, households and businesses can be adversely impacted not just through higher bills, but also through greater volatility in bills placing stress on budgeting.

1.2 Background – AMI customer impacts study: Stage 1

In September 2010, Deloitte was appointed by the Department of Primary Industries (DPI) to determine the potential impacts on domestic and small business consumers of new pricing arrangements enabled by AMI in Victoria, with a focus on those vulnerable customers who may be affected by changing electricity tariffs.

As part of DPI’s request for Deloitte to undertake this work, DPI described the purpose of this AMI customer impact study (the study) as follows:

This project is concerned with determining the potential impacts on domestic and small business consumers of new pricing arrangements enabled by Advanced Metering Infrastructure in Victoria.

The results of this project may be an input into decision-making about interventions to reduce possible financial pressure on some households or businesses. Possible interventions may include concessions, regulatory changes, education and energy efficiency options.

This project was proposed to be undertaken in two stages, the first of which (Stage 1) was completed in 2011. There were two main objectives of Stage 1:

1. To identify which groups are vulnerable to potential increases in energy prices
2. To provide a scenario tool with maximum flexibility for testing a range of ToU simulations.

Stage 1 involved acquiring a significant amount of consumption data from DBs and retail businesses (RBs) in order to identify likely usage patterns under the base case and Flexible Pricing scenarios. Deloitte developed a methodology that would protect the commercial interests of the data providers, whilst still providing useable findings. In Stage 1, we identified the demographic factors most highly correlated with vulnerability, identified areas with the highest percentage of household spend on electricity and then determined which groups (based on the identified demographic factors) were most likely to be living in each area.

Stage 1 identified the following customer groups that are likely to indicate vulnerability:

• Elderly
• Low household net worth
• Low Income
• Needs assistance at home
• Single parent households
• Single income households
• Regional households
• Healthcare concession card holders
• Pension concession card holder
• Single operator business
• Businesses with turnover less than $200,000 per annum.

Using the consumption data provided, average profiles for the vulnerable sections of each customer group could be tested on any number of ToU scenarios⁸. The six Flexible Pricing scenarios that were tested against customer profiles in Stage 1, both with and without applying elasticity sensitivities, include:

• 1A. Single part tariff including CPP component
• 1B. Two part tariff with no seasonal element
• 1C. Two part tariff with no seasonal element but including a CPP component
• 1D. Three part tariff with no seasonal element with off-peak charges fixed at controlled load rate
• 1E. Three part tariff with no seasonal element with off-peak charges fixed at controlled load rate and also including a CPP component
• 1F. Three part tariff with seasonal variation in the application of peak, shoulder and off-peak times.
  

The key findings from Stage 1 relating to residential customer groups are summarised as follows:

• Flexible Pricing will change the existing allocation of electricity costs across customer groups. The customer groups that are ‘winners’ and ‘losers’ are highly dependent on the structure and level of tariffs that are applied, existing tariff levels, whether customers currently have a controlled load off peak tariff ⁹, and how much customers alter their consumption in response to price changes. It is not the case that all customers within certain groups will always be better off or worse off as a result of Flexible Pricing; our analysis demonstrated the likely mean effect of Flexible Pricing on the bills of each group.
• For any particular tariff structure, retailers are likely to be able to adjust the details of the tariff structure (relative level of fixed, peak, off-peak, shoulder and CPP tariffs, and the time periods to which they apply) to increase or decrease the impact on particular customer groups, according to their own business model and objectives and the competitive market environment.
• Under most tariff scenarios, the most marked difference in the impact on customers occurs between single and dual element meter customers (that is, single flat tariff customers and customers who currently have an additional controlled load off peak tariff along with a flat rate ‘peak’ tariff), rather than between vulnerable and non-vulnerable customers. However, this depends on the tariff structure itself and the assumptions applied in the calculation. The impact on dual element meter customers very much depends on how close the new off-peak rate is to the current dual element meter tariff (i.e. current off peak hot water tariffs).
• Average annual bill changes under most of the modelled scenarios are relatively modest and fall in the range of +2% to -4% for the vulnerable groups, assuming zero elasticity.
• If non-zero elasticity is applied then all modelled customer groups experience bill reductions of up to 9% under the tariff scenarios.¹⁰
• Reductions in off-peak rates tend to benefit regional households as they tend to have relatively heavy overnight consumption.
• Under Scenario 1A, all dual element meter residential customer groups modelled are worse off and all single element customer groups are better off.
• Households with individuals requiring disability assistance are usually better off and are only significantly worse off under Scenario A where they have dual element meters.
• Regional households and health care card holders are better off under most modelled scenarios with the exception of Scenario 1A where they have dual element meters.
• Impacts on single income households and low net worth households, including whether they are better or worse off, vary quite markedly depending on the scenario modelled.

As part of Stage 1, we also produced a Microsoft Excel-based simulator that can be used for testing a range of Flexible Pricing scenarios and the effect they might have on electricity prices and bills for the vulnerable customer groups.

1.3 Limitations of the analysis

It is important to note that the Flexible Pricing structures modelled in this report reflect a plausible, although hypothetical, set of tariffs that deliver revenue neutrality for “all residential customers,” being the total group of customers whose data this study is based on. It is also important to note that different outcomes may occur depending on assumptions made with respect to the various tariff structure parameters.

It is for this reason we believe judicious and informed use of the accompanying tariff modelling tool developed as part of Stage 1 is a very important component of being able to derive additional value from this study. The modelling tool allows users to explore the sensitivities of the various parameters to change and how much change to each parameter is required to create degrees of advantage and disadvantage among the various customer groups. The results in this report reflect the outcome of our own assumptions, however where we have adopted assumptions for each tariff part, these are able to be adjusted using the scenario tool to test other assumptions and tariff scenarios.

In performing our work for Stage 2 we have not performed any review or audit work as defined under Australian Audit Standards and consequently no assurance is provided in this report.

1.4 Terms of reference – Stage 2

In accordance of the Acceptance of Quote letter from DPI, dated 28th March 2012, Deloitte has been appointed to carry out Stage 2 of the AMI customer impacts study to validate and conduct further research on the impact of Flexible Pricing on vulnerable customer groups.

As set out in DPI’s Request for Proposal, the objectives of Stage 2 are as follows:

• Validation of the findings of Stage 1 of the study: Stage 1 was undertaken using an aggregation of data, based on the prevalence of certain characteristics at Census Collection District (CCD). Through in-depth primary research and the use of ABS Household Expenditure Survey data, Stage 2 should validate and give further credence or new insight to the findings of Stage 1 in regard to vulnerability of demographic groups.
• Determine consumers’ ability to shift or shed load: In-depth primary analysis should establish how consumers might shift or shed load. What steps would they actually take to respond to new Flexible Pricing plans?
• Identify price sensitivities: Price sensitivity is a measure of how sensitive a customer is to an increase in their bill from different bill levels, for example, whether they consider a particular bill to be too expensive (and therefore unaffordable). Through new research, Stage 2 should identify the degree of price sensitivity expressed by Victorian residential customers and each vulnerable group in particular.
• Bill analysis impacts by quarterly bill: The analysis in Stage 1 focused on changes in annual bills. Stage 2 should analyse in more detail possible seasonal variability that may arise with Flexible Pricing for various groups and scenarios.
• Case studies: Development of a series of case studies to illustrate the impact of Flexible Pricing on a range of vulnerable and non-vulnerable groups.

1.5 Report structure

The remainder of this final report is structured as follows:

• Section 2, "Objective and methodology", describes the approach used to analyse billing data, utilise ABS Household Expenditure Survey (HES) data, and conduct market research in the forms of a telephone survey and focus group discussions
• Section 3, "Impact of ToU scenarios", presents further validation of the vulnerability identified in Stage 1 and explores these issues more fully
• Section 4, "Validation of vulnerable groups using HES data", presents the results of independently validating the presence of vulnerable groups using ABS HES data
• Section 5, "Focus group sessions", provides an overview of the focus groups and summaries of the discussion in each focus group
• Section 6, "Price sensitivity analysis", presents the findings of the quantitative phone surveys
• Section 7, "Case studies", contains six case studies of vulnerable customers developed using observations and quotes taken from the focus groups
• Section 8, "Conclusions and recommendations" summarises our findings and presents some key recommendations for Government relating to the introduction of Flexible Pricing.

2 Objective and methodology

2.1 Overview of approach

In order to meet the objectives of Stage 2, Deloitte carried out a series of in-depth analyses of the data provided by RBs and DBs in Stage 1, in addition to carrying out new primary research.

Stage 2 is divided into the following four main elements:

Quantitative analysis:

• Analysing data collected for Stage 1 to better understand the distribution of effects for defined ToU scenarios
• Utilising the ABS HES data to verify vulnerability identified in Stage 1
• Analysing the sensitivity of vulnerable groups to changes in their electricity bills to further inform the estimates of elasticity used in the Scenario tool produced during Stage 1 (telephone surveys).

The methodology adopted for each of these elements is provided below.

2.2 Development of new tariff scenarios

Since the Stage 1 report was finalised, some progress has been made towards lifting the current moratorium on ToU tariffs through consultation among government, RBs, DBs and customer representatives. Accordingly, it is appropriate to update the ToU tariff scenarios used in Stage 1 that may be introduced to the market once the moratorium is lifted. The Stage 1 Representative tariff (base case) was also updated to reflect retail prices as at January 2012.

All new scenarios modelled are three part tariffs (peak, shoulder and off peak) with varying time periods and price ratios (peak : shoulder : off peak). One scenario (2B) incorporates CPP, with six events occurring in summer and four in winter.

2.3 Analysis of billing data

Stage 1 billing data can be used to understand the impact of ToU on each vulnerable group based on their own profile of electricity use. This is achieved by determining average usage profiles and calculating the distribution of price impact of each of the five ToU scenarios chosen for Stage 2 analysis.

These distributions can then be applied to the ToU simulator to report the distribution of variance, split into the following matrix of results:

• Six ToU Scenarios
• Seven reportable vulnerable groups, plus a general residential group
• Dual element and single element metered households
• Winter, summer and annual bills.

It is noted that, as discussed in the Stage 1 Report, while we were able to identify Single Parent Families and Aged and Veterans Affairs Concession Card Holders as vulnerable, we did not have a statistically significant sample of distributor data to analyse the 24 hour distribution of consumption and was therefore not able to report if these groups would be better off or worse off under different ToU scenarios.

2.4 Analysis of ABS HES data

2.4.1 Objective

The 2009-10 ABS HES is a survey of usual residents of private dwellings in urban and rural areas of Australia covering 97% of the Australian population. It includes data on income and expenditure on electricity. This data is therefore ideal for independently testing the definitions of vulnerability identified in Stage 1 and confirming the existence of vulnerable groups. This has been achieved through the following analysis of ABS HES data:

• Categorising HES households into demographic groups of interest
• Confirming the existence and size of the vulnerable group within each demographic group of interest
• Verifying that vulnerable groups are significantly worse off than the rest of the general population.

The HES data consists of responses from 1,852 Victorian households and is the sample that forms the basis of our analysis¹¹.

2.4.2 Methodology

In order to conduct our analysis, we accessed the CURF (Confidentialised Unit Record File) for the HES survey which allowed us to calculate vulnerability at the household level and perform an analysis of the demographic factors that underlie vulnerability for the purposes of validating findings from Stage 1. The HES collects information on the expenditure, income, net worth and other characteristics of household residents in private dwellings throughout Australia. This survey therefore contains the data needed to independently test the definitions of vulnerability identified in Stage 1 and confirm the existence of vulnerable groups.

Although the Stage 1 vulnerability metric was calculated as the proportion of income spent on the winter electricity bill, there are inherent difficulties in identifying HES expenditure data that pertains to the electricity bill for a household in the winter. For example, when respondents are asked about their last electricity bill, their answer may cover a period of time that includes usage during part of autumn and part of winter. Consequently, all 1,852 Victorian HES records were used for our analysis regardless of the season in which they were collected.

The ABS notes that in the 2009-10 ABS HES, an additional sample of metropolitan households was included in the survey, whose main source of income was a government pension, benefit and/or allowance. Some of the vulnerable groups identified in Stage 1 are likely to be associated with a higher proportion of individuals that receive government assistance, and accordingly the additional sample in the ABS HES helps ensure that a sufficiently large sample for each vulnerable group is available for analysis.

2.4.3 Other analysis

Additional data modelling was performed to explore other factors recorded in the ABS HES data that are associated with the vulnerability measure. These techniques include:

• Correlation analysis: identifies demographic factors and HES metrics that are strongly correlated with vulnerability.
• Regression decision trees: builds a series of rules that identifies clusters of households with a high degree of vulnerability.
• Self-organising maps: through the clustering of households on the basis of similarity across all HES metrics and factors, this technique tests whether vulnerability is a defining characteristic of any particular cluster or subset of households.

Each technique has its own strengths and using all three provides a comprehensive exploration of factors associated with vulnerability.

2.4.4 Identifying Stage 1 Demographic Groups

In Stage 1, entire CCDs were characterised as exemplifying a vulnerable group. However, the ABS HES data does not include the CCD of participants in the survey and instead individual households can only be associated with the demographic groups identified in Stage 1. The following table indicates the criteria used to create this association.

Table 3: Criteria for identifying ABS HES households by demographic group

Stage 1 Demographic Group	Criteria 12
Elderly	The age of the individual participating in the survey is 65 years old or above (AGERHBC).
People requiring disability assistance	Someone in the household aged 15 and over has a disability or long-term health condition (DISBHH1).13
Single income households	There is only one income earner in the household (NOEARNBC).
Low income households	The current "maximum gross income" criterion for a Low Income Health Care Card is met for the household14. This is based on weekly household income from all sources (INCTSCH8): Single, no children: $483 per week Couple, no children: $838 per week Single, 1 dependent child: $838 per week An extra $34 for each additional child has been intentionally omitted in the above criteria. Given the potential measurement error in the recording of weekly income in the ABS HES, the value of this adjustment could result in some households being overlooked as low income households.
Regional	Unable to be reliably determined from HES data15.
Single parent households	Any of the family types where there is a lone parent with dependent children (FAMTYPE).
Low net worth households	Households that are in the bottom 20% of values for net wealth of household (i.e. less than $57,600) as determined from the entire 9,774 households in the ABS HES (WEALTHH).
Health care card holders	There are one or more people in the household with a current health care card (DNHCCBC).

It is noted that a household can satisfy the criterion for more than one demographic group. For analytical completeness, ABS HES household data will be included in the analysis for each applicable demographic group to ensure statistically representative results are obtained. This differs from Stage 1 where a CCD was exclusively assigned to a particular vulnerable group.

2.4.5 Measuring Income and Expenditure

All ABS HES data on income and expenditure was recorded as an average weekly figure. All possible sources of income were included, such as superannuation, rent, investments, etc. To determine electricity expenditure, the survey respondent is asked the value of the last electricity bill and the period of time that the bill covers, enabling a weekly electricity expenditure figure to be calculated. If the last electricity bill could not be confidently recalled or determined, no value was recorded.

2.4.6 Defining vulnerability

In this study, ‘vulnerability’ refers to ‘vulnerability to bill increases’, that is, should Flexible Pricing have a negative impact (an increase) on electricity bills, the focus of this study is on those households that are least able to afford such an increase. In Stage 1, the vulnerability metric was defined as the average household percentage of income spent on electricity, calculated at the level of CCD. For the ABS HES data, the vulnerability metric is instead calculated for each household as the percentage of income spent on electricity.

Once vulnerability factors had been identified, CCDs containing threshold representation of households exhibiting the relevant vulnerability factor also had to be flagged. This was achieved through a two-step process:

• Identifying vulnerability – We determined a level of CCD average Winter expenditure on electricity above which customers in that CCD would be considered to be “vulnerable” by inspecting a distribution of vulnerability by CCD.
  • For the purposes of identifying vulnerable residential customers, the point at which expenditure on electricity exceeds that the expenditure of the bulk of the population was deemed to be 4% of income (approximately two standard deviations from the mean).
• Identifying cut-off representation of vulnerability factors – Each CCD has a proportion of households exhibiting the relevant vulnerability factor. In order to determine the appropriate point of cut-off for inclusion of a CCD within the population of CCDs relevant to a vulnerability factor, a simple linear equation was derived – see the example in Figure 2 where:
  • the proportion of expenditure on electricity was the dependent variable (vertical axis) and
  • the proportion of households exhibiting the relevant vulnerability factor was the independent variable (horizontal axis).

Figure 1: Distribution of residential winter vulnerability

Figure 2: Selecting CCDs for inclusion in analysis of household type – overview

With respect to residential customers, in most cases, the 4% threshold for residential electricity spending was too high to deliver a sufficient number of CCDs to support the analysis, so the threshold was lowered to 3% in order to find an appropriately fitted linear model. It should be noted that this is a conservative threshold for defining vulnerability to Flexible Pricing, which in turn was used to ensure there was sufficient sample sizes for analysis. A higher threshold would mean that there would not be enough data to conduct a meaningful analysis. Table 4 shows the result of this analysis and the determined vulnerability factor cut offs – that is, the representation of households exhibiting the relevant vulnerability factor in order for the CCD to be incorporated in analysis of the relevant group of interest.

For each demographic group of interest, vulnerability can be validated by comparing the distribution of the vulnerability metric against the Stage 1 vulnerability criteria. The following table lists the vulnerability criteria determined in Stage 1, where a vulnerable value greater than this cut-off resulted in the demographic group of a CCD to be classified as vulnerable.

Table 4: Stage 1 criterion for defining vulnerability by demographic group

Stage 1 Group	Criterion*
Elderly	3.0%
People requiring disability assistance	3.0%
Single income households	3.0%
Low income households	3.0%
Regional	N/A
Single parent households	2.8%
Low net worth households	2.4%
Health care card holders (referred to in Stage 1 as ‘Aged and veteran pensions’)	2.6%

* The criterion is a percentage of household income spent on electricity bills, based on 2009-2010 electricity billing data. The percentages shown for each group were determined in Stage 1.

2.5 Quantitative research of sensitivity to ToU pricing

In order to better understand how different groups of vulnerable consumers will be affected by electricity price changes, Deloitte conducted price sensitivity analysis on responses to a telephone survey. This provided an insight into ‘tipping points’ where consumers are likely to make changes to energy consumption behaviour, such as opting in for new pricing such as ToU tariffs, CPP and incentives, direct load control, etc. The telephone survey also included questions on the ability of consumers to shift or reduce consumption in peak times.

In order to perform this analysis, we designed a survey of 3,000 households designed on a methodology defined in the Van Westendorp Price Sensitivity Meter. This approach is ideal for highly commoditised products such as electricity where the consumer can make a relatively clear judgement between price and utility. Amongst other questions, the survey asked respondents to indicate hypothetical electricity prices they would characterise as:

• "too cheap" – for example, that they would be concerned an error was made by the electricity company;
• "cheap" – for example, their bill would be less than expected but that they could explain the reduction through their efforts to conserve electricity;
• "expensive" – for example, their bill was higher than expected but that they could attribute the increase to extra electricity use;
• "too expensive" – for example, their bill was so expensive they would be concerned a mistake had been made by the electricity company and/or they would have trouble paying.

The results were then collated into a price sensitivity meter which shows the range of acceptable prices to the given group under study.

Figure 3: Example graphical representation of analysis of Van Westendorp study*

* Sensitivity to price changes is based on the average gradient of the two thick lines shown in the diagram

The price sensitivity meter is used to calculate the sensitivity of consumers to price increases. The advantage of the Van Westendorp approach is that it allows the visual representation of how the rate of price sensitivity changes over a large range of price. The sensitivity score is calculated as the average gradient of the thick lines shown in Figure 3 (see Appendix B for details on how this is calculated).

The sensitivity score is an indication of affordability on the basis that the electricity bill has already increased to a value where households start to consciously change their behaviour. In other words, the situation being considered is that bills are already at a threshold that would cause concern for households.

The score is interpreted as the decrease in the percentage of consumers who consider they could afford a 1% increase in bill within the "stress zone" illustrated in Figure 3. The “stress zone” is the region around the mean electricity price for the group defined as the price range between the marginal cheap price and the marginal expensive price. For example, a sensitivity score of 0.5 means that for every 1% increase in electricity bill, 0.5% fewer consumers consider they are able to afford such an increase.
Comparing sensitivity scores between vulnerable groups and against the ‘Residential’ customer group confirms that the consumers in the vulnerable groups perceive themselves as being more vulnerable to the impact of price increases relative to the rest of the population. We note that the relationship between bill increase and decrease in customers is one-way, that is, it specifically relates to the situation where bills are already at a threshold that would cause concern for households should the electricity bill increase further. It does not indicate the possible increase in affordability amongst households if electricity bills were to be decreased.

Other questions posed in the telephone survey include asking consumers how easy it would be to shift or reduce electricity usage out of peak times. Analysed in conjunction with the Van Westendorp results, vulnerable groups can be classified on their vulnerability to bill increases as well as their ability to accommodate those bill increases through changing their consumption behaviour.

2.6 Qualitative market research and case study development

As part of Stage 2, Deloitte conducted a series of focus groups to investigate and carry out the following three tasks:

1. Validate the demographic assumptions developed in Stage 1 and investigate the attitudes of these members of these groups to Flexible Pricing
2. Explore the possibilities for members of these groups to shift their demand to non-peak periods
3. Test the impact of changes to electricity tariff structures on these households
4. Identify drivers of higher electricity costs.

Given that we have confidence in the statistics of our identified vulnerable groups, although less confidence in the assumptions we have made about their characteristics, the focus groups provided a good opportunity to give these vulnerable consumers their own voice within the findings of the report.

The focus groups provided us with the capability to informally validate the assumptions drawn from the data in Stage 1 and to triangulate results from other analyses produced by both stages of the study.

Participants were chosen in order to make sure we had representation of the types of customers that were identified as vulnerable in Stage 1. This was done in order to confirm findings from data analysis and also to explore issues for these groups that could not be discovered through data analysis alone.

