Chapter 5: Household Affordability Danish Environmental Protection Agency

Water Prices in CEE and CIS Countries. Volume I: Main Text

Chapter 5
Household Affordability

This chapter and the preceding chapter have a joint purpose namely to provide guidance as to how to assess the demand for water and wastewater services. The demand for services is a function of relative prices, income and preferences of the customers.

The present chapter provides guidance as to methods to carry out an analysis of the affordability in relation to water services.

5.1.2 Target group

The methods and tools presented in the current chapter are relevant for cities in the CEE and in the CIS countries. The methods and tools could also (with minor changes) be used in large cities in developing countries. Parts of the chapter target experts within water utility planning, whereas other parts target experts within market research. In addition, the chapter allows other stakeholders to obtain a grasp of the methods presented and to assess the level of analysis needed for a specific case.

5.1.3 The notion of affordability

The affordability is closely linked to the willingness to pay. However, the customers' own perception of the water cost and service does not influence the analysis in this chapter. Whereas the willingness to pay survey gives information on whether the households are prepared to pay more for improved service or to avoid a worsening of the service, the affordability analysis gives information on the ability to pay. The data collected on affordability is "objective", but it has to be subjectively interpreted by the consultant. This differs from the willingness to pay assessments that are based on the consumers' own subjective assessments of utility and budget constraint.

Society can be seen as the sum of households, enterprises and public institutions. The affordability issues discussed in this chapter concern household affordability.

The notion of affordability in households is related to the upper limit of expenditure on water and wastewater services. In the CEE region, and the better-off cities of the CIS, the percentage of income spent on water and wastewater services should not imply that water service becomes a major post on the household budget. This would indicate over-investment in these services. Furthermore, such charges may result in an expenditure level that is not affordable for lower income families and may lead to lower collection rates.

An upper limit of expenditure on water and wastewater for a household cannot be objectively established. Therefore, it is necessary to apply tools to obtain an overview of the average income and the distribution of income, and hence provide indicators as to when there is a problem of affordability that needs to be addressed. Such indicators must be derived from income, expenditure patterns and water consumption.

A household is assumed to be unable to pay if it cannot pay the water and wastewater bills without having to cut down significantly on basic needs, such as food, and other public services, such as heating.

Another definition of ability to pay could be related to the behaviour if non-paying customers were cut off. In this strict sense, few customers would have affordability problems, as water is a necessity good. However, the latter definition is relevant only after the investment decision has been made, and affordability issues cannot influence the level of service provided.

The result of the affordability analysis should influence the target service and the tariff level. Furthermore, it can provide input for the design of subsidy schemes to poor households. The analysis might influence the political acceptability, especially when poor households constitute a major rating block, or if their plight is of concern to the electorate.

Tools are outlined in order to:

Assess the household's ability to pay for improved service.

Identify the proportion of households that are likely to have difficulties in paying for improved service.

An overall assessment of household affordability can be based on macro - economic data on average income, expenditure for water services and food expenditure as a share of the total expenditure. In addition, a rough estimation of the income distribution is needed. Macro-economic income data in the CEE and CIS countries is often subject to uncertainty. Therefore, the reliability of the assessment depends on the quality of available data. The collection rates in relation to water and wastewater bills (and in particular recent changes herein) may also provide an indication of affordability. In both cases the data can give a first indication as to whether a group of inhabitants will have difficulties in paying the water bill.

If this is the case, further analysis at household level is recommended in order to obtain detailed information on the nature and size of the problem. This can be done using detailed household data on income and expenditures.

For enterprises and public institutions, consideration of the affordability in relation to water tariffs is not helpful. The key issues here relate to options for substitution of public service provision and to cross-subsidisation of household consumers. A critical review is needed of tariff structures which imply higher tariffs for enterprises than for households with no reference to the higher costs associated with servicing enterprises. However, we shall not deal further with these issues in this toolkit.

Public institutions have their expenses covered by public budgets. Budget constraints may imply that public institutions are not able to pay for water and wastewater services. This raises a number of issues related to the allocation of public funds to public institutions and to the enforcement of payments from these institutions. However, we shall not deal further with these issues.

5.2 Suggested approach, timing and resources

5.2.1 Approach

We suggest carrying out the analysis in three steps, as illustrated in Text Box 20 below. Each step is described in a subsection.

Text Box 20:
The recommended approach for analysing household affordability

Affordability analysis using macro-economic data (Section 5.3)

	water and wastewater expenditure in terms of share of average household income (Section 5.3.1)
	food expenditure as a share of total expenditure/disposable income (Section 5.3.2)
	distribution of income (Section 5.3.3)
	simultaneity of price increases (food, heat, electricity etc.) (Section 5.3.4)
	changes in collection rates over time (Section 5.3.5)

Affordability analysis using household data (Section 5.4)

	analysis of income and expenditure data (Section 5.4.1)
	collection of income and expenditure data when unavailable (Section 5.4.2)

Tariff design and financial transfers (Section 5.5)

	tariff base (Section 5.5.1)
	eligibility for subsidies (Section 5.5.2)
	budget effects (Section 5.5.3)

We suggest that the analysis regarding affordability using macro-economic data, Section 5.3, should be carried out first. If the analysis indicates that a group of households have an affordability problem, the more detailed analysis at household level, Section 5.4, will be carried out.