Participants were paid $75 each to attend the sessions. In total, 42 people attended six sessions – four sessions were held in Melbourne and one session each was held in Bendigo and Traralgon. The size of each focus group was determined by the optimum number and background mix of participants to produce the appropriate level of discussion and engagement. We also conducted a mix of daytime and evening groups to ensure that obtained sufficient representation of groups with different time of day availability.

The following table shows the number of participants and location of the focus groups conducted.

Table 5: Focus groups conducted for Stage 2 qualitative market research.

Group	Location	Number of Participants
Elderly; aged over 65 years to 80 years or Disabled	From Inner to Mid distance suburbs of Melbourne	7
Low income/single income with some first home buyers/Heath care card	Even mix across inner, mid and outer suburbs	7
Single parents	Drawn from outer suburbs of Melbourne	6
Elderly; aged over 65 years to 80 years or Disabled	Drawn from outer suburbs of Melbourne	8
Low income/single income with some first home buyers/ health care card - at least two farm workers	Bendigo region	8
Single parents -at least two farm workers	Latrobe valley (½ from Traralgon and ½ from surrounding towns)	6

3 Impact of ToU scenarios

3.1 Development of new tariff scenarios

3.1.1 Base case

Our analysis in Stage 1 incorporated base case (business as usual) residential and small and medium enterprises (SME) electricity tariffs for comparison against the six new tariff scenarios. The base case tariffs were needed in order to model the change in likely customer bills attributable to the new tariff scenarios.

The Stage 1 ‘Representative Tariffs’ were developed using:

• Known state-wide market shares of each RB¹⁶
• Standing and market offer tariffs published by each RB in each DB region
• ESC estimates that 65% of Victorian customers have entered competitive market contracts with either their own or a new retailer¹⁷
• Anecdotal evidence suggesting it is possible to save around 10% annually on electricity if shifting from a standing offer to a market offer.

The Representative Tariffs used in Stage 1 were based on a weighted average of retail prices as at January 2011.¹⁸

The Stage 2 analysis was focussed only on residential customers and accordingly no SME Representative tariff or new tariff scenarios were developed.

In order to update the base case residential tariff to 2012 for Stage 2, we have used updated data on retail electricity tariffs as at January 2012, which was collated by the St Vincent de Paul Society as part of the Victorian Tariff Tracking Project.¹⁹ We have used the same methodology used in Stage 1 to determine the Representative Tariffs, being to develop a weighted average tariff based on known RB and DB market shares published by the ESC.²⁰ The updated Representative Tariffs are presented in the following table.

Table 6: Representative tariffs – March 2012 (including GST)

	Daily supply charge	Peak rate / kWh	Off-peak (controlled load) rate / kWh
Residential	89.61 cents	22.30 cents	13.06 cents

3.1.2 ToU tariff scenarios

As discussed above, for Stage 2, six new tariff scenarios were developed for analysis. The new tariff scenarios stem from consultation between DPI and industry stakeholders and therefore better reflect the likely tariffs that will be introduced into the market once the moratorium on Flexible Pricing is lifted.

The new tariff scenarios tested in Stage 2 are all three-part tariffs and do not contain any seasonal variations (that is, neither rates nor times for peak/off-peak/shoulder will vary with seasons). Most of the variation between these tariffs stems from the different peak/off-peak/shoulder ratios (with peak rates either three or two times the off peak rates), as well as the different times they will apply (i.e. narrow peak period, wide peak period). The peak/off peak/shoulder rate differentials were selected on the basis that they reflect the current ToU rate differentials on the market. Only one tariff contains CPP (tariff 2B).

The tariffs and assumptions used are presented in the following tables.

Table 7: Summary: Tariff scenarios modelled in Stage 2 (cents, including GST)*

Tariff scenario	Standing (daily) charge c/day	Peak rate c/kWh	Shoulder rate c/kWh	Off peak rate c/kWh
Base case (current representative tariff)	89.61	22.30	n/a	13.06 (controlled load rate)
2A – Three part tariff with narrow peak times	89.61	38.73	22.59	12.91
2B – Three part tariff with narrow peak times and CPP	89.61	32.16	21.00	12.00
2C – Three part tariff with wide peak times	89.61	37.26	21.74	12.42
2D – Three part tariff with wide peak times and lower peak price, relative to off peak and shoulder	89.61	30.32	22.74	15.16
2E – Three part tariff with wide peak times, weekends at off peak rate	89.61	40.51	23.63	13.50
2F – Three part tariff with wide peak times which starts and ends one hour later than Scenario C	89.61	36.94	21.55	12.31
*Note: Both the level of the tariff (c/kWh) and the time in which peak/off peak/shoulder tariffs apply will affect the overall revenue and bills.

Table 8: Tariff 2A – three part tariff with narrow peak times

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	38.73 cents, including GST
	Shoulder rate charge (c/kWh)	22.59 cents, including GST
	Off-peak charge (c/kWh)	12.91 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	4pm to 8pm weekdays
	Shoulder	7am to 4pm & 8pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	n/a
	Energy used at CPP peak times compared standard peak times	n/a
	Calls per year	n/a
	Own-price elasticity	n/a
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

Table 9: Tariff 2B – three part tariff with narrow peak times and CPP

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	32.16 cents, including GST
	Shoulder rate charge (c/kWh)	21.00 cents, including GST
	Off-peak charge (c/kWh)	12.00 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	4pm to 8pm weekdays
	Shoulder	7am to 4pm & 8pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	CPP = 5 x peak charge
	Energy used at CPP peak times compared standard peak times	CPP energy = 2 x standard peak energy
	Calls per year	6 Summer; 4 Winter
	Own-price elasticity	-0.025
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

Table 10: Tariff 2C – three part tariff with wide peak times

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	37.26 cents, including GST
	Shoulder rate charge (c/kWh)	21.74 cents, including GST
	Off-peak charge (c/kWh)	12.42 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	2pm to 8pm weekdays
	Shoulder	7am to 2pm & 8pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	n/a
	Energy used at CPP peak times compared standard peak times	n/a
	Calls per year	n/a
	Own-price elasticity	n/a
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

Table 11: Tariff 2D – three part tariff with wide peak times and lower peak price, relative to off peak and shoulder

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	30.32 cents, including GST
	Shoulder rate charge (c/kWh)	22.74 cents, including GST
	Off-peak charge (c/kWh)	15.16 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	2pm to 8pm weekdays
	Shoulder	7am to 2pm & 8pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	n/a
	Energy used at CPP peak times compared standard peak times	n/a
	Calls per year	n/a
	Own-price elasticity	n/a
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

Table 12: Tariff 2E – three part tariff with wide peak times, weekends at off peak rates

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	40.51 cents, including GST
	Shoulder rate charge (c/kWh)	23.63 cents, including GST
	Off-peak charge (c/kWh)	13.50 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	2pm to 8pm weekdays
	Shoulder	7am to 2pm & 8pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	n/a
	Energy used at CPP peak times compared standard peak times	n/a
	Calls per year	n/a
	Own-price elasticity	n/a
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

Table 13: Tariff 2F – three part tariff with wide peak times

Parameter type	Element	Description
Basic charges - residential
	Daily supply charge (c/day)	89.61 cents, including GST (as per base case)
	Peak rate charge (c/kWh)	36.94 cents, including GST
	Shoulder rate charge (c/kWh)	21.55 cents, including GST
	Off-peak charge (c/kWh)	12.31 cents, including GST
Seasonal application
		No variation by season**
Weekday / weekend application of charges
	Peak	3pm to 9pm weekdays
	Shoulder	7am to 3pm & 9pm to 10pm weekdays; 7am to 10pm weekends
	Off-peak	All other times
Own-price elasticity
	Peak	-0.15
	Shoulder	-0.15
	Off-peak	-0.05
Substitution effects*
	Peak to shoulder	0.10
	Peak to off-peak	0.15
	Shoulder to off-peak	0.20
Critical peak price
	CPP as multiple of peak price	n/a
	Energy used at CPP peak times compared standard peak times	n/a
	Calls per year	n/a
	Own-price elasticity	n/a
* Percentage of load reduced as a consequence of own-price elasticity effect that is shifted to another time of day. ** Tariff segments have the same commencement and end times in each season in accordance with Daylight Savings.

The assumptions underpinning the new tariffs are unchanged from Stage 1. Definitions of the various elasticities are presented in the following box.

As discussed in the Stage 1 report, the elasticity assumptions we adopted were developed on the basis of a literature study of ToU trials and studies.²¹

The relative tariffs were generated on the basis of determining an appropriate peak-shoulder-off peak ratio and times, assuming revenue neutrality with the base case across the population (being the sample of all residential customers).²²

Consistent with the approach taken in Stage 1, for tariff scenario 2B which includes a CPP, we have used a CPP elasticity that results in the CPP impact being an overall reduction in energy usage of less than 20%. This is approximately equal to the change in CPP usage recorded by residential customers in trials by Ausgrid.23 Also consistent with Stage 1, CPP events are modelled to occur where in the base case consumption profile, peak energy use in a day (that is, total energy used over the determined peak period) is equal to double the peak time energy use in that period on the average day, before any elasticity (or ToU tariff) is applied.

It is important to note that in calculating the tariff impacts for Stage 1, the results were presented in the report were both with and without elasticity assumptions. Results presented with an elasticity effect were calculated such that revenue was held constant before elasticity assumptions were applied. The same approach has been carried out for Stage 2.

For simplicity we have assumed the same elasticities across all customer groups. In reality this may not be the case – for example pensioners may have a lower ability to reduce (or shift) usage compared to the broader residential group given their relatively low base level of consumption. However, there is insufficient data and no relevant empirical studies that allowed an accurate estimation of elasticities for individual customer groups.

3.2 Distribution of Impacts of Flexible Pricing Scenarios

In general, the distribution of impact is quite broad for all groups under all scenarios. While particular groups may be better or worse off, on average there is still a broad range of price effects in each group.

Figure 4 presents the annual bill change for customers under the various scenarios. The graphs identify the different impacts of the tariff scenarios on each vulnerable group. These graphs demonstrate that, on average, most groups will typically experience no more than a 5% increase in annual bills under the various scenarios.

While there are variable average impacts across all groups, the following points can be drawn from the results:

• Without applying an elasticity effect (that is, assuming customers do not change their behaviour in response to Flexible Pricing), single element meter customers are likely to experience a greater increase in bills (or, in most cases, a smaller decrease in bills) than dual element meter customers under most tariff scenarios. This result occurs because single element meter customers spend proportionally more on peak energy than off peak energy than dual element meter customers. Although under some of the Flexible Pricing scenarios, the off peak price is increasing relative to the base case, the peak price increase is substantially greater than the off peak price increase. This means that the relative bill change for single element meter customers is likely to be a greater increase or smaller decrease than that for dual element meter customers.
• However, when applying an elasticity effect and allowing total revenue to change (fall), the vast majority of groups will be, on average, better off.
• Low income households, people requiring disability support, health care card holders and regional groups are likely to be, on average, better off under each tariff scenario than under the base case, even assuming that they don’t respond to pricing incentives (i.e. even without an elasticity effect)
• Low net worth households and single income households are, on average, worse off (or in most cases, less better off) under each tariff scenario than the other vulnerable groups.
• Adding a CPP (in Tariff 2B) does not make a significant difference to the impact of ToU on each customer group, aside from low net worth households for which it has a small negative effect.
• Tariff 2D (wide peak times, weekend at off peak rates) will result in the smallest changes to customer bills. Allowing for customer response (elasticity assumptions) to Tariff 2D provides an average reduction in bills of less than 1%.

Regarding the graphs in Figure 4:

• Consistent with our approach in Stage 1, the graphs indicate the variation in the annual electricity bill for each group of customer following the application of various tariff scenarios under two different circumstances:
• 1. The darker (blue) bars indicate the variation in the annual electricity bill assuming zero elasticity and revenue neutrality
  2. The lighter (green) bars indicate the variation in the annual electricity bill assuming non-zero elasticity – revenue neutrality is not imposed.
• Slightly different billing breakdowns were supplied by different RBs, therefore for customers on a concession the billed amount was defined as the total bill, including GST and all discounts. Therefore the electricity bill data used for analysis included applicable concessions and rebates.

Figure 4: Average annual price change for customers in each group under the six ToU scenarios

Figure 5 shows the distribution of annual bill effects for each single element group under the six scenarios with elasticity effects. Figure 6 shows the corresponding graphs for dual element groups with elasticity effects. (The values used to create these graphs are shown in Appendix A, where corresponding values for no elasticity effect can also be examined).

Price effects have been graphed using box-and-whisker charts that indicate the extent of each demographic group clustering around the average. The “whisker” line shows the extent to which 90% of the given population is spread, while the box shows the spread of effect for the middle 50% of the population.

The blue dot represents the average annual bill impact for the group, which may be pulled from the exact centre of the box if there a few bills with an extremely large or small change in bill under the scenario.

As an illustrative example (exact values were determined from the table in Appendix A): under Scenario A, for the ‘All residential’ single element meter group, the average result was a decrease of -1.2%, while the middle 50% of the population having a bill change in the range of a deduction of -24.3% to an increase of 12.0%. Looking at it another way, 50% of the population will experience either a reduction in bills of greater than 24.3%, or an increase in bills of more than 12.0%.

For ‘Regional’ customers with single element meters, the average result of Scenario A was a decrease in annual bills of -4.7%, with the middle 50% of the population having a bill change in the range of a reduction of -35.6% to an increase of 15.6%.

Two general conclusions can be draw from these graphs. First, the exact design of the ToU scenarios makes little difference to the overall distribution of bill impact.

Second, there is a very broad range of effect for each group which indicates that belonging to a particular group is of itself not the most important factor in determining bill impact; rather, there are a range of factors that contribute to bill impact in combination with general demographics. These conditions are explored further in the qualitative analysis conducted in Section 5, “Focus group sessions”.

Figure 5: Annual change in electricity bill for single element customers, with elasticity effect

Figure 6: Annual change in electricity bill for dual element meter customers, with elasticity effect

Table 14 and Table 15 summarise the proportion of each vulnerable group that would be better off under each of the ToU scenarios, for both single and dual element meter customers respectively. As discussed in section 1.2 above, dual element meter customers are those with a controlled load, such as electric storage hot water, for which they currently receive an overnight off peak rate. For dual element meter customers, shifting to Flexible Pricing will mean that their hot water energy use is charged at the new off peak rate, which may be higher than their current controlled load off peak rates.

However, as discussed above, depending on the proportional changes in peak and off peak prices, single element meter customers may experience a more significant annual bill change than dual element customers if more of their electricity use is at peak times.

Generally, 50% or more of each single-element and dual-element group is expected to experience a reduction in annual electricity bills under each ToU scenario, although only 40% of single-element health care card holders would experience a benefit.

It is noted that while these tables show that some customers may be worse off under Flexible Pricing scenarios, the average annual bill increase typically does not exceed 2%, even assuming there is a zero elasticity effect and holding RB revenue constant. In most cases, a decrease in the average annual bill occurs particularly when allowing for elasticities.

A full listing of average bill change is presented in Appendix A for all the combinations of element type, season, customer group and tariff scenario. The estimated proportion of customers in each group expected to be better off is also included.

Table 14: Estimated proportion of single element meter groups better off under each ToU Scenario (with elasticity effect)*

Group	Element	ToU Scenario A	ToU Scenario B	ToU Scenario C	ToU Scenario D	ToU Scenario E	ToU Scenario F
All residential	Single	59%	58%	58%	58%	60%	59%
Non-vulnerable	Single	57%	55%	57%	57%	59%	57%
Elderly	Single	55%	54%	55%	54%	56%	55%
Needs assistance**	Single	49%	48%	50%	49%	51%	50%
Single income households	Single	58%	57%	58%	57%	59%	58%
Low income	Single	59%	59%	59%	58%	60%	59%
Regional	Single	60%	60%	60%	59%	61%	60%
Low net worth	Single	58%	57%	58%	58%	61%	58%
Healthcare card holders	Single	41%	40%	41%	41%	43%	41%
* The results in this table do not reflect the magnitude of annual bill changes, which are typically no more than a 2% annual bill increase, on average, for most combinations, although as noted there is much variation around the mean. ** The sample size for the “Needs Assistance” group was smaller than required to consider these results statistically representative of the population of Victorian households needing assistance. These results should be used with caution.

Table 15: Estimated proportion of dual element meter groups better off under each ToU Scenario (with elasticity effect)*

Group	Element	ToU Scenario A	ToU Scenario B	ToU Scenario C	ToU Scenario D	ToU Scenario E	ToU Scenario F
All residential	Dual	56%	56%	57%	53%	57%	57%
Non-vulnerable	Dual	59%	58%	60%	56%	60%	60%
Elderly	Dual	64%	64%	64%	62%	66%	65%
Needs assistance**	Dual	>95%	>95%	>95%	>95%	>95%	>95%
Single income households	Dual	58%	58%	58%	54%	58%	59%
Low income	Dual	54%	54%	54%	51%	54%	55%
Regional	Dual	61%	61%	62%	59%	62%	62%
Low net worth	Dual	51%	51%	52%	49%	52%	52%
Healthcare card holders	Dual	66%	67%	67%	62%	67%	67%
* The results in this table do not reflect the magnitude of annual bill changes, which are typically no more than a 2% annual bill increase, on average, for most combinations, although as noted there is much variation around the mean. ** The sample size for the “Needs Assistance” group was smaller than required to consider these results statistically representative of the population of Victorian households needing assistance. These results should be used with caution.

3.3 Quarterly impact of ToU Pricing

The distribution analysis of Flexible Pricing scenario impacts in section 3.2 is focused on the overall annual change in electricity bills.

However, household electricity bills are typically quite varied across quarters, with summer and winter bills reflecting the increased heating and cooling loads in those quarters, depending on the appliance mix of households. We have carried out some analysis of the quarterly bill impacts of tariff scenario 2F on each group to give an indication of the general impact of Flexible Pricing throughout the year.

In general, the distribution of bill impacts across quarters is quite varied for all groups, with differences between dual and single element meter customers reflective of the same annual bill outcomes. The results show that summer bills are the most negatively affected by Flexible Pricing in that the bill change is either a greater increase or smaller decrease than other quarters. With the exception of the elderly group which shows autumn bills as being more negatively affected than spring bills, autumn and spring bills are affected to a similar degree. Winter bill changes are quite varied across all groups and dual/single meter customers.

Figure 7 illustrates the quarterly analysis results.

Figure 7: Comparison of change in electricity bill across quarterly bills for Scenario F

4 Validation of vulnerable groups using HES data

4.1 Prevalence of Demographic Groups

The following table indicates the percentage of the 1,852 ABS HES Victorian survey respondents that fall into each of the Stage 1 demographic groups of interest. It is noted that many ABS HES households satisfy the criteria for more than one vulnerable group. As these households can be assigned to more than one group, the percentages shown in the table across the groups do not add up to 100%.

Table 16: Percentage of Victorian HES data records by demographic group

Demographic Group	Total number of records in ABS HES	% of Victorian ABS HES records
Elderly	1,071	28.6%
People requiring disability assistance	1,877	45.1%
Single income households	770	25.6%
Low income households	1,908	46.8%
Regional	N/A	N/A
Single parent households	348	7.9%
Low net worth households	794	14.1%
Health care card holders	2,416	61.5%
Non-vulnerable (i.e. not assigned to any of the above groups)	287	15.5%

The above table highlights the prominence in the ABS HES of Victorian-based low income households, people requiring disability assistance and health care holders. Single parent households and low net worth households based in Victoria have a small representation amongst ABS HES households. Results pertaining to groups of such small representation may be not representative of the demographic as a whole and reliable conclusions cannot be made about these groups.

It is noted that:

• The percentage shown in Table 16 does not reflect the prevalence of these groups in the Victorian population. For example, whilst 61.5% of households in the ABS HES had health care cards, this does not suggest that a similar percentage of Victorians have health care cards. This is not an issue for this analysis as the groups are not being combined to form a ‘whole of Victoria’ view and are only being analysed separately.
• All demographic groups shown in Table 16, except for “Non-vulnerable” (households that do not fit into one of the specific demographic groups), have a sufficiently large sample size to consider them representative of the Victorian population.
• The “Non-vulnerable” group has significantly less than the minimum statistical representative baseline of 380 households. The suitability of this demographic group to represent the corresponding population of Victorian households will be explored in the analysis of ABS HES data in the following section.

4.2 Confirmation of vulnerable groups

The methodology for identifying vulnerable groups in Stage 1 cannot be used as the ABS HES records do not identify the location (e.g. no CCD information) of households. Instead, comparisons were made between each of the demographic groups of interest and the “Non-vulnerable” group (i.e. the rest of households that had not been identified with any of the demographic groups of interest from Stage 1). Each of these groups were also assessed against the vulnerability criterion determined in Stage 1, enabling the relative size of the vulnerable group within each demographic to be determined.

The following graph presents the weekly income versus weekly electricity bill for the “Non-vulnerable” group. It reveals almost all households below the red line, indicating few households exceeding the vulnerability criterion of 3%. Further investigation of the few points that do exist above the 3% line does not reveal any unifying demographic characteristics. There are also no distinguishable groups of households close to the 3% criteria that might suggest a demographic with “borderline” vulnerability.

Figure 8: Scatterplots of income versus expenditure for the "Non-vulnerable" group

The above graph is consistent with the findings from Stage 1 that amongst the non-vulnerable population the proportion of income spent on electricity typically does not exceed 3%. This also verifies that this “Non-vulnerable” group in the ABS HES is representative of Victorian households in the general population, despite not meeting the minimum statistical representative baseline as noted above. The data for this demographic can therefore be reliably used for comparison against vulnerability evident in the other demographic groups.

The following graphs present the weekly income versus weekly electricity bill of the remainder of the demographic groups. From these graphs it is evident that each group has a significant vulnerable group of households, that is, households that lie in the region above the 3% vulnerability criterion. As expected, the households that fall in this area of each graph are those with the lowest income in each demographic. Of particular note is the low income demographic, which has distinctly more households above the vulnerable criterion than other demographics.