Section 5.5 gives ideas as to tariff design and how transfers can be made to households unable to pay for the water and wastewater services.

5.2.2 Timing

Affordability ought to be an important factor in determining the level of the technical solution, especially when it is likely that households' costs rise above an unacceptable level. Therefore, optimal timing of the affordability analysis is early in the planning process. This applies whether a traditional investment project in a public utility or a concession agreement is planned.

For a traditional investment project in a public utility, part of the affordability analysis should be undertaken during the pre-feasibility phase. In this phase, overall estimates should be produced of the share that the cost of water constitutes of the household income, whereas a detailed analysis of the income distribution and expenditure can be carried out along with the willingness to pay analysis.

For a PPP arrangement, the optimal timing is likely to be prior to or simultaneous with the announcement of the selected short list of firms invited to submit a bid for the lease, concession etc. The short-listed firms will find it useful to have an indication of affordability in order to assess the tariff that is likely to be acceptable at a given future service level.

Furthermore, the municipality may find the affordability analysis useful in order to anticipate subsidies to poor households.

5.2.3 Resources required

The resources required are given in the table below. If no household level data is available on income or expenditure, a separate survey should be carried out in order to make the affordability analysis at household level.

Table 5.1:
Resources required for the affordability tasks

Person Weeks

Calendar Weeks

Affordability in society in general,

Household affordability

1-4

2-4

If register data or survey data is available

If expenditure survey is to be carried out

Tariff design and financial transfers

Total

3-7

6-8

Disclaimer: The resource requirement has been estimated by COWI and is not necessarily endorsed by DEPA or EBRD

If the household expenditure survey is needed, the affordability analysis is expected to require approximately seven person weeks and eight calendar weeks.

If data on household income or expenditure data is already available, the affordability analysis can be carried out with an input of approximately three person weeks and over a period of six weeks.

5.3 Affordability analysis using macro-economic data

This section suggests an affordability analysis that can be made when micro-economic data on households is not readily available¹⁹. Macro-economic data available from national statistical sources and from international databases.

Based on macro-economic data, we suggest the following analyses:

Data on average income is used to investigate whether an average household can afford the expenditure that results from the investment in improved services. An example of such an analysis is provided in Section 5.3.1. The composition of household expenditure can give an insight into whether the households will have difficulty in paying the water bill. Section 5.3.2 discusses how to do this.

Income is often distributed unequally among the households in society. Affordability can be a severe concern for segments of the population also when there is no affordability issue for the average household. In order to evaluate the proportion of households that may experience difficulties in paying for the services, the distribution of income is of significant interest. Income distribution is discussed in Section 5.3.3.

Further, it is important to include simultaneous price increases for water and other goods and services in the analysis. If the price of heating or electricity goes up in the same period of time as an increase in the cost of water and wastewater services is planned, this is likely to influence the affordability negatively, see Section 5.3.4.

Finally, changes in the collection rate can be used as a very rough measure of affordability, as a high collection rate indicates that customers, all things equal, are able to pay, see Section 5.3.5.

5.3.1 Cost and average income

An average measure of the cost of future water is useful in order to obtain an overall impression of whether the society can afford the planned investment in improved water and wastewater services. The cost of water and wastewater services as a percentage of the average household income or the cost of water and wastewater services as a percentage of the GDP per capita constitutes useful measures.

The maximum affordable level of combined water and wastewater services is often based on the rule of thumb that the average payment for these services should not exceed 4% of the average household income. This is as a first indicator of affordability.

When analysing an investment scenario, the future share of household income needed to pay for water services should be considered. This is done by using an official forecast of the income per capita and the expected increase in tariffs in different scenarios. In this analysis, it is suggested that the tariff depends on the level of investment, only, and not on macro-economic fluctuations affecting the loan. This means that it is assumed that the loan is secured in such a way that the increases in tariffs can be forecasted without risk.

The analysis should be made for each scenario considered, i.e. for each level of the tariff. Text Box 21 gives an example from Kaliningrad. In the example, the project has already been chosen. Therefore, there is only one scenario.

The rule of thumb mentioned above cannot stand alone. While an average exceeding 4% indicates potential affordability problems, a low share of water cost is not sufficient to conclude that affordability problems will not occur. The reason is that the income variation may be large, thus making it difficult for some households to pay the water and wastewater bills. Furthermore, there may be simultaneous price increases in other household expenditure such as electricity and heating. In this case, even the 4% threshold may represent an unacceptable burden.