Figure 9: Scatterplots of income versus expenditure for each demographic group

The contrast between these graphs (Figure 9) and that for the “Non-vulnerable” group (Figure 8) is evidence of the vulnerability that exists within these specific demographic groups in the community. The actual proportion of Victorian ABS HES households above the vulnerability criterion is shown in Table 17 below. These figures correspond to the percentage of data points above the red vulnerability line in Figure 8 and Figure 9.

Table 17: Percentage of vulnerable Victorian ABS HES households in each demographic

Demographic Group	% of vulnerable Victorian ABS HES households
Elderly	46.4%
People requiring disability assistance	43.5%
Single income households	17.5%
Low income households	58.1%
Regional	N/A
Single parent households	40.6%
Low net worth households	42.0%
Health care card holders	45.0%
Non-vulnerable (i.e. not assigned to any of the above groups)	1.8%

4.3 Distribution of vulnerability

The graphs on the following page explore the distribution of vulnerability in each demographic group against the “Non-vulnerable” group. They use the same data presented in Figure 8, but shown instead as a boxplot to illustrate the distribution of values and being side-by-side to the “Non-vulnerable” group to make for easy comparison.

Figure 10: Distribution of vulnerability compared between each demographic group and the “Non-vulnerable” group

Except for ABS HES households identified as having only a single income:

• 75% of households in each of the demographics of interest spend a greater percentage of their income on electricity bills than 75% of the “Non-vulnerable” group
• Approximately 50% of the households in each of the demographics of interest exceed the vulnerability criteria identified for that group in Stage 1
• The percentage of income spent on electricity bills for the top 25% of households in each of the demographics of interest is typically between 4% and 7%.

For single income households, only about 50% spend a greater percentage of their income on electricity bills than the “Non-vulnerable” group. Less than 25% of single income households exceed the vulnerability criteria determined for them in Stage 1. The top 25% of single income households spend between 2.5% and 5% of their income on electricity bills.

This analysis has confirmed the ‘degree of vulnerability’ among the vulnerable groups identified during Stage 1, being defined as expenditure on electricity representing more than 3% of their household income. This gives confidence to our analysis of the impact of Flexible Pricing on vulnerable groups by confirming that the majority of the household consumption data used in our analysis is indeed drawing from customers considered to be vulnerable.

4.4 Further analysis of vulnerability

Further modelling to uncover associations between vulnerability and other factors and metrics recorded in ABS HES data was performed using the techniques described in Section 2.4, “Analysis of ABS HES data”:

• Correlation analysis
• Regression decision trees
• Self-organising maps

Through the application of the above techniques to the HES data, it would be expected that any vulnerable groups not identified in Stage 1 could be highlighted. However, a rigorous application of these techniques did not identify any new vulnerable groups or subsets of the population for which the impact of Flexible Pricing would be of concern. It is therefore concluded that the vulnerable groups identified in Stage 1, which are examined in this report, are the only noteworthy groups to explore at this time.

5 Focus group sessions

5.1 Overview of approach

In the time period allowed for this study, comprehensive primary research on elasticity of demand for electricity among Victorian customers was not possible. Instead, in order to identify the potential behavioural responses to Flexible Pricing and the likely attitude of different vulnerable groups to changes in electricity prices, a qualitative method was applied.

To investigate the findings of Stage 1, we conducted a series of focus group sessions to introduce and discuss the concepts of Flexible Pricing, including Critical Peak Pricing and direct load control with vulnerable customers. Due to the nature of focus groups, no quantitative analysis of the outputs is possible, however, the discussions were designed to contribute qualitatively to our analysis and test our understanding of how consumers might respond to Flexible Pricing.

The aim of the focus groups was to:

1. Validate the demographic assumptions developed in Stage 1 and investigate the attitudes of these members of these groups to ToU tariffs
2. Explore the possibilities for members of these groups to shift their demand to non-peak periods
3. Test the impact on these households of changes to electricity tariff structures
4. Identify drivers of higher electricity costs.

5.2 Focus groups conducted

Six focus group sessions of two hours each were conducted over a fortnight in April 2012²⁴. Participants were selected through telephone interviews randomly selected from the Victorian population. Telephone calls were made at various times of the day and week to ensure that the participants would be as representative as possible of the group they were being recruited for.

Of the 42 participants:

• 35 indicated they live in a single income household
• 20 stated their range of annual household income as “Less than $25,000 per annum”
• Only two participants indicated they had an annual household income of greater than $45,000
• 32 indicated they held a Health Care Card and/or Pensioner Card.
  The following table presents an overview of the demographic features of each session.

Table 18: Focus group sessions conducted

Session	Location and time	Demographic groups represented
1	Melbourne, weekday lunchtime	Elderly; aged over 65 years to 80 years or Disabled
2	Melbourne, weekday evening	Low income/single income with some first home buyers/Heath care card
3	Melbourne, weekday morning	Single parents
4	Outer Melbourne, weekday lunchtime	Elderly; aged over 65 years to 80 years or Disabled
5	Bendigo, weekday evening	Low income/single income with some first home buyers/ health care card - – at least two farm workers
6	Latrobe Valley (half from Traralgon, half from other areas), weekday evening	Single parents -at least two farm workers.

5.3 Discussion guide

The purpose of each session was explained to participants as ‘helping us to explore new ideas about electricity use and pricing’. It was not immediately made clear to participants that the conversation was related to smart metering or Flexible Pricing; in many cases smart metering was not mentioned until participants had had an opportunity to discuss their energy bills and the drivers of increased costs.

Discussions commenced with an introduction by each participant and a description of their household and lifestyle with regards to electricity use. Participants were asked to list their major appliances (including whether they are driven by gas or electricity) and when they are generally used.

Questions around participants’ electricity and gas bills were then asked. Those who could recall their most recent and their highest bills were asked to do so, which sparked some conversations comparing bills between participants. People were asked to comment on whether they thought their bills were high and what was driving any increase in costs.

The concepts of smart meters and Flexible Pricing were then introduced and initial feelings on the AMI Program were sought from participants, including their thoughts on the reasons for the program. The costs of peak demand were discussed; in most sessions this involved drawing a graph to explain peak times. In general, peak demand was described to the group as ‘inefficient’ and participants were asked to indicate whether they could identify any benefits from spreading their consumption over the day.

Flexible Pricing was described as off peak charges that are lower than current flat tariffs and peak charges that are higher than current flat tariffs. Participants were asked for their initial impressions of such pricing structures, including whether any problems spring to mind and whether it would change their household’s behaviour. The groups were asked to identify ways they could shift their energy use away from peak times and at what price differential (peak to off peak) would give them an incentive to change their energy consumption.

More detailed aspects of Flexible Pricing were then introduced, including seasonal tariff variations, critical peak pricing incentives and direct load control of air conditioning. In-home displays were also discussed in the context of ways to better understand energy use. Participants were asked to indicate whether they would be willing to pay for such devices.

To finish each session, participants were asked to list the elements of the new pricing structures that appealed most to them and those that most worried them. In some sessions, participants were asked to describe the obligations that electricity suppliers and government have in presenting such new ideas to consumers, including their view on how best to communicate complex ideas to the majority of people.

5.4 Key insights from discussions

The six focus group sessions involved largely positive discussions and attitudes to energy use and bills, revealing some useful insights into the level of understanding of energy in the community. The groups typically exhibited a low level of understanding of their household energy use, particularly in terms of which appliances contribute most to their bills. However, in each session there was at least one well informed consumer who had participated in the Energy Saver Incentive scheme and was well aware of their energy use and the contribution of different appliances.

On Flexible Pricing and related issues, participants exhibited attitudes ranging from excitement at the potential for them to learn more about their energy use and lower their bills, to defeatist ‘there’s nothing I can do, prices always seem to go up anyway’ views. In general, once the concept of peak demand and its associated costs were explained to each group, participants seemed to view the new pricing arrangements as ‘fair’ but in some cases ‘difficult’.

Peak demand and its costs proved a difficult concept for the groups to understand, with many participants questioning why their own peak and off peak tariff times could not be better structured around their particular needs. The use of analogies such as peak hour road traffic assisted the explanation, however, the challenge of explaining the reasons for the costs of peak demand without needing to explain how energy is produced and transported should not be underestimated. While constraints in water supply can be visualised by the majority of the population, the drivers of costs in energy supply are not well understood. A common response to the idea of incentives to shift people’s energy use was that this would create the same network problems and costs, just at different times of the day.

Interestingly, many participants were more compelled by the idea that elderly or sick people might suffer during brown outs unless the general population became more aware of the impacts of their peak demand. Viewed in the context of the success of water saving campaigns in recent years, such community-aware attitudes could be important considerations in the context of customer engagement in Flexible Pricing.

Other key learnings from the sessions include:

Household appliance use

• Many vulnerable households have high numbers of laptops and televisions used primarily during peak times, which were the first appliances that participants mentioned in relation to their energy use. While these appliances do not use a large proportion of household energy, this is not always understood by participants, and participants generally considered use of these discretionary items would be the first consumption they would shift in response to Flexible Pricing. More information on what drives energy bills would clearly assist people to respond to new ToU pricing signals.
• Many participants across the groups expressed that due to their high bills, they were in a cycle or ‘trap’ of not being able to afford new, more efficient appliances which would lower their bills.
• Participants from households with young children reported that they use electric clothes dryers regularly and acknowledged that these appliances were having a large impact on their energy bills.
• Air conditioners were not ‘front of mind’ when people reported their energy use, although most participants indicated that they restricted their use of air conditioning to extreme heat in order to reduce their energy bills.
• In many cases, participants suggested they could shift their energy use by charging laptops, phones, portable media players, etc. overnight, instead of during the day. Others indicated they were already charging such items overnight.

Attitudes to energy bills

• Participants exhibited a high degree of distrust of energy retailers and distributors and consistently stated that the government should provide more information to assist households in managing their energy use. Corporate greed and the impact of privatising Victorian electricity assets were repeatedly cited as reasons for rising electricity costs.
• Bills were consistently described as ‘high’, however, when put in the context of a ‘daily cost to run a household’ (being approximately $3 per day), participants agreed that energy is actually relatively cheap.
• Estimated bills (where a meter cannot be read for one period and the corresponding bill is estimated based on previous usage) were viewed as problematic, causing financial difficulty where usage is significantly over or under-estimated.
• Several people expressed concerns that they feel they have no control over their energy bills and that they could not see a connection between their behaviour and their bills.
• Some discussion around the differences between owner-occupiers and renters raised the notion that renters have fewer opportunities to manage their energy use and bills due to their inability to invest in gas heating, insulation, awnings or solar panels.

Attitudes to Flexible Pricing (ToU)

• As the notion of lower off peak charges was introduced before the notion of higher peak prices, participants’ initial response to ToU tariffs was positive. A common response among older group members was that the differential between peak and off peak would need to be more than 50% to encourage them to shift their consumption.
• There was a degree of scepticism expressed around the notion that the new tariffs would need to be structured such that overall retail and distribution revenue was no higher as a result.
• Many participants recommended that the new tariffs be determined according to each individual household’s daily schedule, rather than generic timeframes for peak and off peak. The notion that each household has a different schedule was a common issue discussed.
• Despite this, simplicity in the tariff structure was viewed as important. Participants continually stated that they needed to be able to understand their contract without too much work or reading in order to agree to sign up for Flexible Pricing. Seasonal tariff variations were unpopular, in favour of simple, easy to remember peak and off peak periods.
• Several participants across the groups raised concerns that enabling their appliances to operate overnight unsupervised (such as using a timer for a washing machine or dishwasher) ran the significant risk of flooding and even house fires.

Attitudes to Critical Peak Pricing, Direct Load Control and In Home Displays

• Critical Peak Pricing, where tariffs are set at 5 times the peak rate for a few hours per year, was generally an unpopular notion. Participants were concerned about how they would receive messages from energy retailers and distributors and that at the time of the CPP event they would forget.
• However, Critical Peak Incentives or Rebates, where customers would receive a discount or rebate on their bill for using less power during the Critical Peak time, appeared to be attractive to most participants. A range of rebate levels were discussed, with some participants prepared to respond for a rebate of two days’ of their energy use, others requiring more than $40 to respond. The idea of going to a shopping centre or movie during the peak event was discussed, as participants seemed to consider that obtaining a rebate equivalent to the amount they might spend in the two hour period was reasonable.
• Direct load control of air conditioning seemed to be unpopular amongst most participants, associated with the idea of ‘big brother’ controlling appliances. There was a general belief among the groups that cycling an air conditioner’s compressor on and off could not be carried out without feeling some effects in the household. Participants were also concerned that this might damage or shorten the life of their appliances.
• Many participants responded positively to the notion of In Home Displays, immediately seeing value in the information that it could provide to assist their energy management. Around half of the participants expressed a willingness to pay a one-off fee of between $50 and $100 for an In Home Display (payment plans or rental charges were unpopular).

Responsibility for information

• Most participants were unaware of the rationale for installing smart meters and expressed annoyance at the lack of information distributed to them about the AMI Program. Several participants said that the smart metering brochures distributed at the time of installation were inadequate and confusing.
• Participants consistently expressed the need for information regarding household energy use, particularly which appliances contribute most to energy consumption.
• In general, government was viewed as the logical, appropriate primary source for this information (noting the scepticism of energy companies discussed above). Participants suggested that brochures or bill inserts were the best ways for this information to be distributed.
• A few participants indicated that they regularly compare their bills with friends and neighbours. The notion of household comparators (similar to the Target 155 water information) was popular among some groups.

The following table presents an outline of the general attitudes to Flexible Pricing that was expressed by differing vulnerable groups, where they differed from the views discussed above.

Table 19: Summary of attitudes by vulnerable group

Vulnerable group	Attitudes expressed during discussions
Elderly or aged pensioners	Discussions revealed that the older participants seem to have lower energy bills than other groups. Perhaps due to this, they appeared less interested in or less able to change their lifestyles to reduce their energy bills. Adult children and other family members visiting often was a concern, driving higher bills in certain periods. Ways to respond to Flexible Pricing that were mentioned included using slow cookers to cook main meals (spreading the power use over the day), using a thermos instead of boiling the kettle frequently, and cooking a midday or evening meal during off peak times.
Rural Victorians	Groups outside of Melbourne seemed to have relatively high bills. Most people use electric clothes dryers; birds and weather conditions were cited as common reasons for not hanging clothes outside. Participants who are at home during the day were more open to the idea of Flexible Pricing; they consider they have an ability to change their daytime energy use.
Single parents	Participants consistently reported difficulties in communicating the need to save energy to their children, particularly parents of teenage or dependant adult children. There was a general feeling that the bill payers did not have a high degree of control over the household energy use, as they were working outside of the home for most of the time.
Disability pensioners/carers	General view expressed that they have a limited ability to shift their energy use due to special needs of disabled persons. Air conditioning was viewed as a necessity.
Low income / low wealth	Several participants indicted they are currently on a payment plan for electricity and/or gas which involves a monthly, pre-determined bill. At least five participants stated that they have live-in boarders (generally international students) paying rent that includes fixed energy bills and who are difficult to communicate with. They expressed concern as to how they would communicate Flexible Pricing to these household members.

5.5 Conclusion – focus group sessions

Our findings from the focus group sessions are summarised as:

• There is a lack of understanding among vulnerable customers as to what drives household electricity bills
• Vulnerable customers generally consider that it is the government’s responsibility to provide more information about household energy use and future Flexible Pricing structures, particularly as they are sceptical of energy companies
• When it is explained to them in detail and justified on the basis of improving efficiency in supply costs, vulnerable customers generally understand that Flexible Pricing may present opportunities for them to lower their electricity bills, if they can respond to the new pricing signals
• Simplicity in tariff structure, contracts and any information is considered very important in getting vulnerable customers to consider Flexible Pricing
• Critical Peak Incentives or Rebates are viewed more favourably than Critical Peak Pricing
• Direct load control of air conditioning is unpopular
• In Home Displays are generally considered useful, with around half the participants indicating they would be willing to pay a one off fee of between $50 and $100 to receive one.

We note that these findings are based only on the attitudes of the 42 participants of the focus group sessions, as at the time they were interviewed. The sessions were particularly focussed on views of the individuals at that time in their lives and not their views as they might change over time as their homes and lives change. A wide range of both external and inherent factors affect people’s views on energy bills and Flexible Pricing, for example, whether people have young children, or have a disability.

While our qualitative analysis makes it possible to identify common themes and potential impacts among similar people, it was not possible in the time period for our study to determine the precise impact of each factor that impacts household energy bills and attitudes to energy use. Instead, the value of this qualitative analysis is in the insights it presents for government as to the policies surrounding the introduction of Flexible Pricing and the information that should be provided to support it.

6 Price sensitivity analysis

6.1 Overview of respondents

A total of 3,003 respondents participated in a telephone survey we conducted to quantify consumer sensitivity to bill increases. Questions in the survey such as income band, age band, household situation and location enabled respondents to be categorised either into one of the vulnerable groups (developed in Stage 1 of the study), or as a“non-vulnerable” group of customers.

Many respondents were found to satisfy more than one vulnerable category. For example, a respondent may have been over 65 years of age and earning less than $45,000 a year, categorising them as both ‘Elderly’ and ‘Low Income’. In such instances, the data from respondents was used in the analysis of both vulnerable groups. The table below shows the number of respondents classified into each category.

Table 20: Number of telephone survey respondents by group

Group	Number of respondents**	Percentage of respondents***
All Residential*	3,003	--
Non-vulnerable	627	10%
Elderly	682	11%
Needs Assistance	277	5%
Single Income	1033	17%
Low Income	1001	16%
Regional	875	14%
Low Net Worth****	422	7%
Health Care Card Holder	964	16%
Single Parent	231	4%
* The ‘All Residential’ group refers to all the respondents to the telephone survey and therefore there is no applicable value for ‘percentage of respondents’ ** As respondents met the criteria for more than one vulnerable group, their data was used to analyse each group they belonged to, resulting in the actual number of records being used for analysis greater than the 3,003 respondents participating in the survey. *** The percentage shown is based on the total number of records (6,112) created by the multiple assignment of respondents to vulnerable groups. **** Due to the difficulty of including a question that would ascertain net worth of respondents, the postcode of the respondent was asked and matched to postcodes in the Stage 1 billing data where the average net worth for that postcode had been previously determined from other data sources.

We note that the percentage of respondents in each vulnerable group does not reflect the proportion of these groups in the Victorian population. However, these groups are being analysed separately and not aggregated to provide an overall population result. Therefore the representativeness of the groups in the general population is not material to our results. However, the number of respondents for the ‘Needs assistance’ and ‘Single parents’ groups is short of an ideal target (being 385 respondents, based on standard market research benchmarks). The sample for these individual groups is therefore not strongly statistically representative, but large enough for their results to be indicative of the sentiment and perception of these vulnerable groups.

To understand how electricity was used by respondents, questions were asked about the use of electricity for heating. As illustrated in Figure 11, less than a quarter of each vulnerable group indicated that they use electricity for heating purposes, which is consistent with the Victorian average penetration of reticulated gas²⁵. As demonstrated in Figure 12, a very small percentage of respondents use electric-boosted solar hot water. For both questions, respondents in the Regional group reported higher use of electricity. Use of electricity is higher in regional Victoria where there is generally less access to natural gas.

Figure 11: Percentage of each group using electricity for heating

Figure 12: Percentage of each group using electricity for electric-boosted solar hot water

6.2 Results

6.2.1 Van Westendorp study

Using the methodology described in Section 2.5, “Quantitative research of sensitivity to ToU pricing”, Van Westendorp price sensitivity meter analysis was conducted for each group. Using each group’s Van Westendorp diagram (see Appendix B), the sensitivity score for that group was calculated. Figure 13 shows the scores for each group, where the larger the score, the more sensitive the group is to increases in electricity bill. For example, a sensitivity score of 0.5 means that for a 1% increase in electricity bills, the percentage of respondents indicating this to be affordable decreases by 0.5%.

A larger sensitivity score means that affordability decreases more rapidly as the electricity bill increases. We note that the relationship between bill increase and decrease in customers is one-way, that is, it specifically relates to the situation where bills are already at a threshold that would cause concern for households should the electricity bill increase further. It does not indicate the possible increase in affordability amongst households if electricity bills were to be decreased.

Figure 13: Van Westendorp sensitivity scores to bill increases for each group

The above diagram illustrates that households classified as Regional or Single Parent are 70% to 100% more sensitive (i.e. about twice as sensitive) to increases in electricity bills as ‘Non-vulnerable’ households. Overall, all vulnerable groups except those classified as having Health Care Cards indicated they were less able to afford increases in electricity bills than ‘Non-vulnerable’ group respondents. However, the difference between Health Care Card households and Non-vulnerable households is not significant.

6.2.2 Ease of changing electricity consumption

Heating and cooling the home are the main uses of energy that households have direct control over. The quantitative survey included questions on how easily respondents would be able to shift the time they used heating and cooling appliances in their home from peak times to other times, as well as the ease with which they could similarly shift consumption for other uses. A question was also posed on the ease with which households could reduce electricity consumption in peak times. Respondents answered these questions on a 1 to 5 rating scale, with 1 representing that it would be easy to shift or reduce consumption, and 5 representing that it would be very hard or impossible to shift or reduce consumption.

Figure 14 presents the average response for each group on how easily they could shift electricity consumption for heating and cooling uses from peak times. With average values around 4 on the 1 to 5 rating scale, most respondents indicated that it would be difficult to shift electricity, but not hard or impossible (which would have been reflected if the average response was closer to 5). Interestingly, the vulnerable customer groups did not indicate it was any more difficult for them to shift load than the broader non-vulnerable group.

Figure 14: Average rating for ‘ease of shifting heating and cooling use’ by group

Figure 15 presents the average response for each group on how easily they could shift electricity consumption for uses other than for heating and cooling from peak times. With average values around 3 on the 1 to 5 rating scale, most respondents indicated that to shift electricity for other uses would be neither easy nor hard. This included respondents in the ‘Non-vulnerable’ group, which reported a level of difficulty not significantly different from all other groups.

Figure 15: Average rating for ‘ease of shifting other use’ by group

Finally, Figure 16 presents the average response for each group on how easily they could reduce electricity consumption during peak times. With average values around 3.5 on the 1 to 5 rating scale, most respondents indicated that they would face some difficulty in reducing electricity use during peak times. This included respondents in the ‘Non-vulnerable’ group, which reporting a level of difficulty not significantly different from all other groups.

Figure 16: Average rating for ‘reducing electricity use during peak times’ by group

6.3 Summary

Overall, all vulnerable groups (except Health Care Card Holders) identified in Stage 1 indicated that they would be less able to afford increases in electricity bills than ‘Non-vulnerable’ respondents, with Regional and Single Parent households reporting 70% to 100% greater sensitivity than the average ‘Non-vulnerable’ household. Health Care Card holders did not indicate significantly different responses in affordability of electricity bill increases as ‘Non-vulnerable’ households.

All respondents, including those in the ‘Non-vulnerable’ group, indicated it would be difficult to shift electricity for heating and cooling purposes from peak times. All respondents also indicated it would be difficult to cut down electricity usage during peak times, but not as difficult as shifting that usage to another time. Respondents generally indicated that shifting their electricity for uses other than heating and cooling would be neither easy nor difficult, rather somewhere in between.

7 Case studies

7.1 Overview of approach

Six case studies have been developed to draw together the analysis undertaken as part of this study and illustrate the options and decisions likely to be faced by customers following the introduction of Flexible Pricing. These case studies are designed to represent the general attitudes of vulnerable customers to energy use, Flexible Pricing and associated issues. They represent the overarching findings of our research into how vulnerable consumers might be affected by and respond to Flexible Pricing.

Given our findings that there is a wide variation in the distribution of impacts of Flexible Pricing on customer bills, it is important to note however that the case studies are examples of potential attitudes and responses rather than the average customer response.

We have relied on quotes obtained during the Focus Group Sessions discussed above to illustrate ‘real world’ scenarios; however, we note that most of the case study personas are based on characteristics revealed by several similar participants across the groups held. As such, the case studies are best described as ‘archetypes’ of real people and do not accurately reflect any single individuals.

7.2 Case studies developed

The following table lists the case studies that have been developed based on the discussions within the focus groups, results from the telephone survey, and insights derived from analysing the billing data and ABS HES data.

Table 21: List of case studies

	Description of Case Study
1	Single parent with young / teenage children, working but with a low income, renter (possibly in government housing). Concerned about bills, already taking action to lower consumption.
2	Elderly person living alone in their own home, receiving an old age pension. Old appliances. Already working hard to save energy and money, low or no capital to invest in solar panels or other energy saving items or new appliances.
3	Person who needs assistance, taken this is from the perspective of a disabled person. Requires constant air conditioning for health reasons.
4	Single person with a low income (under-employed or unemployed), renting, no capital to buy new appliances, but very high bills and little control over household energy use.
5	Average family unit, relatively low income (single income family), with two children. Struggling with energy bills, on a payment plan.
6	Aged concession holder with high net wealth but low disposable income, supporting adult children or with a boarder living in the house (international student)

7.2.1 Case Study 1 – "Jenny"

Single parent with young and teenage children, working but with a low income. Renting (possibly in government housing). Concerned about bills, already taking action to lower consumption.

"Jenny"

Jenny is a single mother living in Shepparton with her two children (one young, one teenager). She works for a building company so has some good insights into energy saving features of houses. Jenny is a low income earner and holds a Health Care Card which entitles her to a discount on her annual electricity bills of 17.5%.²⁶

Jenny has a range of appliances in her home including gas heating and gas cooking appliances. Some of Jenny’s appliances are fairly old.

“I can’t afford new appliances. My fridge and washing machine are 16 years old and my freezer is 25 years old.”

Jenny is a well-informed energy consumer, she regularly calls EnergyWatch to find out the best available rates, and if her electricity rates are higher than the market prices, she pushes the retailer to match the best price or she’ll switch. She says this method is always successful, even when arguing against annual CPI adjustments to prices.

Jenny has participated in the Victorian Energy Efficiency Target (VEET) scheme. An independent operator recently came around to install Standby Power Controllers in her home. While she thought the installer should have given her more information about how to use the device, she thinks it has helped her to lower her bills. Jenny also installed new low energy light globes, which she says helped bring her bills down.

Jenny likes the idea of time-of-use tariffs. She understands that peak electricity costs a lot of money to supply, and is worried about the impact of brown-outs and black-outs, particularly for elderly and sick people who need heating and cooling to survive. Jenny thinks that she’ll be able to shift some of her consumption outside of peak times and even save money, thanks to lower off-peak rates.

In particular, Jenny thinks she’ll be able to cook outside of peak times using a slow-cooker. Jenny could also shift her use of the washing machine to off-peak times. Jenny doesn’t have a dishwasher or a dryer, and rarely uses her air conditioner, even on hot days.

She also likes the idea of Critical Peak Pricing Rebates, where she’d receive advanced notice of a critical peak event, allowing her to plan to shift or reduce her electricity for a few hours to receive a discount on her monthly bills. Jenny says “I could chuck my Ugg boots on and have the doona on for a day, as long as it wasn’t too many days.”

She also said: “I could take the kids to the pool for a few days in the year, when it’s really hot.”

Jenny is a bit worried about how she’ll teach her teenage son to shift his energy use. “He’s always on the laptop or watching TV. Sometimes I wake up in the middle of the night and go around turning the TV and computer off. He doesn’t understand that energy is precious and costs money.”

Jenny thinks that an In Home Display could be useful for her in teaching her son about the cost of electricity at peak times. She would be willing to pay a one off fee of around $50 to receive an In Home Display.

Overall, if moved onto a time-of-use tariff, Jenny would have an opportunity to lower her electricity bills by moving some of her consumption into off peak times, and participating in critical peak pricing rebates.

7.2.2 Case Study 2 – "Barbara"

Elderly person living alone in their own home, receiving an old age pension. Old appliances. Already working hard to save energy and money, low or no capital to invest in solar panels or other energy saving items.

"Barbara"

Barbara is an elderly lady who lives alone in her two-bedroom home within a retirement village in Melbourne. Barbara holds a Health Care Card and a Pensioner Card which entitles her to a 17.5% discount on her electricity bills.

She has a gas hot water system and heating, however has electric cooking appliances. She uses a radio most days and watches some TV programs too. She has a fridge and a chest freezer in the garage. Barbara doesn’t use an air conditioner.

Barbara is very careful with her energy use so she can save money: she always turns off appliances at the wall and is very strict with lights. She only turns on the lights in her living room late in the evening, preferring to leave the curtains open and use the streetlights when she’s watching TV.

Barbara usually cooks dinner at 5:30pm, she always cooks more than one meal so she can re-heat meals and save cooking time. If using the oven, Barbara will always cook a few things at once to save energy.

Barbara is excited about the idea of time-of-use pricing. She thinks she could save money by shifting her cooking and washing times to off-peak.

Although she is quite excited about Critical Peak Pricing Rebates, she is not sure how she’ll get the day-ahead messages, as Barbara doesn’t have a computer or a mobile phone. She does like the idea of being able to ‘earn refunds’ on her bills by shifting some consumption.

Barbara says “I’d take it. Everyone’s going to take a discount”

As she doesn’t have an air conditioner, Barbara doesn’t use as much energy at critical peak times as other people. To be rewarded for using even less, she’d need to have reasonable targets set on her consumption, based on her actual usage, not an average household.

Barbara is not sure she would use an In Home Display, however would like to know more about how much energy her appliances are actually using. She thinks that the Government should give people more information about how to save energy and money, for example, she would like to know how to use a timer to control her appliances.

7.2.3 Case Study 3 – "Tony"

Person who needs assistance, this is taken from the perspective of a disabled person. Requires air conditioning for health reasons.

"Tony"

Tony is in his mid-40s and lives alone in Government housing in the La Trobe Valley. Tony has a disability which means he finds it more difficult than most people to maintain an even body temperature, particularly in extreme heat or cold weather. Tony holds a Pensioner Card, a Health Care Card and receives a Medical Cooling Concession for his disability, which means that he is eligible for a 17.5% discount on his electricity bills, plus an additional 17.5% discount during the warmer months (1 November to 30 April). ²⁷

Tony has a range of electric appliances, including a washing machine, dryer, kettle, toaster, microwave and a TV. Tony has gas heating, hot water and cooking facilities in his home.

Tony says ‘Well I’m on a disability pension, my health is really bad. I’m qualified to have an air conditioner concession, but I can’t afford to run it much anyway. It’s just a case of not having enough money, with all my other expenses, so when it’s hot I go to the shopping centre instead. It’s cooler there.’

Tony is on a payment plan with his electricity company and pays around $80 a month for gas and electricity, after his concessions. Tony is really worried about electricity prices going up. He says:

"If the price is higher, the money’s got to come from somewhere and the only place I could get extra money is taking it away from food. And I’m lucky if I buy $50 worth of food a week as it is."

Tony uses his clothes dryer around four times a week. He doesn’t hang his clothes outside because the birds make a mess of them.

When told that his clothes dryer uses a lot of energy, Tony is surprised. He is also surprised that his kettle uses a lot of energy, because he often keeps it full of hot water so he doesn’t have to wait long to heat it up again. When told that he should just boil water for one cup at a time or even use the microwave to make a cup of tea to save energy, Tony is interested to find out more.

"The government should probably send out stuff that explains a little bit about it so we can go, ‘okay, I won’t put so much water in the kettle.’ Maybe I would hang the clothes outside a bit more often, or get one of those clothes hangers for inside."

Even though he thinks that he probably could use less energy at peak times, Tony is worried about Flexible Pricing. He’s not sure he’ll be able to understand enough about it and will end up paying more.

Tony likes the idea of getting a rebate for turning things off at critical peak times, but doesn’t like the idea of getting a critical peak price. On the critical peak price, he says:

"If I was really sick and I couldn’t go outside all day and I had to have air con on, that critical peak would annoy me because I would still have to pay the high price, even though I had no choice."

7.2.4 Case Study 4 – "Malcolm"

Single person with a low income (under-employed or unemployed), renting, no capital to buy new appliances, but very high bills and little control over household energy use.

"Malcolm"

Malcolm lives in Dandenong in a large share-house with his 18 year old daughter and three other housemates. Malcolm is underemployed, working only 8 hours a week, though he is hoping to work more hours soon. Malcolm holds a Health Care Card, and because the electricity bill is in his name, the household receives a 17.5% discount on their electricity bills.

Malcolm’s household has a wide range of appliances, including two fridges (one old, one newer), five TVs and five computers which are used every night. He also has a waterbed which he heats every day and an air conditioner used sometimes in summer. Malcolm has a gas stove and gas hot water, but has electric heating.

Malcolm’s household has very high electricity bills, the biggest he can remember was $1200 for one quarter and they are typically around $700 a quarter.

"We use a lot of electricity because we have a lot of computers in the house, so computers and the evil dryer, which is a necessity these days. I have to wash my daughter’s uniform every day. I could dry it outside I guess, but it’s been raining so much that you don’t really have a choice. Right now, I can’t afford to buy her another uniform."

Despite being home more often than his other housemates, Malcolm doesn’t have a lot of control over his household’s electricity use. His housemates often have their TVs, computers and blow heaters going in their rooms all night.

Malcolm has recently participated in the Victorian Energy Saver Incentive scheme, having a tradesman come around to install a Standby Power Controller in his home. A Standby Power Controller is plugged into a number of appliances (typically home entertainment equipment or computers) and automatically switches them off after a period of non-use.

"They came around with these switches, depending on where the plug is, there is one that can have continuous power and then which controls standby, after a certain amount of time it just shuts off."

Malcolm is yet to see the results of this controller on his electricity bills, but is determined to work to try and save money by using less energy.

Malcolm knows that electricity prices have been rising because of the costs of generation and running the network are going up.

"I have seen them working on the power lines."

When Flexible Pricing is explained to Malcolm, he understands that if people can shift their energy use to off peak times, it will cost less to maintain the whole system. Malcolm would turn his air conditioner off during peak times, if he could get a discount on his energy bills in return.

“Well, who doesn’t like the idea of having cheaper anything? Most of my stuff I don’t use during the day, only first thing in the morning with the kettle, toaster, but everything else is off. At night, when we all come back, that’s the late part of the peak time when I’m going to be getting hit. And we can’t shift heating use when it’s cold, that’s just not possible.”

Malcolm knows that his daughter would not be able to stop using her laptop at night during peak times.

“She gets her homework emailed to her, so she has to do it on the computer.”

Malcolm says that he could make sure all his laptops, phones, ipods etc. were charged during off-peak times, to try and take advantage of cheaper energy and lower peak use.

Malcolm likes the idea of using timers to control his appliances, though he’s less sure about his washing machine:

“Doing the washing late at night would be difficult. I can’t use a timer on my machine, because at the moment it doesn’t work properly, so we have a problem with the water. I can’t have it running when I’m not there because if I do, it will flood.”

"I can’t afford a new washing machine because at the moment I am only working 8 hours a week. I am barely paying bills as it is, without doing maintenance and buying new appliances. When I am working fully again, it’s not a problem. It’s a small thing that I can manage now by turning the tap on and off."

7.2.5 Case Study 5 – “Sarah”

Average family unit, relatively low income (single income family), with two children. Struggling with energy bills, on a payment plan.

"Sarah"

Sarah is a full time student and mother of two young children (aged two and three). Her partner works as a paramedic and together they rent a home in Melbourne’s outer west.

Sarah’s home has air conditioning and ducted gas heating and cooking facilities. When asked what appliances use most of the electricity, she said:

"I have a lot of computers! And the kids insist on having a light on during the night, so that chews it up. I have a big plasma TV and a PlayStation as well. We also have a fish tank and fridges that probably need replacing. Also, with two young kids, we use the old clothes dryer incessantly!"

"My computer is almost on 24/7 and then the PlayStation is usually on at the same time, and we have three TVs in the house. It’s out of control now, it’s ridiculous."

"When we had our light bulbs changed over to energy efficient ones, I found out that we have 109 lights in the house. 109! It’s mostly down lights."

Sarah thinks her family’s energy bills are too high, with electricity being on average $420 a quarter and gas being around $280 per quarter during summer and $400 per quarter during winter.

“I am on payment plans because it is just insane. I try to reduce electricity, but as hard as I try, it’s just chewing through it, so I think I need to just replace the old appliances and get new ones.”

Sarah is concerned about the rising cost of electricity.

“I feel I am stuck with my provider because I am on a payment plan so every fortnight I have to pay a certain amount regardless. I was on the phone only a few weeks ago, asking if there are any cheaper rate plans that I could go on, and they basically said that unless you actually pay your bill on time you are not going to get any discount. I am kind of stuck because I constantly owe this company money.”

”Five years ago, we were better off, and the electricity bill didn’t cause a heart attack, but now you just dread it when you get that bill.”
Sarah is trying to use the computers less to save energy, but says that her husband often has their PlayStation on for five or six hours when he is home from work.

“I tell him to stop using it, to turn off the lights, but you feel like a nag!”

Sarah likes the idea of finding new ways to try to lower her household’s energy bills using Flexible Pricing. She says that if she had the opportunity, she would sometimes stay up at night to wait for the off peak rates, and run her dishwasher and maybe do the washing as well.

"I don’t always want to stay up til midnight, though I can put them on a time delay."

When asked if she would participate in a critical peak rebate scheme over summer, where she might receive $20 for turning off her air conditioner for a few hours, Sarah says it depends on the amount the electricity company would pay:

"So, stay hot and sticky for $20? I don’t know, how much do you really want that rebate? It comes down to personal comfort."

Sarah likes the idea of an In Home Display as she doesn’t feel she has a good handle on what appliances are using the most electricity. She thinks that the government should provide information to households on their energy use and what is driving their bills.

"I want to see how everything is energy rated. I would pay $50 for a device that could do that."

7.2.6 Case Study 6 – "Andrew"

Aged concession holder with high net wealth but low disposable income, supporting adult children and with a boarder living in the house (international student).

“Andrew”

Andrew is retired and lives with his wife and adult son in Melbourne, in a low-energy using house that he built six years ago. A young student from a developing country lives in a small flat underneath Andrew’s house and pays him rent which incorporates his energy bills. Due to his low income, Andrew has a Pensioner Card and therefore qualifies for a 17.5% rebate on all his electricity bills.

“We have the basic electricity for lighting and TV and that sort of thing. We have gas appliances for cooking, both top of stove and oven are gas, heating is gas and the hot water service is gas. We have an evaporative cooling system out on the roof for keeping cool in the summer.”

Andrew’s household electricity bills are around $300 per quarter, which he thinks is surprisingly high, considering the features of the house.

“The house is very well insulated, it works extremely well. About a year ago we got solar panels, only a 1.5 kilowatt system. I am investing in our old age.”

“I have just got my first bill which relates to the solar, and a quarter’s use was about $300. I don’t understand how it can be so high. I can’t put a finger on that particularly because I don’t work, the main power I use is probably the oven and I don’t cook in that very often.”

Andrew’s son runs a part-time mechanic shop in their garage and uses a number of power tools and machinery; however Andrew doesn’t know how much energy they use. Andrew also does not know how much energy his student boarder is using. The student boarder has his own TV, computer, lighting and electric hot plates for cooking.

”It’s very hard to communicate with him. He just doesn’t understand the concept of saving energy. He didn’t have electricity in the village where he grew up. He often leaves the lights on when he goes out.”

When asked whether he would respond to Flexible Pricing, Andrew wasn’t sure.

“Whether I would respond depends on the relative price increase and decrease and peak and off peak times.”

Andrew does like the idea of receiving critical peak rebates for turning energy use down on extremely hot days.

“Between $5 and $10 is to me, a measurable benefit for a day because if you look at your consumption, that’s most probably what it is per day. I would respond to that. But the price should vary according to how much you usually use. You don’t want to punish the lower users because they can’t shift their energy use because it’s already low.”

When asked whether he would participate in Direct Load Control, where the electricity company puts a device on his air conditioner to cycle it on and off less frequently during times of extreme peak demand, Andrew said no.

“I don’t want anyone controlling my air conditioner. I want to decide if I will turn it on or off. I think it would damage my air conditioner.”

Andrew was very interested in the concept of In Home Displays. He wants to understand more about his household energy use and would be willing to pay up to $100 for a device to help him. He thinks an In Home Display would help him communicate with the student boarder and his son, teaching them how to save money on electricity bills.

7.3 Conclusion – case studies

The case studies presented above reflect the ideas and attitudes expressed by vulnerable customers during the focus group sessions, as well as generally reflecting our findings from the other quantitative analysis of the impacts of Flexible Pricing. The case studies demonstrate that the individual effects of Flexible Pricing are likely to be highly varied and dependant on a wide range of external and inherent factors, in particular the number, age and occupation of household members, the stock of appliances and how they are used and core features of the house, such as insulation or solar panels.

Consistent themes expressed within these case studies (and within the other analysis conducted in Stage 2) include:

• A lack of understanding of what drives household electricity bills and, in many cases, difficulty for bill payers in controlling the household electricity use or communicating with other household members on the cost of using electricity
• Belief that the government needs to provide more information on energy use and needs to provide information to assist households in considering Flexible Pricing offers
• A general willingness to investigate and even try Flexible Pricing once the concepts are explained in detail, along with the reasons for the price incentives.

8 Conclusions and recommendations

8.1 Summary of Stage 2 analysis results

Stage 2 of the Customer Impact Study involved updating and expanding the analysis of Flexible Pricing scenarios and conducting three new streams of research, building on the findings of Stage 1 and seeking to validate the initial results. The conclusions reached in this report are summarised below.

8.1.1 Flexible Pricing scenarios and distributional effects

Incorporating new Flexible Pricing tariff structures developed through consultation between DPI and industry, we used the scenario tool developed as part of Stage 1 to identify the impact of various new pricing scenarios on the vulnerable groups. We also identified the distribution of the annual bill impacts across each tariff scenario and vulnerable group.

The following conclusions can be drawn from this analysis:

• All pricing scenarios tested resulted in average bill changes of between -6.1% and 1.8%.
• However, the distribution of the impact is substantially greater than the mean effect. Across all groups (“All residential”), 50% of the population could experience either a reduction in bills of greater than 24.3%, or an increase in bills of more than 12.0%. This result indicates the high degree of variability in energy use, even among customers of similar vulnerability characteristics.
• While some customers may be worse off under Flexible Pricing, the average annual bill increase typically does not exceed 1.8%, even assuming there is a zero elasticity effect and holding RB revenue constant. In most cases, a decrease in the average annual bill occurs, particularly when allowing for elasticity.
• Within the relatively narrow set of tariffs we examined, the exact design of the Flexible Pricing scenarios made little difference to the overall distribution of the price impact. Varying peak, shoulder and off peak times by one or two hours (that is, starting peak at 4pm instead of 5pm) does not have a substantial impact on the average vulnerable customers’ bills.
• There is a very broad range of effect for each group which indicates that belonging to a particular group of itself is not the most important factor in determining bill impact; rather, there are a range of factors that contribute to vulnerability in combination with general demographics.
• Without applying an elasticity effect (that is, assuming customers do not change their behaviour in response to Flexible Pricing), single element meter customers are likely to experience a greater increase in bills (or, in most cases, a smaller decrease in bills) than dual element meter customers under most tariff scenarios.
• However, when applying an elasticity effect and allowing total revenue to change (fall), on average the vast majority of groups will be better off.
• Low income households, people requiring disability support, health care card holders and regional groups are likely to be, on average, better off under each tariff scenario than under the base case.
• Low net worth households and single income households are, on average, worse off (or in most cases, less better off) under Flexible Pricing than the other vulnerable groups.
• As modelled, adding a CPP does not make a significant difference to the impact of ToU on each customer group, aside from low net worth households for which it has a small negative effect.
• Having wider peak times (that is, 6 hours of peak on each weekday instead of 4 hours) and having weekends at off peak rates will result in the smallest changes to customers’ annual bills.
• Analysis of the impacts of Flexible Pricing scenario 2F across quarterly bills highlights that summer bills are likely to be most negatively affected (i.e. summer bills will increase by more or decrease by less in response to Flexible Pricing than other quarters). Spring and autumn bills are likely to be affected to a similar extent, with the exception of the elderly group whose autumn bills are more negatively affected than spring bills. Winter bill changes are quite varied across all groups and dual/single meter customers.

8.1.2 Confirming vulnerability

Through analysis of ABS data developed through the HES, we were able to confirm that the vulnerable groups identified in Stage 1 are indeed among those Victorians spending the greatest proportion of their income on electricity use (being more than 3% of total household income).

This result gives confidence that we are analysing the energy use data of the most vulnerable people in Victoria and that our findings are relevant to the objective of the Customer Impact Study, being to understand the impact of ToU tariffs on vulnerable customers.

8.1.3 Focus group sessions

Over two weeks in April 2012, we conducted six focus group sessions of two hours each in four locations across Melbourne and regional Victoria, gathering the views of 42 participants on Flexible Pricing and energy affordability issues.

The groups typically exhibited a low level of understanding of their household energy use, particularly in terms of which appliances contribute most to their bills.

Many participants said that they have difficulty understanding the connection between their own appliance use and their electricity bills. Participants consistently expressed the view that government (rather than energy companies) needs to provide them with more information on how to save money on their energy bills.

On Flexible Pricing and related issues, participants exhibited attitudes ranging from excitement at the potential for them to learn more about their energy use and lower their bills, to defeatist ‘there’s nothing I can do, prices always seem to go up anyway’ views. In general, once the concept of peak demand and its associated costs were explained to each group (which proved difficult to understand for many people), participants seemed to view the new pricing arrangements as ‘fair’ but in some cases ‘difficult’.

A range of ideas on how to shift energy use were expressed during the sessions, including using electronic timers to turn appliances on overnight and taking children to the local pool during critical summer peak periods. We have developed six case studies representing archetypes of the various participants which include views on how they might respond to Flexible Pricing.

8.1.4 Price sensitivity analysis

Following a methodology based on Van Westendorp Price Sensitivity Analysis, we conducted over 3000 telephone surveys to identify the sensitivity of residential customers to increases in their electricity bills. We also asked questions regarding the level of difficulty people feel they face in trying to shift electricity use away from peak times and reduce peak electricity use. This analysis was intended to identify groups that are particularly sensitive to increases in their bills, such as those which might occur for some customers entering into Flexible Pricing.

The results of our analysis suggest that households classified as Regional or Single Parent are approximately twice as sensitive to increases in electricity bills as ‘Non-vulnerable’ consumers. Residential consumers indicated that every 1% increase in electricity bill would reduce the percentage of the group that could ‘afford’ it by 0.4%. In contrast, the percentage of Regional households that can afford a 1% increase in electricity bill decreases by 0.7%, and amongst single parent households affordability decreases by 0.66%.

When asked how easy or difficult they would find it to shift their use of electric heating and cooling appliances from peak times, most respondents indicated that it would be difficult to shift electricity use, but not hard or impossible. Most respondents indicated that to shift electricity use for appliances other than heating and cooling from peak times would be neither easy nor hard. Reducing electricity use during peak times was considered somewhat difficult, but not as difficult as shifting that usage from peak times.

Importantly, there was no statistically significant difference in the results for the vulnerable groups; meaning that in the quantitative survey, no one group expressed significantly more or less difficulty in shifting or reducing peak energy use than any other group, including residential consumers not in a vulnerable group themselves.

8.2 Verification of Stage 1 findings

As discussed in section 1 above, our final report for Stage 1 made 10 key findings.²⁸ Part of our objective in carrying out Stage 2 of this study was to further verify and build upon the Stage 1 results. The following table presents a summary of the Stage 1 findings and our comments on these following the analysis under Stage 2.

Table 22: Stage 1 key findings and comments

	Stage 1 findings	Stage 2 comments
1	Flexible Pricing will change the existing allocation of electricity costs across customer groups. The customer groups that are ‘winners’ and ‘losers’ are highly dependent on the structure and level of tariffs that are applied, existing tariff levels, whether customers currently have a controlled load off peak tariff, and how much customers alter their consumption in response to price changes. It is not the case that all customers within certain groups will always be better off or worse off as a result of Flexible Pricing; our analysis demonstrated the likely mean effect of Flexible Pricing on the bills of each group.	Stage 2 has confirmed that the effects of Flexible Pricing on household electricity bills are likely to be highly varied and dependent upon many factors. Our analysis has focused on the average (or mean) effect of Flexible Pricing on the identified vulnerable customer groups, however as is discussed throughout the report, there is significant variation around the average bill impact. The Stage 2 focus groups and quantitative phone surveys confirmed the high variability of household behaviours and attitudes to electricity use, which lend further support to this general finding (as well as providing additional insights into potential responses to Flexible Pricing).
2	For any particular tariff structure, retailers are likely to be able to adjust the details of the tariff structure (relative level of fixed, peak, off-peak, shoulder and CPP tariffs, and the time periods to which they apply) to increase or decrease the impact on particular customer groups, according to their own business model and objectives and the competitive market environment.	This point remains relevant to our analysis, although we note that the Government may elect to constrain the relative level of fixed, peak, off peak, shoulder and CPP tariffs and time periods in the initial years following the end of the current moratorium on ToU tariffs.
3	Under most tariff scenarios, the most marked difference in the impact on customers occurs between single and dual element meter customers rather than between vulnerable and non-vulnerable customers. However, this depends on the tariff structure itself and the assumptions applied in the calculation. The impact on dual element meter customers very much depends on how close the new off-peak rate is to the current dual element meter tariff (i.e. current off peak hot water tariffs).	In Stage 2 we tested a different set of ToU tariffs (tariffs 2A to 2F) to those tested in Stage 1 (tariffs 1A to 1F). Tariffs 2A to 2F were developed on the basis of further consultation between DPI and industry stakeholders. Consistent with Stage 1, the Stage 2 tariff analysis demonstrates significant variation in the effects between single and dual element meter customers. The higher the off peak price, the greater the increase (or smaller the decrease) in dual element meter customer bills. Stage 2 tariff analysis has also pointed out some differences in the effects on different vulnerable customer groups, as discussed in section 8.1.1 above.
4	Average annual bill changes under most of the modelled scenarios are relatively modest and fall in the range of +2% to -4% for the vulnerable groups, assuming zero elasticity.	All pricing scenarios tested in Stage 2 resulted in average bill changes of between -6.1% and +1.8% (for scenarios tested both with and without elasticity assumptions). This confirms that the average bill change resulting from Flexible Pricing is likely to be relatively modest.
5	If non-zero elasticity is applied then all modelled customer groups experience bill reductions of up to 9% under the tariff scenarios.29	For Stage 2 tariff scenarios tested with elasticity assumptions applied, all groups resulted in bill reductions of up to 6.1%.
6	Reductions in off-peak rates tend to benefit regional households as they tend to have relatively heavy overnight consumption.	The Stage 2 tariff tests indicated that low income households, people requiring disability support, health care card holders and regional groups are likely to be, on average, better off under each Stage 2 tariff scenario than under the base case. The Stage 2 tariff scenario results confirm that regional households are, on average, likely to face greater bill decreases than the other vulnerable groups if off peak rates are relatively low (that is, if the ratio of peak:off peak is higher than 2x, as in tariffs 2A, 2B, 2C, 2E, 2F).
7	Under Scenario 1A, all dual element meter residential customer groups modelled are worse off and all single element customer groups are better off.	In Stage 1, tariff scenario 1A was a single part tariff with CPP. We did not test single part tariffs in Stage 2. However, Stage 2 tariff 2B tested a three part ToU tariff with CPP. The results indicate that the CPP is likely to have highly variable impacts on the different vulnerable groups (such that it is not possible to identify precisely which vulnerable customer groups will be most affected by CPP in itself, however, that the impacts are likely to be variable among and within groups).
8	Households with individuals requiring disability assistance are usually better off and are only significantly worse off under Scenario 1A where they have dual element meters.	Testing our Stage 2 tariffs on households with people requiring disability assistance confirmed the Stage 1 finding that, on average, they are likely to have lower electricity bills under Flexible Pricing, regardless of whether they have a single element or dual element meter. We note that there were no single rate tariffs like 1A tested in Stage 2.
9	Regional households and health care card holders are better off under most modelled scenarios with the exception of Scenario 1A where they have dual element meters.	Stage 2 has confirmed these results: Regional households and health care card holders are likely, on average, to face lower electricity bills under Flexible Pricing, regardless of whether they have a single element or dual element meter. We note that there were no single rate tariffs like 1A tested in Stage 2.
10	Impacts on single income households and low net worth households, including whether they are better or worse off, vary quite markedly depending on the scenario modelled.	Stage 2 findings suggest that Flexible Pricing is likely to have particularly variable average results on elderly, single income and low net worth households. The impacts are varied between the various tariff scenarios tested, as well as between single and dual meter customers. However, given the tariffs tested in Stage 2 featured less variation than those tested in Stage 1, the differences between the tariff scenario results were not as significant as those found for Stage 1 tariffs. In summary, Stage 2 has found that varying peak, shoulder and off peak times by one or two hours (that is, starting peak at 4pm instead of 5pm) does not have a substantial impact on the average vulnerable customers’ bills.

In summary, Stage 2 has verified and built upon many of the findings of Stage 1. While Stage 2 involved testing a different set of tariff scenarios to those tested in Stage 1, the results of Stage 2, in terms of the variability of impacts, the average overall impacts, and the different impacts between vulnerable groups, are largely similar and lend support to our Stage 1 findings.

In addition, Stage 2 involved new primary research into customer attitudes to electricity bills and Flexible Pricing. This new analysis has provided valuable insights that will assist the Victorian Government to engage with the community on the introduction of Flexible Pricing and in managing household energy use.

8.3 Recommendations

The analysis and results outlined in this report serve to highlight and reinforce a consistent underlying theme of Stages 1 and 2 of the Customer Impact Study, being that the effects of Flexible Pricing on vulnerable customers are likely to be highly variable both among and within different groups.

However, our analysis suggests that, on average, if vulnerable people elect to take up Flexible Pricing, they will be better off than they are currently in that their total electricity bills over a year will be lower. Following this, if they also respond to the incentives created by shifting or lowering their peak consumption in response to the price incentives, our analysis concludes that on average they will face even lower electricity bills over the year.

This result suggests that it is not necessarily the case that all customers who identify as vulnerable will find Flexible Pricing results in higher electricity bills; indeed, the majority of customers are likely to be better off or to face very little changes to their electricity bills. This is not to say that some vulnerable customers won’t face increases in bills.

While on average there will be reductions, a proportion of customers will face increases if they elect to move to Flexible Pricing. The qualitative messages stemming from our research suggest that if people are given a clear, meaningful introduction to Flexible Pricing and if time is taken to explain the complexities of the arrangements and the reasons for the changes, then people are willing to and, in many cases, able to respond positively and embrace the changes.

However, it is clear that the level of information distributed to customers on the AMI rollout and Flexible Pricing to date has led to a level of confusion and distrust among those most vulnerable in the community. It is also clear that the level of understanding of what drives energy bills at present is very low among those vulnerable customers.

As noted above, we have found that some people in each vulnerable group will face significant increases in their electricity bills if they switch to Flexible Pricing, while others will achieve significant reductions in their bills. It is critical that customers are given enough information to make an informed, intelligent choice regarding whether they take up Flexible Pricing offers.

After considering the results of our analysis, we have developed three core recommendations relating to the introduction of Flexible Pricing.

8.3.1 Access to Flexible Pricing

The current moratorium on distribution ToU pricing is due to end on 1 January 2013. Based on the results of our analysis which show that Flexible Pricing will have a wide range of impacts on vulnerable customers’ electricity bills, we support the government’s position that the introduction of ToU pricing, including CPP and other incentives, should be on a voluntary or ‘opt in’ basis. Whilst customers should be given the choice to undertake Flexible Pricing, government and industry should take action to promote the benefits of ToU pricing to the Victorian community in order to achieve the efficiency that lower peak demand can deliver.

We understand that the Government is considering a regime to monitor the customer response to the AMI program, including the number of customers who sign up to ToU pricing. As part of this monitoring regime, we recommend:

• After a certain time has elapsed (say, three years), the Government consider whether the results of this monitoring indicate that a compulsory reassignment to ToU distribution tariffs is appropriate.

Consideration should also be given to how government can assist customers who are currently on payment plans or who find themselves locked into a ToU pricing structure. We recommend:

• Specifying a maximum contract lock-in period for ToU pricing of 6 months, or potentially providing a ‘cooling off’ period to enable vulnerable or even all customers to return to flat tariff rates within a certain period of time could lower the risk of significant hardship.

Lock-in periods on Victorian electricity retail market offers are currently up to 3 years, meaning that if a customer decides to switch retailers in that period they face an exit fee which can be more than $130.³⁰ In general, longer contract periods have higher exit fees, but often have more competitive c/kWh tariffs.

The longer a customer is able to trial a new tariff, the better understanding they will have of the impact of the new tariff on their bills. However, the longer the lock-in contract period, the greater the potential for them to suffer financial difficulties due to the new tariff. We therefore recommend:

• Consideration be given to the appropriate length of time needed to trial Flexible Pricing, which is particularly important given quarterly bills for some customers can vary significantly (for example, a winter bill might be 30% higher than an autumn bill for customers with electric heating).

Customers are generally aware of which time in the year they receive higher bills due to their own stock of heating and cooling appliances. A six month maximum lock-in contract would enable customers to receive two quarterly bills under the new tariff before exiting the contract without paying an exit fee, which should present enough information to enable an informed decision as to whether they are better off or worse off under Flexible Pricing over the whole year.

Alternatively (or in addition) to the maximum contract lock-in period, a shorter, 3 month ‘cooling off’ period could be applied to Flexible Pricing offers made to vulnerable (concessions) customers, to enable customers to switch back to their previous tariff after the first quarterly bill without incurring an exit fee. If applied in addition to a maximum contract period, consideration would need to be given to how both measures would operate in tandem and how any adverse implications could be avoided.³¹

The separate introduction of monthly billing would support a shorter ‘cooling off’ period (such as three months), as three separate monthly bills would have been received under the new tariff. However, we note that the introduction of monthly billing could itself have a positive impact on people’s perception of their electricity costs and their ability to pay their bills, which may distort their view of Flexible Pricing and prevent appropriate behavioural responses (whether to revert from the tariff or respond to the Flexible Pricing Incentives). We recommend the best way to minimise this impact is by:

• The design of information packages that supports Flexible Pricing, as discussed in the next section.

8.3.2 Information

Discussions in focus groups revealed that vulnerable customers feel they do not have enough information on:

• Drivers of energy use in the home
• Ways to save on their energy bills
• The purpose of smart meters and how they work
• Flexible pricing structures.

In approximately half of the focus group sessions, participants were asked who should provide such information and in what format it would best be received and understood. Participants consistently viewed government as the logical, appropriate primary source for such information, with bill inserts or mailed-out brochures viewed as the most appropriate format to explain the issues.

Given our findings that the impact of Flexible Pricing on vulnerable customers is likely to be highly variable with some customers likely to find they are significantly worse off, it is critical that customers understand the likely impacts before committing to a ToU pricing structure. We consider that the best way for this to occur is either:

• Via an online pricing comparator, similar to the ESC’s Your Choice website, using a consumption profile estimated for each customer based on their home appliances and their average hours of use per week and per season. This would need to be simple to use, yet sufficiently detailed to provide a meaningful result and would need to draw from actual ToU market and standing offers. Customers could also be given a chance to indicate the degree to which they can shift their appliance use under each tariff scenario (i.e. they could be asked to nominate how they would shift their energy use to identify the result).
• Via a mail-out (or emailed) information package, including the customers’ actual yearly average consumption profile data sourced from their current retailer to enable them to independently determine the likely impact using their own calculations. The package could also include some directions on how to calculate the bill impact using the consumption profile data.

We note that in order for such online comparators or mail out guides to be useful, customers would need to have access to their consumption data in reasonable detail (ideally half hourly data, or alternatively, consumption aggregated in line with the time periods of the Flexible Pricing tariffs).

In addition to information on Flexible Pricing, we consider that substantial savings could be achieved by providing clear, high quality information to customers on energy use in the home. We have found that there is a very low level of understanding amongst vulnerable customers of what is driving their energy bills. As a result, poor choices in appliance purchase and use are evident, driving very high bills for some customers. The government should seek to provide information to customers on usage levels for various appliances, either independently of information about Flexible Pricing itself, or more generally as part of a Flexible Pricing information package. It is our view that the successful delivery of this information is critical to the success of Flexible Pricing in Victoria.

8.3.3 Concessions

While a comprehensive review of the Victorian electricity concessions framework was beyond the scope of this study, our analysis has suggested that, on average, there is no single group of vulnerable customers that are likely to face more difficulties and hardship under Flexible Pricing structures than any other vulnerable group. We have noted that many of the focus group participants who were not elderly expressed particular concern for elderly people dealing with Flexible Pricing. However, the elderly people participating in focus groups and surveys did not generally indicate any more concerns with Flexible Pricing than other groups.

There is nothing in our findings to suggest that changes to the broad electricity concessions framework currently operating in Victoria are needed to protect particular customer groups. In our view, determining concessions on the basis of a percentage discount on overall energy bills continues to be relevant. However, this finding relies on clear, consistent information being provided to vulnerable customers on how Flexible Pricing will affect them, as discussed above.

We note that the introduction of Flexible Pricing on a voluntary or ‘opt-in’ basis adds an additional protection to customers receiving electricity concessions, provided the information provided to customers is sufficient to enable them to make informed decisions on whether to take up the new tariffs.

9 Limitation of our work

General use restriction

This report is prepared solely for the internal use of Department of Primary Industries Victoria. This report is not intended to and should not be used or relied upon by anyone else and we accept no duty of care to any other person or entity. The report has been prepared for the purpose set out in our proposal dated 28th March 2012. You should not refer to or use our name or the advice for any other purpose.

Appendix A

The following table presents the distribution of change of electricity bill for each combination of season, vulnerable customer group, element type, elasticity type and ToU scenario. Column “Percentile 5”, “Percentile 25”, “Percentile 75”, “Percentile 95” indicates the change in bill for the bottom 5%, 25%, 75% and 95% of customers respectively. “Mean” shows the average change of electricity bill. The final column indicates the estimated proportion of customers that would benefit from the ToU scenario, that is, would experience a decrease in electricity bill.

Table 23: Distribution of change of electricity bill and % customers better off under each scenario and against season, vulnerability group and element type³²

Bill season	Customer Group	Element	Scenario	Elasticity	Percentile 5	Percentile 25	Mean	Percentile 75	Percentile 95	Better off %
Annual	All residential	Single	A	Without	-42.7%	-23.1%	0.8%	14.5%	60.8%	55.7%
Annual	All residential	Single	B	Without	-42.6%	-23.0%	1.0%	14.6%	61.1%	55.6%
Annual	All residential	Single	C	Without	-42.7%	-23.0%	1.0%	14.7%	61.3%	55.5%
Annual	All residential	Single	D	Without	-43.0%	-23.5%	0.1%	13.4%	59.5%	56.9%
Annual	All residential	Single	E	Without	-42.8%	-23.2%	0.7%	14.3%	60.6%	55.9%
Annual	All residential	Single	F	Without	-42.6%	-23.0%	1.0%	14.7%	61.2%	55.4%
Annual	All residential	Dual	A	Without	-48.1%	-27.0%	-1.8%	18.8%	76.2%	54.5%
Annual	All residential	Dual	B	Without	-48.3%	-27.3%	-2.2%	18.1%	75.7%	55.0%
Annual	All residential	Dual	C	Without	-48.2%	-27.2%	-2.2%	18.2%	75.2%	55.0%
Annual	All residential	Dual	D	Without	-47.6%	-26.1%	-0.3%	21.0%	79.1%	52.7%
Annual	All residential	Dual	E	Without	-48.0%	-26.8%	-1.5%	19.3%	76.6%	54.0%
Annual	All residential	Dual	F	Without	-48.3%	-27.3%	-2.2%	18.1%	75.3%	55.0%
Annual	Elderly	Single	A	Without	-31.7%	-24.7%	0.6%	20.2%	30.5%	52.5%
Annual	Elderly	Single	B	Without	-31.4%	-24.6%	0.6%	20.1%	30.8%	52.5%
Annual	Elderly	Single	C	Without	-31.4%	-24.3%	0.9%	20.2%	30.6%	52.3%
Annual	Elderly	Single	D	Without	-31.7%	-24.3%	0.0%	18.6%	28.9%	53.4%
Annual	Elderly	Single	E	Without	-31.6%	-24.5%	0.1%	19.5%	29.5%	52.8%
Annual	Elderly	Single	F	Without	-31.6%	-24.7%	0.5%	19.9%	30.6%	52.7%
Annual	Elderly	Dual	A	Without	-31.1%	-29.4%	-1.6%	10.3%	12.2%	62.0%
Annual	Elderly	Dual	B	Without	-31.0%	-29.3%	-2.1%	9.8%	11.9%	62.5%
Annual	Elderly	Dual	C	Without	-30.9%	-29.3%	-1.9%	10.1%	12.0%	62.1%
Annual	Elderly	Dual	D	Without	-30.6%	-28.9%	-0.4%	11.2%	12.9%	61.0%
Annual	Elderly	Dual	E	Without	-31.4%	-29.8%	-2.0%	9.6%	11.3%	62.8%
Annual	Elderly	Dual	F	Without	-31.1%	-29.4%	-1.9%	9.9%	11.9%	62.5%
Annual	Needs assistance	Single	A	Without	-44.7%	-22.1%	-1.7%	26.9%	85.7%	47.5%
Annual	Needs assistance	Single	B	Without	-43.7%	-21.4%	-1.5%	27.4%	85.2%	46.9%
Annual	Needs assistance	Single	C	Without	-45.1%	-22.0%	-1.7%	26.7%	84.7%	47.6%
Annual	Needs assistance	Single	D	Without	-45.0%	-22.3%	-2.0%	26.3%	87.5%	47.9%
Annual	Needs assistance	Single	E	Without	-44.9%	-22.0%	-1.6%	26.7%	86.3%	47.6%
Annual	Needs assistance	Single	F	Without	-44.7%	-22.0%	-1.6%	26.9%	86.2%	47.5%
Annual	Single income	Single	A	Without	-43.0%	-27.1%	0.6%	16.7%	61.1%	56.0%
Annual	Single income	Single	B	Without	-42.8%	-27.0%	0.6%	16.9%	60.7%	55.8%
Annual	Single income	Single	C	Without	-42.6%	-26.6%	1.1%	17.4%	62.5%	55.3%
Annual	Single income	Single	D	Without	-43.3%	-27.6%	0.0%	16.2%	60.1%	56.5%
Annual	Single income	Single	E	Without	-42.6%	-26.4%	1.0%	17.2%	63.2%	55.2%
Annual	Single income	Single	F	Without	-42.8%	-26.8%	0.7%	16.8%	61.6%	55.8%
Annual	Single income	Dual	A	Without	-43.7%	-21.5%	-2.4%	13.5%	45.4%	55.7%
Annual	Single income	Dual	B	Without	-43.8%	-22.0%	-3.3%	12.5%	43.7%	56.9%
Annual	Single income	Dual	C	Without	-43.6%	-21.5%	-2.8%	13.1%	44.2%	56.1%
Annual	Single income	Dual	D	Without	-43.2%	-21.1%	-1.0%	16.0%	49.4%	53.4%
Annual	Single income	Dual	E	Without	-44.0%	-21.3%	-2.2%	14.3%	45.7%	54.9%
Annual	Single income	Dual	F	Without	-43.8%	-21.7%	-2.8%	12.8%	43.8%	56.5%
Annual	Low income	Single	A	Without	-42.9%	-26.3%	-0.4%	14.8%	49.5%	57.0%
Annual	Low income	Single	B	Without	-42.9%	-26.3%	-0.6%	14.5%	49.7%	57.3%
Annual	Low income	Single	C	Without	-42.7%	-25.9%	0.1%	15.7%	50.2%	56.1%
Annual	Low income	Single	D	Without	-43.0%	-26.4%	-0.6%	14.5%	49.2%	57.3%
Annual	Low income	Single	E	Without	-43.1%	-25.8%	-0.2%	15.3%	49.0%	56.4%
Annual	Low income	Single	F	Without	-42.9%	-26.3%	-0.3%	15.1%	49.7%	56.7%
Annual	Low income	Dual	A	Without	-54.1%	-24.9%	-3.3%	21.1%	84.5%	52.1%
Annual	Low income	Dual	B	Without	-54.0%	-25.2%	-3.7%	20.0%	82.3%	52.9%
Annual	Low income	Dual	C	Without	-54.1%	-25.1%	-3.7%	20.5%	82.7%	52.5%
Annual	Low income	Dual	D	Without	-54.0%	-24.4%	-1.4%	24.6%	91.3%	49.9%
Annual	Low income	Dual	E	Without	-53.9%	-24.9%	-2.8%	22.3%	86.2%	51.4%
Annual	Low income	Dual	F	Without	-54.3%	-25.2%	-3.7%	20.1%	82.1%	52.8%
Annual	Regional	Single	A	Without	-63.4%	-34.6%	-2.9%	17.8%	76.4%	58.0%
Annual	Regional	Single	B	Without	-63.0%	-34.8%	-3.1%	17.2%	74.5%	58.4%
Annual	Regional	Single	C	Without	-63.4%	-34.3%	-2.9%	17.9%	76.6%	57.9%
Annual	Regional	Single	D	Without	-63.7%	-34.8%	-2.8%	18.4%	77.1%	57.7%
Annual	Regional	Single	E	Without	-63.4%	-34.5%	-3.2%	17.6%	75.8%	58.1%
Annual	Regional	Single	F	Without	-63.5%	-34.4%	-3.0%	17.5%	76.1%	58.2%
Annual	Regional	Dual	A	Without	-57.2%	-34.5%	-2.7%	15.4%	89.1%	59.6%
Annual	Regional	Dual	B	Without	-57.2%	-34.7%	-3.0%	14.8%	88.7%	60.1%
Annual	Regional	Dual	C	Without	-57.3%	-34.7%	-3.1%	14.9%	88.0%	60.0%
Annual	Regional	Dual	D	Without	-56.8%	-33.5%	-1.2%	17.2%	92.4%	58.0%
Annual	Regional	Dual	E	Without	-57.2%	-34.3%	-2.4%	15.8%	89.4%	59.2%
Annual	Regional	Dual	F	Without	-57.3%	-34.8%	-3.2%	14.7%	87.9%	60.1%
Annual	Low net worth	Single	A	Without	-37.4%	-17.5%	1.3%	12.1%	40.5%	54.5%
Annual	Low net worth	Single	B	Without	-37.1%	-17.2%	1.8%	12.4%	40.7%	54.0%
Annual	Low net worth	Single	C	Without	-37.3%	-17.3%	1.7%	12.6%	41.3%	53.9%
Annual	Low net worth	Single	D	Without	-37.8%	-18.1%	0.4%	10.9%	39.4%	56.2%
Annual	Low net worth	Single	E	Without	-37.6%	-17.6%	1.2%	12.0%	40.5%	54.8%
Annual	Low net worth	Single	F	Without	-37.3%	-17.2%	1.7%	12.6%	41.1%	53.9%
Annual	Low net worth	Dual	A	Without	-44.1%	-25.3%	-2.1%	25.7%	64.6%	49.8%
Annual	Low net worth	Dual	B	Without	-43.9%	-25.2%	-2.5%	24.7%	64.3%	50.3%
Annual	Low net worth	Dual	C	Without	-44.4%	-25.4%	-2.4%	25.3%	63.7%	50.1%
Annual	Low net worth	Dual	D	Without	-44.2%	-24.8%	-0.7%	28.3%	67.8%	48.4%
Annual	Low net worth	Dual	E	Without	-44.3%	-25.4%	-1.7%	26.6%	65.3%	49.4%
Annual	Low net worth	Dual	F	Without	-44.2%	-25.4%	-2.5%	24.9%	63.7%	50.2%
Annual	Health Care Card Holders	Single	A	Without	-31.5%	-9.7%	-0.7%	26.4%	77.7%	38.4%
Annual	Health Care Card Holders	Single	B	Without	-31.4%	-9.5%	-0.5%	26.6%	77.9%	38.1%
Annual	Health Care Card Holders	Single	C	Without	-31.3%	-9.3%	-0.6%	26.8%	78.6%	37.9%
Annual	Health Care Card Holders	Single	D	Without	-31.9%	-10.5%	-1.2%	24.9%	76.2%	39.8%
Annual	Health Care Card Holders	Single	E	Without	-31.6%	-9.7%	-0.7%	26.2%	78.0%	38.4%
Annual	Health Care Card Holders	Single	F	Without	-31.4%	-9.4%	-0.6%	26.7%	78.1%	38.1%
Annual	Health Care Card Holders	Dual	A	Without	-56.0%	-25.5%	-2.3%	7.1%	101.1%	64.1%
Annual	Health Care Card Holders	Dual	B	Without	-56.3%	-25.8%	-3.4%	5.9%	100.6%	65.6%
Annual	Health Care Card Holders	Dual	C	Without	-56.1%	-25.6%	-2.8%	6.6%	99.7%	64.8%
Annual	Health Care Card Holders	Dual	D	Without	-55.2%	-24.2%	-0.8%	9.2%	103.6%	61.3%
Annual	Health Care Card Holders	Dual	E	Without	-55.9%	-25.0%	-1.9%	7.6%	101.2%	63.3%
Annual	Health Care Card Holders	Dual	F	Without	-56.2%	-25.7%	-2.8%	6.5%	100.3%	64.9%
Winter	All residential	Single	A	Without	-44.2%	-23.0%	-0.2%	16.2%	66.0%	54.4%
Winter	All residential	Single	B	Without	-42.0%	-19.6%	4.3%	21.6%	74.0%	48.8%
Winter	All residential	Single	C	Without	-44.3%	-23.3%	-0.5%	15.8%	65.4%	54.8%
Winter	All residential	Single	D	Without	-44.6%	-23.7%	-1.0%	14.9%	64.5%	55.7%
Winter	All residential	Single	E	Without	-44.7%	-23.7%	-0.9%	15.3%	64.9%	55.4%
Winter	All residential	Single	F	Without	-44.1%	-22.9%	-0.1%	16.3%	66.1%	54.2%
Winter	All residential	Dual	A	Without	-51.6%	-29.1%	-2.4%	19.3%	77.9%	55.0%
Winter	All residential	Dual	B	Without	-49.8%	-26.3%	1.3%	23.6%	84.7%	51.4%
Winter	All residential	Dual	C	Without	-51.9%	-29.6%	-3.4%	18.1%	75.8%	56.1%
Winter	All residential	Dual	D	Without	-51.2%	-28.3%	-1.2%	21.4%	80.9%	53.5%
Winter	All residential	Dual	E	Without	-51.8%	-29.3%	-2.8%	19.1%	77.4%	55.3%
Winter	All residential	Dual	F	Without	-51.7%	-29.3%	-2.9%	18.5%	76.8%	55.7%
Winter	Elderly	Single	A	Without	-35.3%	-24.8%	-1.7%	17.1%	28.4%	54.6%
Winter	Elderly	Single	B	Without	-33.4%	-22.9%	1.7%	21.8%	34.0%	50.6%
Winter	Elderly	Single	C	Without	-35.3%	-24.8%	-1.8%	16.5%	27.6%	55.0%
Winter	Elderly	Single	D	Without	-35.4%	-24.5%	-1.8%	16.8%	27.6%	54.6%
Winter	Elderly	Single	E	Without	-35.7%	-25.4%	-2.3%	16.7%	27.1%	55.1%
Winter	Elderly	Single	F	Without	-35.2%	-24.9%	-1.9%	16.7%	27.8%	54.9%
Winter	Elderly	Dual	A	Without	-37.3%	-35.7%	-2.0%	5.5%	7.4%	68.3%
Winter	Elderly	Dual	B	Without	-35.7%	-34.1%	1.1%	8.7%	10.9%	64.8%
Winter	Elderly	Dual	C	Without	-37.3%	-35.8%	-2.5%	4.9%	6.8%	68.9%
Winter	Elderly	Dual	D	Without	-36.7%	-35.1%	-0.9%	6.3%	7.9%	67.4%
Winter	Elderly	Dual	E	Without	-37.9%	-36.4%	-2.8%	4.2%	5.9%	69.8%
Winter	Elderly	Dual	F	Without	-37.4%	-35.9%	-2.4%	5.0%	6.9%	68.9%
Winter	Needs assistance	Single	A	Without	-9.0%	-6.2%	-1.7%	60.2%	63.4%	29.7%
Winter	Needs assistance	Single	B	Without	-4.9%	-2.1%	2.6%	67.3%	71.1%	26.5%
Winter	Needs assistance	Single	C	Without	-9.4%	-6.7%	-2.1%	58.7%	61.8%	30.1%
Winter	Needs assistance	Single	D	Without	-9.3%	-6.5%	-2.2%	60.3%	63.0%	29.9%
Winter	Needs assistance	Single	E	Without	-9.4%	-6.7%	-1.9%	59.4%	62.2%	30.1%
Winter	Needs assistance	Single	F	Without	-9.0%	-6.3%	-1.6%	59.8%	63.1%	29.7%
Winter	Single income	Single	A	Without	-42.9%	-25.6%	0.0%	24.1%	58.3%	50.8%
Winter	Single income	Single	B	Without	-41.0%	-22.4%	3.9%	28.6%	65.2%	47.0%
Winter	Single income	Single	C	Without	-42.8%	-25.5%	0.3%	24.9%	59.2%	50.3%
Winter	Single income	Single	D	Without	-43.4%	-26.2%	-0.4%	24.7%	58.0%	50.8%
Winter	Single income	Single	E	Without	-43.1%	-25.2%	0.3%	25.3%	59.1%	49.9%
Winter	Single income	Single	F	Without	-42.6%	-25.3%	0.2%	24.2%	59.3%	50.6%
Winter	Single income	Dual	A	Without	-49.5%	-27.6%	-3.2%	5.7%	50.2%	66.4%
Winter	Single income	Dual	B	Without	-48.8%	-25.2%	0.1%	9.1%	53.2%	61.8%
Winter	Single income	Dual	C	Without	-49.8%	-28.0%	-4.1%	4.7%	48.0%	67.8%
Winter	Single income	Dual	D	Without	-48.7%	-26.9%	-2.0%	7.6%	54.4%	64.0%
Winter	Single income	Dual	E	Without	-50.7%	-28.0%	-3.6%	5.5%	48.9%	66.8%
Winter	Single income	Dual	F	Without	-49.9%	-28.0%	-3.8%	4.9%	48.0%	67.6%
Winter	Low income	Single	A	Without	-46.0%	-32.0%	-0.1%	12.5%	56.0%	60.9%
Winter	Low income	Single	B	Without	-44.6%	-29.6%	3.8%	17.0%	63.1%	56.8%
Winter	Low income	Single	C	Without	-45.8%	-32.1%	0.1%	13.5%	56.8%	60.2%
Winter	Low income	Single	D	Without	-45.9%	-32.3%	-0.5%	12.1%	55.2%	61.4%
Winter	Low income	Single	E	Without	-46.0%	-32.1%	-0.2%	13.0%	55.2%	60.6%
Winter	Low income	Single	F	Without	-45.9%	-32.2%	0.0%	13.0%	55.8%	60.6%
Winter	Low income	Dual	A	Without	-60.9%	-27.3%	-4.3%	25.7%	112.2%	50.7%
Winter	Low income	Dual	B	Without	-59.6%	-24.2%	-1.3%	29.2%	118.6%	47.6%
Winter	Low income	Dual	C	Without	-61.1%	-27.7%	-5.0%	24.9%	109.8%	51.3%
Winter	Low income	Dual	D	Without	-60.9%	-27.0%	-2.4%	29.7%	120.1%	48.8%
Winter	Low income	Dual	E	Without	-61.1%	-27.5%	-4.2%	26.6%	114.0%	50.4%
Winter	Low income	Dual	F	Without	-61.0%	-27.4%	-4.8%	24.8%	109.2%	51.2%
Winter	Regional	Single	A	Without	-60.3%	-33.9%	-3.6%	13.7%	73.4%	60.6%
Winter	Regional	Single	B	Without	-58.8%	-30.9%	0.2%	17.9%	80.1%	56.6%
Winter	Regional	Single	C	Without	-60.2%	-34.2%	-4.1%	12.9%	72.6%	61.3%
Winter	Regional	Single	D	Without	-60.6%	-34.6%	-3.5%	14.3%	74.3%	60.4%
Winter	Regional	Single	E	Without	-60.2%	-34.6%	-4.3%	12.6%	72.3%	61.7%
Winter	Regional	Single	F	Without	-60.3%	-33.8%	-3.8%	13.1%	72.7%	61.0%
Winter	Regional	Dual	A	Without	-57.9%	-35.5%	-3.3%	18.3%	84.7%	58.0%
Winter	Regional	Dual	B	Without	-56.5%	-33.2%	0.2%	22.8%	91.8%	54.6%
Winter	Regional	Dual	C	Without	-58.2%	-36.1%	-4.3%	17.2%	82.6%	58.9%
Winter	Regional	Dual	D	Without	-57.5%	-34.6%	-1.9%	20.2%	87.9%	56.6%
Winter	Regional	Dual	E	Without	-58.1%	-35.7%	-3.6%	18.1%	83.9%	58.2%
Winter	Regional	Dual	F	Without	-58.1%	-35.9%	-3.9%	17.5%	83.5%	58.6%
Winter	Low net worth	Single	A	Without	-35.4%	-15.7%	0.1%	14.9%	41.7%	50.6%
Winter	Low net worth	Single	B	Without	-33.0%	-12.0%	4.4%	20.1%	48.2%	43.7%
Winter	Low net worth	Single	C	Without	-35.7%	-16.0%	-0.2%	14.7%	41.3%	51.0%
Winter	Low net worth	Single	D	Without	-35.7%	-16.4%	-0.8%	13.7%	40.7%	52.3%
Winter	Low net worth	Single	E	Without	-36.0%	-16.5%	-0.8%	14.0%	40.5%	52.0%
Winter	Low net worth	Single	F	Without	-35.4%	-15.5%	0.3%	15.3%	41.9%	50.2%
Winter	Low net worth	Dual	A	Without	-48.0%	-27.4%	-2.8%	26.4%	80.3%	50.5%
Winter	Low net worth	Dual	B	Without	-45.9%	-24.7%	0.7%	30.1%	87.7%	47.6%
Winter	Low net worth	Dual	C	Without	-48.9%	-28.1%	-3.8%	25.8%	77.9%	51.0%
Winter	Low net worth	Dual	D	Without	-48.0%	-26.7%	-1.4%	29.8%	82.3%	48.6%
Winter	Low net worth	Dual	E	Without	-49.1%	-28.0%	-3.1%	27.2%	78.6%	50.4%
Winter	Low net worth	Dual	F	Without	-48.3%	-27.6%	-3.3%	25.5%	79.5%	51.0%
Winter	Health Care Card Holders	Single	A	Without	-35.1%	-11.1%	-1.2%	27.9%	80.7%	39.2%
Winter	Health Care Card Holders	Single	B	Without	-32.7%	-6.7%	2.9%	34.1%	89.4%	33.2%
Winter	Health Care Card Holders	Single	C	Without	-35.1%	-11.3%	-1.6%	27.4%	80.1%	39.6%
Winter	Health Care Card Holders	Single	D	Without	-35.4%	-12.3%	-1.8%	26.1%	79.1%	41.0%
Winter	Health Care Card Holders	Single	E	Without	-35.5%	-11.6%	-1.8%	26.8%	79.8%	40.1%
Winter	Health Care Card Holders	Single	F	Without	-35.0%	-10.9%	-1.2%	28.0%	80.6%	39.0%
Winter	Health Care Card Holders	Dual	A	Without	-51.0%	-2.4%	-1.6%	37.4%	229.4%	28.0%
Winter	Health Care Card Holders	Dual	B	Without	-49.9%	1.3%	2.1%	42.5%	244.4%	23.4%
Winter	Health Care Card Holders	Dual	C	Without	-51.3%	-3.0%	-1.5%	36.1%	225.4%	28.9%
Winter	Health Care Card Holders	Dual	D	Without	-50.2%	-0.9%	-0.1%	39.5%	232.1%	26.1%
Winter	Health Care Card Holders	Dual	E	Without	-51.2%	-2.0%	-0.5%	37.2%	228.0%	27.6%
Winter	Health Care Card Holders	Dual	F	Without	-51.3%	-2.8%	-1.8%	36.6%	227.9%	28.5%
Summer	All residential	Single	A	Without	-44.4%	-25.4%	1.8%	11.3%	55.4%	59.6%
Summer	All residential	Single	B	Without	-39.0%	-17.3%	13.5%	24.4%	74.2%	45.8%
Summer	All residential	Single	C	Without	-44.0%	-24.8%	2.8%	12.5%	57.3%	58.2%
Summer	All residential	Single	D	Without	-44.4%	-25.5%	1.3%	10.5%	54.5%	60.4%
Summer	All residential	Single	E	Without	-43.8%	-24.8%	2.6%	12.2%	56.5%	58.6%
Summer	All residential	Single	F	Without	-44.2%	-25.2%	2.2%	11.8%	56.2%	59.0%
Summer	All residential	Dual	A	Without	-46.0%	-26.2%	-0.5%	18.5%	77.4%	54.3%
Summer	All residential	Dual	B	Without	-41.5%	-19.0%	9.6%	30.8%	97.7%	44.1%
Summer	All residential	Dual	C	Without	-45.8%	-25.8%	0.0%	19.0%	78.4%	53.8%
Summer	All residential	Dual	D	Without	-45.3%	-25.2%	1.2%	20.5%	79.6%	52.6%
Summer	All residential	Dual	E	Without	-45.6%	-25.4%	0.7%	20.0%	79.3%	52.9%
Summer	All residential	Dual	F	Without	-46.1%	-26.4%	-0.7%	18.1%	77.1%	54.6%
Summer	Elderly	Single	A	Without	-26.1%	-20.6%	2.0%	14.2%	30.2%	54.6%
Summer	Elderly	Single	B	Without	-20.1%	-14.4%	12.1%	27.3%	44.5%	42.3%
Summer	Elderly	Single	C	Without	-25.2%	-19.6%	3.1%	15.0%	31.7%	53.3%
Summer	Elderly	Single	D	Without	-25.9%	-20.1%	1.5%	12.5%	28.9%	55.8%
Summer	Elderly	Single	E	Without	-25.3%	-19.7%	2.2%	13.9%	30.0%	54.3%
Summer	Elderly	Single	F	Without	-25.9%	-20.5%	2.1%	14.0%	31.4%	54.7%
Summer	Elderly	Dual	A	Without	-22.7%	-20.7%	-0.7%	17.6%	19.6%	52.1%
Summer	Elderly	Dual	B	Without	-16.1%	-13.9%	7.6%	27.6%	30.2%	41.8%
Summer	Elderly	Dual	C	Without	-22.1%	-20.1%	-0.4%	18.2%	20.3%	51.3%
Summer	Elderly	Dual	D	Without	-22.4%	-20.4%	0.7%	18.8%	20.5%	51.0%
Summer	Elderly	Dual	E	Without	-22.5%	-20.5%	-0.3%	17.6%	19.5%	51.9%
Summer	Elderly	Dual	F	Without	-22.5%	-20.6%	-0.6%	17.5%	19.6%	52.0%
Summer	Needs assistance	Single	A	Without	-59.7%	-25.2%	-1.0%	9.1%	89.1%	61.7%
Summer	Needs assistance	Single	B	Without	-56.5%	-17.7%	9.6%	21.6%	109.7%	47.5%
Summer	Needs assistance	Single	C	Without	-60.1%	-24.5%	-0.3%	10.2%	88.6%	60.3%
Summer	Needs assistance	Single	D	Without	-60.2%	-25.3%	-1.1%	8.5%	91.6%	62.5%
Summer	Needs assistance	Single	E	Without	-59.9%	-24.4%	-0.1%	10.1%	90.1%	60.4%
Summer	Needs assistance	Single	F	Without	-59.6%	-25.1%	-0.7%	9.5%	90.6%	61.3%
Summer	Needs assistance	Dual	A	Without	-37.5%	-35.7%	-1.1%	0.8%	2.6%	73.9%
Summer	Needs assistance	Dual	B	Without	-31.1%	-29.2%	9.3%	11.8%	14.1%	60.6%
Summer	Needs assistance	Dual	C	Without	-36.9%	-35.2%	-0.7%	1.4%	3.3%	73.1%
Summer	Needs assistance	Dual	D	Without	-37.3%	-35.5%	0.1%	1.2%	2.8%	73.3%
Summer	Needs assistance	Dual	E	Without	-36.8%	-35.1%	0.0%	1.9%	3.6%	72.4%
Summer	Needs assistance	Dual	F	Without	-37.4%	-35.6%	-1.1%	0.8%	2.8%	74.0%
Summer	Single income	Single	A	Without	-42.6%	-27.7%	1.2%	12.4%	53.6%	59.5%
Summer	Single income	Single	B	Without	-37.8%	-21.6%	11.7%	25.5%	71.6%	47.9%
Summer	Single income	Single	C	Without	-41.7%	-26.8%	2.2%	13.2%	55.5%	58.5%
Summer	Single income	Single	D	Without	-42.5%	-27.6%	0.7%	11.2%	52.4%	60.6%
Summer	Single income	Single	E	Without	-41.1%	-26.7%	2.1%	13.1%	56.3%	58.5%
Summer	Single income	Single	F	Without	-42.2%	-27.5%	1.4%	12.6%	53.8%	59.3%
Summer	Single income	Dual	A	Without	-35.0%	-14.7%	-0.9%	25.2%	51.4%	43.5%
Summer	Single income	Dual	B	Without	-28.6%	-6.6%	7.9%	36.5%	64.3%	32.7%
Summer	Single income	Dual	C	Without	-34.0%	-14.0%	-0.2%	26.1%	52.0%	42.4%
Summer	Single income	Dual	D	Without	-35.1%	-14.4%	0.7%	28.1%	56.1%	42.0%
Summer	Single income	Dual	E	Without	-33.5%	-14.2%	0.2%	27.5%	53.9%	42.0%
Summer	Single income	Dual	F	Without	-34.6%	-14.7%	-0.9%	24.7%	50.4%	43.6%
Summer	Low income	Single	A	Without	-37.8%	-18.5%	-0.5%	15.4%	45.9%	52.2%
Summer	Low income	Single	B	Without	-33.9%	-12.1%	8.1%	26.2%	61.0%	40.8%
Summer	Low income	Single	C	Without	-37.6%	-17.4%	0.4%	16.5%	46.9%	50.6%
Summer	Low income	Single	D	Without	-37.8%	-18.1%	-0.3%	15.5%	45.5%	52.0%
Summer	Low income	Single	E	Without	-38.0%	-17.4%	0.2%	16.2%	46.0%	50.9%
Summer	Low income	Single	F	Without	-37.9%	-18.4%	-0.4%	15.6%	46.7%	52.1%
Summer	Low income	Dual	A	Without	-47.9%	-28.2%	-1.9%	14.0%	45.7%	58.4%
Summer	Low income	Dual	B	Without	-43.6%	-21.6%	7.1%	24.0%	58.3%	48.7%
Summer	Low income	Dual	C	Without	-47.3%	-28.0%	-1.6%	14.7%	45.7%	57.8%
Summer	Low income	Dual	D	Without	-47.5%	-27.6%	-0.1%	17.0%	50.5%	55.9%
Summer	Low income	Dual	E	Without	-47.0%	-28.0%	-0.8%	16.0%	47.9%	56.8%
Summer	Low income	Dual	F	Without	-47.8%	-28.4%	-2.2%	13.4%	44.4%	59.0%
Summer	Regional	Single	A	Without	-61.5%	-37.6%	-2.0%	17.8%	65.8%	58.9%
Summer	Regional	Single	B	Without	-58.0%	-31.7%	8.9%	30.5%	82.4%	50.5%
Summer	Regional	Single	C	Without	-61.3%	-36.6%	-1.2%	18.8%	69.0%	58.0%
Summer	Regional	Single	D	Without	-61.7%	-37.0%	-1.7%	18.2%	68.7%	58.5%
Summer	Regional	Single	E	Without	-61.4%	-36.6%	-1.4%	18.1%	68.2%	58.5%
Summer	Regional	Single	F	Without	-61.5%	-37.3%	-1.9%	17.6%	66.9%	59.0%
Summer	Regional	Dual	A	Without	-57.2%	-35.3%	-1.4%	10.9%	94.0%	63.2%
Summer	Regional	Dual	B	Without	-53.4%	-28.8%	9.1%	22.7%	116.9%	53.0%
Summer	Regional	Dual	C	Without	-56.9%	-34.9%	-0.8%	11.6%	95.0%	62.5%
Summer	Regional	Dual	D	Without	-56.9%	-34.5%	0.1%	12.6%	96.4%	61.6%
Summer	Regional	Dual	E	Without	-56.9%	-34.7%	-0.2%	12.3%	95.5%	61.9%
Summer	Regional	Dual	F	Without	-57.3%	-35.5%	-1.6%	10.6%	93.3%	63.5%
Summer	Low net worth	Single	A	Without	-40.5%	-21.0%	2.8%	7.2%	37.0%	62.3%
Summer	Low net worth	Single	B	Without	-34.0%	-11.8%	15.1%	20.1%	53.3%	43.5%
Summer	Low net worth	Single	C	Without	-40.0%	-20.0%	4.2%	8.6%	39.0%	60.0%
Summer	Low net worth	Single	D	Without	-41.1%	-21.5%	1.9%	6.0%	36.0%	64.1%
Summer	Low net worth	Single	E	Without	-39.8%	-20.1%	3.8%	8.2%	38.2%	60.6%
Summer	Low net worth	Single	F	Without	-40.0%	-20.4%	3.5%	7.8%	38.0%	61.1%
Summer	Low net worth	Dual	A	Without	-39.7%	-18.7%	-0.4%	25.9%	64.5%	46.0%
Summer	Low net worth	Dual	B	Without	-32.8%	-9.4%	10.0%	39.0%	81.0%	34.7%
Summer	Low net worth	Dual	C	Without	-38.9%	-17.8%	0.4%	26.2%	67.0%	45.2%
Summer	Low net worth	Dual	D	Without	-39.8%	-18.6%	0.9%	27.7%	68.7%	45.1%
Summer	Low net worth	Dual	E	Without	-38.6%	-17.9%	1.2%	27.3%	69.5%	44.8%
Summer	Low net worth	Dual	F	Without	-39.4%	-18.4%	-0.4%	25.3%	64.3%	46.0%
Summer	Health Care Card Holders	Single	A	Without	-30.6%	-11.9%	0.1%	22.1%	72.4%	42.5%
Summer	Health Care Card Holders	Single	B	Without	-23.8%	-2.5%	11.0%	36.4%	92.6%	28.2%
Summer	Health Care Card Holders	Single	C	Without	-29.9%	-10.9%	1.1%	23.7%	75.2%	40.8%
Summer	Health Care Card Holders	Single	D	Without	-30.9%	-12.2%	-0.2%	21.2%	71.7%	43.3%
Summer	Health Care Card Holders	Single	E	Without	-30.2%	-11.3%	1.0%	23.2%	74.4%	41.4%
Summer	Health Care Card Holders	Single	F	Without	-30.4%	-11.7%	0.4%	22.6%	73.1%	42.0%
Annual	All residential	Single	A	With	-43.3%	-24.3%	-1.2%	12.0%	56.8%	58.5%
Annual	All residential	Single	B	With	-43.0%	-23.8%	-0.5%	12.7%	57.9%	57.6%
Annual	All residential	Single	C	With	-43.3%	-24.3%	-1.2%	12.0%	56.9%	58.5%
Annual	All residential	Single	D	With	-43.1%	-24.0%	-0.8%	12.2%	57.4%	58.2%
Annual	All residential	Single	E	With	-43.9%	-25.2%	-2.5%	10.4%	54.5%	60.4%
Annual	All residential	Single	F	With	-43.3%	-24.3%	-1.2%	12.0%	56.8%	58.5%
Annual	All residential	Dual	A	With	-48.7%	-28.1%	-3.5%	16.7%	72.5%	56.3%
Annual	All residential	Dual	B	With	-48.7%	-28.0%	-3.5%	16.6%	72.8%	56.4%
Annual	All residential	Dual	C	With	-48.8%	-28.4%	-4.1%	15.9%	71.2%	57.0%
Annual	All residential	Dual	D	With	-47.7%	-26.5%	-1.0%	20.1%	77.3%	53.4%
Annual	All residential	Dual	E	With	-49.1%	-28.6%	-4.2%	16.0%	71.0%	57.1%
Annual	All residential	Dual	F	With	-48.9%	-28.4%	-4.1%	15.8%	71.2%	57.2%
Annual	Elderly	Single	A	With	-32.2%	-25.4%	-1.2%	17.6%	27.4%	54.5%
Annual	Elderly	Single	B	With	-31.7%	-25.1%	-0.7%	18.1%	28.4%	54.0%
Annual	Elderly	Single	C	With	-32.0%	-25.2%	-1.1%	17.4%	27.3%	54.6%
Annual	Elderly	Single	D	With	-31.8%	-24.7%	-0.8%	17.4%	27.3%	54.3%
Annual	Elderly	Single	E	With	-32.7%	-26.0%	-2.7%	15.6%	25.0%	56.2%
Annual	Elderly	Single	F	With	32.2%	-25.5%	-1.4%	17.2%	27.3%	54.9%
Annual	Elderly	Dual	A	With	-31.5%	-30.1%	-3.1%	8.6%	10.2%	63.9%
Annual	Elderly	Dual	B	With	-31.3%	-29.8%	-3.1%	8.6%	10.4%	63.8%
Annual	Elderly	Dual	C	With	-31.5%	-30.1%	-3.5%	8.3%	9.9%	64.2%
Annual	Elderly	Dual	D	With	-30.7%	-29.2%	-1.1%	10.5%	11.9%	61.8%
Annual	Elderly	Dual	E	With	-32.5%	-31.1%	-4.4%	6.9%	8.3%	66.0%
Annual	Elderly	Dual	F	With	-31.7%	-30.3%	-3.5%	8.0%	9.7%	64.5%
Annual	Needs assistance	Single	A	With	-45.2%	-23.2%	-3.5%	24.3%	82.1%	49.4%
Annual	Needs assistance	Single	B	With	-44.2%	-22.2%	-2.9%	25.4%	82.4%	48.4%
Annual	Needs assistance	Single	C	With	-45.6%	-23.2%	-3.7%	23.9%	80.8%	49.6%
Annual	Needs assistance	Single	D	With	-45.1%	-22.7%	-2.8%	25.2%	85.9%	48.7%
Annual	Needs assistance	Single	E	With	-45.8%	-24.0%	-4.5%	22.7%	80.6%	50.7%
Annual	Needs assistance	Single	F	With	-45.2%	-23.3%	-3.6%	24.1%	81.9%	49.6%
Annual	Single income	Single	A	With	-43.5%	-28.1%	-1.3%	14.4%	57.2%	58.1%
Annual	Single income	Single	B	With	-43.1%	-27.8%	-0.7%	15.1%	57.9%	57.4%
Annual	Single income	Single	C	With	-43.2%	-27.8%	-1.0%	14.8%	58.1%	57.7%
Annual	Single income	Single	D	With	-43.5%	-28.0%	-0.9%	15.1%	58.0%	57.5%
Annual	Single income	Single	E	With	-43.7%	-28.2%	-2.0%	13.5%	57.1%	58.8%
Annual	Single income	Single	F	With	-43.4%	-28.0%	-1.3%	14.3%	57.3%	58.1%
Annual	Single income	Dual	A	With	-44.0%	-22.5%	-4.0%	11.8%	43.4%	57.8%
Annual	Single income	Dual	B	With	-43.9%	-22.6%	-4.4%	11.3%	42.3%	58.3%
Annual	Single income	Dual	C	With	-44.0%	-22.6%	-4.5%	11.2%	42.0%	58.4%
Annual	Single income	Dual	D	With	-43.2%	-21.4%	-1.7%	15.3%	48.5%	54.1%
Annual	Single income	Dual	E	With	-44.8%	-23.1%	-4.7%	11.5%	42.4%	58.4%
Annual	Single income	Dual	F	With	-44.2%	-22.7%	-4.5%	10.9%	41.6%	58.8%
Annual	Low income	Single	A	With	-43.3%	-27.2%	-2.0%	12.7%	46.3%	59.1%
Annual	Low income	Single	B	With	-43.1%	-26.9%	-1.8%	12.9%	47.2%	58.8%
Annual	Low income	Single	C	With	-43.2%	-27.0%	-1.8%	13.2%	46.5%	58.5%
Annual	Low income	Single	D	With	-43.1%	-26.8%	-1.4%	13.4%	47.5%	58.4%
Annual	Low income	Single	E	With	-43.9%	-27.5%	-3.0%	11.7%	44.0%	60.1%
Annual	Low income	Single	F	With	-43.3%	-27.3%	-2.1%	12.7%	46.1%	59.1%
Annual	Low income	Dual	A	With	-54.8%	-26.3%	-4.7%	19.6%	81.9%	53.7%
Annual	Low income	Dual	B	With	-54.6%	-26.2%	-4.8%	19.0%	80.5%	54.0%
Annual	Low income	Dual	C	With	-54.8%	-26.6%	-5.3%	18.8%	79.9%	54.3%
Annual	Low income	Dual	D	With	-54.3%	-25.1%	-2.0%	24.0%	90.4%	50.5%
Annual	Low income	Dual	E	With	-55.0%	-27.0%	-5.2%	19.5%	82.0%	54.0%
Annual	Low income	Dual	F	With	-55.0%	-26.6%	-5.4%	18.4%	79.3%	54.6%
Annual	Regional	Single	A	With	-63.8%	-35.6%	-4.7%	15.6%	72.5%	59.8%
Annual	Regional	Single	B	With	-63.3%	-35.4%	-4.4%	15.6%	71.8%	59.7%
Annual	Regional	Single	C	With	-63.7%	-35.4%	-4.9%	15.4%	72.2%	59.9%
Annual	Regional	Single	D	With	-63.8%	-35.2%	-3.6%	17.3%	75.1%	58.5%
Annual	Regional	Single	E	With	-64.1%	-36.3%	-6.1%	13.9%	69.7%	61.1%
Annual	Regional	Single	F	With	-63.9%	-35.5%	-5.0%	15.0%	71.8%	60.1%
Annual	Regional	Dual	A	With	-57.7%	-35.4%	-4.4%	13.2%	85.0%	61.4%
Annual	Regional	Dual	B	With	-57.6%	-35.4%	-4.3%	13.2%	85.4%	61.4%
Annual	Regional	Dual	C	With	-57.8%	-35.7%	-5.0%	12.5%	83.6%	62.0%
Annual	Regional	Dual	D	With	-57.0%	-34.0%	-2.0%	16.3%	90.4%	58.8%
Annual	Regional	Dual	E	With	-58.1%	-36.0%	-5.2%	12.4%	83.3%	62.2%
Annual	Regional	Dual	F	With	-57.9%	-35.8%	-5.1%	12.4%	83.4%	62.2%
Annual	Low net worth	Single	A	With	-38.1%	-18.8%	-0.8%	9.6%	37.0%	58.1%
Annual	Low net worth	Single	B	With	-37.5%	-18.2%	0.2%	10.5%	38.1%56.7%
Annual	Low net worth	Single	C	With	-38.1%	-18.7%	-0.6%	9.9%	37.4%	57.7%
Annual	Low net worth	Single	D	With	-37.9%	-18.7%	-0.6%	9.7%	37.6%	57.9%
Annual	Low net worth	Single	E	With	-38.8%	-19.9%	-2.1%	8.0%	35.2%	60.6%
Annual	Low net worth	Single	F	With	-38.0%	-18.7%	-0.6%	9.8%	37.2%	57.8
Annual	Low net worth	Dual	A	With	-44.3%	-26.2%	-3.8%	23.4%	62.5%	51.4%
Annual	Low net worth	Dual	B	With	-43.9%	-25.7%	-3.8%	23.0%	62.9%	51.4%
Annual	Low net worth	Dual	C	With	-44.6%	-26.4%	-4.3%	22.7%	61.3%	51.9%
Annual	Low net worth	Dual	D	With	-43.9%	-25.0%	-1.4%	27.2%	67.2%	49.0%
Annual	Low net worth	Dual	E	With	-45.0%	-27.1%	-4.4%	22.9%	61.4%	52.1%
Annual	Low net worth	Dual	F	With	-44.5%	-26.3%	-4.4%	22.4%	61.2%	52.0%
Annual	Health Care Card Holders	Single	A	With	-32.2%	-11.2%	-2.6%	23.5%	73.3%	41.1%
Annual	Health Care Card Holders	Single	B	With	-31.8%	-10.5%	-1.9%	24.4%	74.4%	40.1%
Annual	Health Care Card Holders	Single	C	With	-32.1%	-11.0%	-2.6%	23.7%	73.6%	40.8%
Annual	Health Care Card Holders	Single	D	With	-32.1%	-11.1%	-2.1%	23.5%	73.9%	41.0%
Annual	Health Care Card Holders	Single	E	With	-33.0%	-12.2%	-3.7%	21.8%	71.2%	43.0%
Annual	Health Care Card Holders	Single	F	With	-32.2%	-11.1%	-2.7%	23.6%	73.2%	41.0%
Annual	Health Care Card Holders	Dual	A	With	-56.1%	-26.3%	-4.0%	5.6%	96.6%	66.2%
Annual	Health Care Card Holders	Dual	B	With	-56.3%	-26.3%	-4.6%	5.0%	97.0%	67.0%
Annual	Health Care Card Holders	Dual	C	With	-56.3%	-26.6%	-4.5%	5.0%	95.0%	67.1%
Annual	Health Care Card Holders	Dual	D	With	-55.1%	-24.5%	-1.5%	8.7%	101.5%	61.9%
Annual	Health Care Card Holders	Dual	E	With	-56.3%	-26.5%	-4.5%	5.1%	94.7%	66.9%
Annual	Health Care Card Holders	Dual	F	With	-56.4%	-26.6%	-4.6%	4.9%	95.3%	67.3%
Winter	All residential	Single	A	With	-44.7%	-24.3%	-2.2%	13.6%	61.9%	57.0%
Winter	All residential	Single	B	With	-42.6%	-20.9%	2.2%	19.0%	69.8%	51.2%
Winter	All residential	Single	C	With	-44.9%	-24.6%	-2.6%	13.1%	61.1%	57.6%
Winter	All residential	Single	D	With	-44.7%	-24.2%	-1.9%	13.8%	62.5%	56.9%
Winter	All residential	Single	E	With	-45.7%	-25.7%	-4.0%	11.4%	58.9%	59.6%
Winter	All residential	Single	F	With	-44.8%	-24.3%	-2.3%	13.5%	61.6%	57.1%
Winter	All residential	Dual	A	With	-52.2%	-30.1%	-4.2%	17.1%	74.1%	56.9%
Winter	All residential	Dual	B	With	-50.5%	-27.5%	-0.5%	21.3%	80.8%	53.2%
Winter	All residential	Dual	C	With	-52.5%	-30.7%	-5.1%	15.8%	72.0%	58.0%
Winter	All residential	Dual	D	With	-51.4%	-28.7%	-1.9%	20.4%	79.1%	54.2%
Winter	All residential	Dual	E	With	-52.8%	-31.0%	-5.4%	15.8%	72.0%	58.1%
Winter	All residential	Dual	F	With	-52.4%	-30.5%	-4.8%	16.1%	72.7%	57.7%
Winter	Elderly	Single	A	With	-35.8%	-25.5%	-3.2%	14.9%	25.7%	56.5%
Winter	Elderly	Single	B	With	-33.9%	-23.6%	0.1%	19.5%	31.1%	52.3%
Winter	Elderly	Single	C	With	-35.8%	-25.6%	-3.5%	14.3%	24.8%	57.1%
Winter	Elderly	Single	D	With	-35.4%	-24.8%	-2.5%	15.9%	26.3%	55.4%
Winter	Elderly	Single	E	With	-36.7%	-26.8%	-4.8%	13.5%	23.2%	58.2%
Winter	Elderly	Single	F	With	-35.8%	-25.7%	-3.5%	14.4%	24.9%	57.0%
Winter	Elderly	Dual	A	With	-37.6%	-36.3%	-3.5%	3.9%	5.4%	70.2%
Winter	Elderly	Dual	B	With	-36.1%	-34.7%	-0.5%	7.1%	8.9%	66.5%
Winter	Elderly	Dual	C	With	-37.7%	-36.4%	-4.1%	3.2%	4.8%	70.9%
Winter	Elderly	Dual	D	With	-36.7%	-35.3%	-1.6%	5.6%	7.0%	68.2%
Winter	Elderly	Dual	E	With	-38.8%	-37.5%	-5.2%	1.7%	3.0%	72.9%
Winter	Elderly	Dual	F	With	-37.9%	-36.5%	-4.0%	3.2%	4.9%	71.0%
Winter	Needs assistance	Single	A	With	-10.0%	-7.6%	-3.5%	57.1%	59.7%	30.9%
Winter	Needs assistance	Single	B	With	-6.0%	-3.6%	0.7%	64.0%	67.1%	27.7%
Winter	Needs assistance	Single	C	With	-10.4%	-8.1%	-4.0%	55.5%	58.1%	31.4%
Winter	Needs assistance	Single	D	With	-9.4%	-7.0%	-2.9%	59.0%	61.3%	30.3%
Winter	Needs assistance	Single	E	With	-11.3%	-9.0%	-4.8%	54.6%	56.9%	32.1%
Winter	Needs assistance	Single	F	With	-10.1%	-7.8%	-3.6%	56.3%	59.1%	31.1%
Winter	Single income	Single	A	With	-43.4%	-26.7%	-1.8%	21.9%	54.8%	52.5%
Winter	Single income	Single	B	With	-41.5%	-23.6%	2.1%	26.5%	61.6%	48.6%
Winter	Single income	Single	C	With	-43.3%	-26.7%	-1.7%	22.3%	55.2%	52.2%
Winter	Single income	Single	D	With	-43.5%	-26.7%	-1.3%	23.6%	56.1%	51.5%
Winter	Single income	Single	E	With	-44.1%	-27.1%	-2.6%	21.5%	53.5%	52.8%
Winter	Single income	Single	F	With	-43.2%	-26.6%	-1.8%	21.7%	55.2%	52.5%
Winter	Single income	Dual	A	With	-49.6%	-28.4%	-4.8%	4.1%	48.1%	68.7%
Winter	Single income	Dual	B	With	-48.8%	-26.1%	-1.5%	7.4%	51.3%	63.9%
Winter	Single income	Dual	C	With	-49.8%	-28.9%	-5.7%	3.1%	45.9%	70.2%
Winter	Single income	Dual	D	With	-48.6%	-27.2%	-2.6%	6.9%	53.4%	64.9%
Winter	Single income	Dual	E	With	-51.1%	-29.5%	-6.0%	2.9%	45.8%	70.5%
Winter	Single income	Dual	F	With	-49.9%	-28.9%	-5.4%	3.1%	45.8%	70.1%
Winter	Low income	Single	A	With	-46.3%	-32.9%	-1.9%	10.4%	52.3%	63.0%
Winter	Low income	Single	B	With	-45.0%	-30.5%	2.0%	14.8%	59.3%	58.7%
Winter	Low income	Single	C	With	-46.2%	-33.0%	-1.8%	10.9%	52.7%	62.5%
Winter	Low income	Single	D	With	-46.0%	-32.6%	-1.4%	11.0%	53.2%	62.4%
Winter	Low income	Single	E	With	-46.8%	-33.5%	-3.1%	9.3%	49.6%	64.2%
Winter	Low income	Single	F	With	-46.3%	-33.1%	-2.0%	10.5%	51.9%	62.9%
Winter	Low income	Dual	A	With	-61.4%	-28.7%	-5.8%	24.3%	108.4%	52.1%
Winter	Low income	Dual	B	With	-60.2%	-25.7%	-2.8%	27.7%	114.7%	49.0%
Winter	Low income	Dual	C	With	-61.6%	-29.1%	-6.5%	23.3%	105.8%	52.8%
Winter	Low income	Dual	D	With	-61.1%	-27.7%	-3.0%	29.2%	118.4%	49.4%
Winter	Low income	Dual	E	With	-61.9%	-29.6%	-6.5%	23.9%	108.4%	52.7%
Winter	Low income	Dual	F	With	-61.6%	-28.9%	-6.4%	23.1%	105.2%	52.8%
Winter	Regional	Single	A	With	-60.7%	-35.0%				62.5%
Winter	Regional	Single	B	With	-59.2%	-32.0%	-1.5%	15.9%	76.5%	58.4%
Winter	Regional	Single	C	With	-60.7%	-35.3%	-5.9%	10.8%	68.7%	63.3%
Winter	Regional	Single	D	With	-60.7%	-35.0%	-4.2%	13.4%	72.5%	61.1%
Winter	Regional	Single	E	With	-61.0%	-36.3%	-7.1%	9.4%	66.8%	64.7%
Winter	Regional	Single	F	With	-60.7%	-35.0%	-5.7%	10.9%	68.8%	63.1%
Winter	Regional	Dual	A	With	-58.5%	-36.5%	-5.0%	16.1%	80.7%	59.7%
Winter	Regional	Dual	B	With	-57.1%	-34.2%	-1.5%	20.4%	87.7%	56.3%
Winter	Regional	Dual	C	With	-58.8%	-37.1%	-6.0%	14.8%	78.5%	60.7%
Winter	Regional	Dual	D	With	-57.7%	-35.0%	-2.6%	19.1%	86.0%	57.3%
Winter	Regional	Dual	E	With	-59.0%	-37.3%	-6.2%	14.7%	78.2%	60.8%
Winter	Regional	Dual	F	With	-58.7%	-36.9%	-5.7%	15.0%	79.2%	60.6%
Winter	Low net worth	Single	A	With	-36.0%	-17.0%	-1.9%	12.4%	38.4%	53.9%
Winter	Low net worth	Single	B	With	-33.7%	-13.4%	2.4%	17.6%	44.7%	46.6%
Winter	Low net worth	Single	C	With	-36.3%	-17.3%	-2.3%	12.1%	37.8%	54.4%
Winter	Low net worth	Single	D	With	-35.8%	-16.9%	-1.6%	12.5%	39.1%	53.7%
Winter	Low net worth	Single	E	With	-37.2%	-18.6%	-3.8%	10.2%	35.7%	57.3%
Winter	Low net worth	Single	F	With	-36.1%	-17.0%	-1.9%	12.5%	38.3%	53.8%
Winter	Low net worth	Dual	A	With	-48.2%	-28.3%	-4.5%	24.2%	76.8%	51.9%
Winter	Low net worth	Dual	B	With	-46.2%	-25.7%	-1.0%	27.9%	84.0%	49.0%
Winter	Low net worth	Dual	C	With	-49.0%	-29.0%	-5.5%	23.4%	74.4%	52.7%
Winter	Low net worth	Dual	D	With	-47.7%	-27.0%	-2.1%	28.7%	80.8%	49.2%
Winter	Low net worth	Dual	E	With	-49.6%	-29.6%	-5.7%	23.6%	73.5%	52.8%
Winter	Low net worth	Dual	F	With	-48.5%	-28.6%	-5.1%	23.1%	75.5%	52.7%
Winter	Health Care Card Holders	Single	A	With	-35.7%	-12.6%	-3.1%	25.0%	76.3%	41.8%
Winter	Health Care Card Holders	Single	B	With	-33.3%	-8.4%	1.0%	31.1%	84.8%	35.6%
Winter	Health Care Card Holders	Single	C	With	-35.7%	-12.9%	-3.5%	24.4%	75.5%	42.2%
Winter	Health Care Card Holders	Single	D	With	-35.5%	-12.9%	-2.6%	24.7%	76.9%	42.1%
Winter	Health Care Card Holders	Single	E	With	-36.7%	-14.1%	-4.6%	22.5%	73.3%	44.3%
Winter	Health Care Card Holders	Single	F	With	-35.7%	-12.6%	-3.3%	24.9%	75.8%	41.8%
Winter	Health Care Card Holders	Dual	A	With	-50.9%	-3.6%	-3.2%	35.5%	221.0%	29.6%
Winter	Health Care Card Holders	Dual	B	With	-49.9%	0.0%	0.3%	40.5%	235.6%	25.0%
Winter	Health Care Card Holders	Dual	C	With	-51.2%	-4.3%	-3.4%	34.1%	217.0%	30.6%
Winter	Health Care Card Holders	Dual	D	With	-49.9%	-1.2%	-0.9%	39.0%	227.9%	26.6%
Winter	Health Care Card Holders	Dual	E	With	-51.4%	-4.1%	-3.3%	34.1%	216.5%	30.4%
Winter	Health Care Card Holders	Dual	F	With	-51.2%	-4.1%	-3.7%	34.4%	218.7%	30.3%
Summer	All residential	Single	A	With	-45.0%	-26.5%	-0.3%	8.8%	51.5%	62.5%
Summer	All residential	Single	B	With	-40.0%	-19.2%	10.4%	20.8%	68.6%	49.0%
Summer	All residential	Single	C	With	-44.7%	-26.1%	0.4%	9.6%	52.7%	61.5%
Summer	All residential	Single	D	With	-44.6%	-26.0%	0.3%	9.2%	52.3%	61.9%
Summer	All residential	Single	E	With	-45.0%	-26.9%	-0.9%	8.0%	50.2%	63.5%
Summer	All residential	Single	F	With	-44.9%	-26.4%	-0.1%	9.0%	51.8%	62.3%
Summer	All residential	Dual	A	With	-46.6%	-27.3%	-2.2%	16.4%	73.5%	56.2%
Summer	All residential	Dual	B	With	-42.4%	-20.6%	7.0%	27.6%	92.0%	46.4%
Summer	All residential	Dual	C	With	-46.5%	-27.0%	-2.0%	16.5%	73.8%	56.0%
Summer	All residential	Dual	D	With	-45.5%	-25.6%	0.3%	19.4%	77.6%	53.4%
Summer	All residential	Dual	E	With	-46.7%	-27.3%	-2.3%	16.3%	73.0%	56.3%
Summer	All residential	Dual	F	With	-46.8%	-27.5%	-2.6%	15.8%	72.8%	56.8%
Summer	Elderly	Single	A	With	-26.7%	-21.4%	0.2%	11.6%	27.1%	57.4%
Summer	Elderly	Single	B	With	-21.2%	-15.7%	9.4%	23.6%	40.2%	45.0%
Summer	Elderly	Single	C	With	-26.0%	-20.7%	0.9%	12.2%	28.1%	56.5%
Summer	Elderly	Single	D	With	-26.1%	-20.5%	0.5%	11.2%	27.2%	57.4%
Summer	Elderly	Single	E	With	-26.8%	-21.5%	-1.0%	9.8%	25.2%	59.3%
Summer	Elderly	Single	F	With	-26.6%	-21.5%	0.1%	11.3%	27.9%	57.7%
Summer	Elderly	Dual	A	With	-23.3%	-21.7%	-2.2%	15.8%	17.5%	53.9%
Summer	Elderly	Dual	B	With	-17.3%	-15.4%	5.4%	25.0%	27.2%	44.1%
Summer	Elderly	Dual	C	With	-22.9%	-21.3%	-2.2%	16.0%	17.7%	53.5%
Summer	Elderly	Dual	D	With	-22.5%	-20.8%	-0.1%	17.9%	19.4%	51.9%
Summer	Elderly	Dual	E	With	-24.0%	-22.3%	-2.9%	14.5%	16.0%	55.4%
Summer	Elderly	Dual	F	With	-23.3%	-21.6%	-2.3%	15.4%	17.2%	54.2%
Summer	Needs assistance	Single	A	With	-60.1%	-26.2%	-2.8%	6.7%	85.4%	64.8%
Summer	Needs assistance	Single	B	With	-57.2%	-19.4%	6.8%	18.1%	104.1%	50.8%
Summer	Needs assistance	Single	C	With	-60.4%	-25.8%	-2.4%	7.5%	84.5%	63.7%
Summer	Needs assistance	Single	D	With	-60.3%	-25.8%	-2.0%	7.3%	89.9%	64.0%
Summer	Needs assistance	Single	E	With	-60.5%	-26.4%	-3.3%	6.2%	84.3%	65.5%
Summer	Needs assistance	Single	F	With	-60.0%	-26.2%	-2.8%	6.9%	86.0%	64.6%
Summer	Needs assistance	Dual	A	With	-38.1%	-36.6%	-2.9%	-1.2%	0.4%	89.9%
Summer	Needs assistance	Dual	B	With	-32.3%	-30.6%	6.6%	8.9%	10.8%	63.8%
Summer	Needs assistance	Dual	C	With	-37.7%	-36.3%	-2.8%	-0.9%	0.7%	86.2%
Summer	Needs assistance	Dual	D	With	-37.4%	-35.9%	-0.7%	0.3%	1.7%	74.6%
Summer	Needs assistance	Dual	F	With	-38.1%	-36.6%	-3.1%	-1.4%	0.3%	91.9%
Summer	Single income	Single	A	With	-43.1%	-28.6%	-0.7%	10.0%	49.9%	62.1%
Summer	Single income	Single	B	With	-38.7%	-23.0%	9.0%	21.8%	66.3%	50.6%
Summer	Single income	Single	C	With	-42.4%	-27.9%	-0.1%	10.4%	51.1%	61.4%
Summer	Single income	Single	D	With	-42.6%	-28.0%	-0.3%	10.0%	50.3%	61.9%
Summer	Single income	Single	E	With	-42.3%	-28.5%	-1.2%	9.2%	50.1%	62.8%
Summer	Single income	Single	F	With	-42.8%	-28.5%	-0.7%	10.0%	49.7%	62.1%
Summer	Single income	Dual	A	With	-35.7%	-15.8%	-2.4%	23.3%	49.4%	45.2%
Summer	Single income	Dual	B	With	-29.9%	-8.3%	5.6%	33.6%	61.1%	34.9%
Summer	Single income	Dual	C	With	-35.0%	-15.3%	-2.1%	23.7%	49.4%	44.6%
Summer	Single income	Dual	D	With	-35.3%	-14.8%	-0.1%	27.2%	55.1%	42.7%
Summer	Single income	Dual	E	With	-35.0%	-16.3%	-2.5%	24.0%	50.0%	45.2%
Summer	Single income	Dual	F	With	-35.4%	-15.9%	-2.7%	22.6%	48.1%	45.6%
Summer	Low income	Single	A	With	-38.2%	-19.4%	-2.0%	13.4%	42.8%	54.5%
Summer	Low income	Single	B	With	-34.6%	-13.5%	5.9%	23.3%	56.5%	43.3%
Summer	Low income	Single	C	With	-38.1%	-18.6%	-1.5%	14.1%	43.3%	53.4%
Summer	Low income	Single	D	With	-37.9%	-18.6%	-1.1%	14.4%	43.8%	53.2%
Summer	Low income	Single	E	With	-38.8%	-19.3%	-2.6%	12.7%	41.0%	55.2%
Summer	Low income	Single	F	With	-38.3%	-19.4%	-2.1%	13.4%	43.1%	54.6%
Summer	Low income	Dual	A	With	-48.7%	-29.4%	-3.4%	12.7%	44.7%	59.9%
Summer	Low income	Dual	B	With	-44.8%	-23.4%	4.8%	21.9%	56.3%	50.9%
Summer	Low income	Dual	C	With	-48.3%	-29.4%	-3.4%	13.0%	44.3%	59.7%
Summer	Low income	Dual	D	With	-48.0%	-28.2%	-0.8%	16.5%	50.5%	56.5%
Summer	Low income	Dual	E	With	-48.3%	-30.0%	-3.5%	13.3%	45.3%	59.7%
Summer	Low income	Dual	F	With	-48.7%	-29.7%	-3.9%	11.9%	43.2%	60.7%
Summer	Regional	Single	A	With	-61.9%	-38.4%	-3.8%	15.5%	62.5%	60.6%
Summer	Regional	Single	B	With	-58.7%	-33.0%	6.1%	27.1%	77.7%	52.5%
Summer	Regional	Single	C	With	-61.7%	-37.7%	-3.4%	16.0%	64.6%	60.1%
Summer	Regional	Single	D	With	-61.8%	-37.4%	-2.6%	17.0%	66.7%	59.4%
Summer	Regional	Single	E	With	-62.1%	-38.3%	-4.6%	14.1%	62.0%	61.6%
Summer	Regional	Single	F	With	-61.9%	-38.3%	-4.0%	15.1%	62.8%	60.9%
Summer	Regional	Dual	A	With	-57.7%	-36.3%	-3.2%	8.8%	89.5%	65.2%
Summer	Regional	Dual	B	With	-54.3%	-30.3%	6.4%	19.6%	110.3%	55.4%
Summer	Regional	Dual	C	With	-57.6%	-36.1%	-2.9%	9.1%	89.8%	64.9%
Summer	Regional	Dual	D	With	-57.1%	-34.9%	-0.8%	11.5%	94.1%	62.6%
Summer	Regional	Dual	E	With	-57.9%	-36.4%	-3.3%	8.7%	88.4%	65.3%
Summer	Regional	Dual	F	With	-57.8%	-36.5%	-3.6%	8.3%	88.4%	65.7%
Summer	Low net worth	Single	A	With	-41.2%	-22.3%	0.5%	4.7%	33.5%	66.4%
Summer	Low net worth	Single	B	With	-35.3%	-14.0%	11.8%	16.5%	48.4%	47.9%
Summer	Low net worth	Single	C	With	-40.8%	-21.6%	1.5%	5.7%	34.8%	64.6%
Summer	Low net worth	Single	D	With	-41.3%	-22.1%	0.7%	4.7%	34.0%	66.3%
Summer	Low net worth	Single	E	With	-41.3%	-22.5%	0.0%	4.0%	32.6%	67.4%
Summer	Low net worth	Single	F	With	-40.9%	-22.0%	1.0%	5.0%	34.0%	65.7%
Summer	Low net worth	Dual	A	With	-39.7%	-19.5%	-2.2%	23.7%	62.9%	47.6%
Summer	Low net worth	Dual	B	With	-33.5%	-11.0%	7.3%	35.7%	78.1%	36.8%
Summer	Low net worth	Dual	C	With	-39.1%	-18.9%	-1.8%	23.7%	64.4%	47.2%
Summer	Low net worth	Dual	D	With	-39.4%	-18.8%	0.0%	26.7%	68.3%	45.7%
Summer	Low net worth	Dual	E	With	-39.4%	-19.9%	-2.0%	23.6%	65.2%	47.9%
Summer	Low net worth	Dual	F	With	-39.6%	-19.4%	-2.5%	22.9%	62.3%	47.9%
Summer	Health Care Card Holders	Single	A	With	-31.3%	-13.3%	-1.8%	19.4%	68.1%	45.3%
Summer	Health Care Card Holders	Single	B	With	-25.1%	-4.7%	8.1%	32.4%	86.6%	31.3%
Summer	Health Care Card Holders	Single	C	With	-30.8%	-12.6%	-1.2%	20.5%	70.0%	44.0%
Summer	Health Care Card Holders	Single	D	With	-31.1%	-12.9%	-1.2%	19.8%	69.2%	44.7%
Summer	Health Care Card Holders	Single	E	With	-31.8%	-13.9%	-2.3%	18.6%	67.3%	46.4%
Summer	Health Care Card Holders	Single	F	With	-31.2%	-13.2%	-1.7%	19.6%	68.3%	45.1%
Summer	Health Care Card Holders	Dual	D	With	-52.5%	-28.5%	-1.2%	-1.2%	54.4%	75.4%
Summer	Health Care Card Holders	Dual	E	With	-53.7%	-30.4%	-4.4%	-4.2%	49.7%	76.5%
Annual	Non-vulnerable	Single	A	Without	-23%	-42%	1%	16%	64%	56%
Annual	Non-vulnerable	Single	B	Without	-23%	-42%	1%	17%	64%	42%
Annual	Non-vulnerable	Single	C	Without	-23%	-42%	1%	17%	64%	55%
Annual	Non-vulnerable	Single	D	Without	-23%	-43%	0%	15%	59%	55%
Annual	Non-vulnerable	Single	E	Without	-23%	-43%	0%	16%	64%	57%
Annual	Non-vulnerable	Single	F	Without	-23%	-42%	1%	17%	64%	55%
Annual	Non-vulnerable	Dual	A	Without	-28%	-47%	-2%	16%	64%	60%
Annual	Non-vulnerable	Dual	B	Without	-28%	-47%	-3%	15%	63%	48%
Annual	Non-vulnerable	Dual	C	Without	-28%	-47%	-3%	15%	63%	59%
Annual	Non-vulnerable	Dual	D	Without	-27%	-47%	-1%	18%	67%	57%
Annual	Non-vulnerable	Dual	E	Without	-28%	-47%	-2%	16%	64%	60%
Annual	Non-vulnerable	Dual	F	Without	-28%	-48%	-3%	15%	63%	60%
Annual	Non-vulnerable	Single	A	With	-24%	-43%	-1%	14%	60%	56%
Annual	Non-vulnerable	Single	B	With	-23%	-42%	0%	16%	63%	42%
Annual	Non-vulnerable	Single	C	With	-24%	-43%	-1%	14%	60%	55%
Annual	Non-vulnerable	Single	D	With	-24%	-43%	-1%	14%	57%	55%
Annual	Non-vulnerable	Single	E	With	-25%	-44%	-3%	12%	57%	57%
Annual	Non-vulnerable	Single	F	With	-24%	-43%	-1%	14%	60%	55%
Annual	Non-vulnerable	Dual	A	With	-29%	-48%	-4%	14%	61%	60%
Annual	Non-vulnerable	Dual	B	With	-28%	-48%	-3%	15%	63%	48%
Annual	Non-vulnerable	Dual	C	With	-29%	-48%	-4%	13%	59%	59%
Annual	Non-vulnerable	Dual	D	With	-27%	-47%	-1%	17%	65%	57%
Annual	Non-vulnerable	Dual	E	With	-29%	-49%	-5%	13%	59%	60%
Annual	Non-vulnerable	Dual	F	With	-29%	-48%	-4%	13%	59%	60%