Text Box 21:
The cost of water and wastewater in the investment period

The tariff has to increase in Kaliningrad in order to cover the planned improvements of the water and wastewater services.

At the end of the investment period, the water and wastewater costs will amount to more than 4% of the inhabitants' income. This may limit the ability to pay of a large group of people.

Expected increase in water tariffs and income as well as the average cost of water services as a percentage of real income

	2000	2001	2002	2003	2004	2005
Increase in the real water tariff	33%	20%	40%	30%	6%	6%
Increase in the real income	3%	3%	3%	3%	3%	3%
Water cost as a share of income	1.9%	2.3%	3.1%	3.9%	4.0%	4.1%

Note: The increases in real water tariff are as proposed in Krüger (1999): Kaliningrad water and wastewater services feasibility study.
Sources: Krüger and the Ministry of Finance of the Russian Federation.

5.3.2 Composition of expenditure

The composition of expenditure provides important information on affordability. The World Bank has suggested that households that use 70% of their disposable income on food are "absolutely poor". Hence, if the average spending on food approximates 70%, there is likely to be a problem of affordability.

Therefore, the composition of household expenditure provides a useful measure of the households' wealth, cf. Text Box 22. In the text box example, total household expenditure has been used instead of disposable income. In a society where there is almost no saving or dissaving, there is little difference between expenditure and disposable income. Thus both measures can be used.

The distribution of expenditure may differ significantly from one region to another because of institutional differences. This can make comparability across regions difficult. For instance rent constitutes a very small part of the household budget in some CIS countries.

Some literature suggests using the share of expenditure for tobacco and alcoholic beverages as a minimum level of acceptable expenditure for water and wastewater services. This measure requires access to a reliable household expenditure survey, and only provides a rough assessment.

Text Box 22:
Composition of expenditure in Kaliningrad, 1991 - 98

The average expenditure on food has increased from 29% of the total expenditure in 1991 to 54% in 1998.

This change in the expenditure composition clearly indicates that the citizens of Kaliningrad have become substantially poorer since 1991. A share of food expenditure amounting to 54% of the total expenditure also indicates that there is not much room for increasing charges for services. This is exacerbated by the expected future price increases for rent and other utilities.

Composition of expenditure in Kaliningrad, 1991 - 98

Note: Non-food includes alcohol and meals outside home
Source: Kaliningrad Statistical Committee.

5.3.3 Income distribution

Great variation in income distribution makes the average income a poor proxy for representative household income. Therefore, the analysis of income and expenditure should be supplemented with measures that take income distribution explicitly into account.

If income distribution data is not available, a summary measure of inequality can be applied. One such measure is the Gini coefficient which can be calculated using income deciles²⁰. The Gini coefficient can be used to evaluate whether the income distribution is unequal compared to other areas. However, this is a very indirect and poor measure of affordability for the lower income groups.

The literature takes income distribution explicitly into account in different ways. Some literature suggests relating water and wastewater expenditure to the income of households at subsistence level. Usually, there is an official measure of per capita subsistence income, e.g. in connection with setting the minimum pension. However, there is no consensus as to what is the appropriate share of income to be used for water and wastewater for households at subsistence level.

If income distribution data is available, the share of income used to pay as an average bill for water and wastewater services can be estimated for each income group. This will give an indication of the proportion of households that spend more than a certain percentage of their total income on water and wastewater services. This is, however, only an indicative measure, as a large proportion of the poorest households are eligible for subsidies.

Furthermore, if consumption is metered, poorer households can, and generally do reduce their water bill. In this way, the households themselves mitigate the affordability issue.

Text Box 23:
Income distribution in Kaliningrad

The average quarterly disposable income per household member is approximately 3,000 roubles.

The equivalent average cost of water services in 1999 was approximately 40 roubles.

For 16% of the inhabitants with an income of less than 1,000 roubles, the cost of water and wastewater services exceeds 4%, at present. This group might have difficulties in paying higher water tariffs without compensation.

Income distribution in Kaliningrad, third quarter 1999

Note: Data is from the third quarter of 1999.
Source: Kaliningrad Statistical Committee.

5.3.4 Simultaneity of price increases

When designing the future scenarios of services and tariffs, it is useful to have information about the expected development in the price of other goods and services.

Municipal authorities should be aware of the total impact on households of simultaneous price increases on various public services such as water services, heating, electricity etc. Thus an assessment of the maximum tariff for water and wastewater services must, therefore, take into account the implications of other price increases. A calculation of residual income after paying for the public services may contribute to the assessment.

An example of this analysis from Sevastopol, the Ukraine is given in Text Box 24.

Text Box 24:
Simultaneous price increases in Sevastopol, the Ukraine

The real income per capita is estimated to increase by 18% over the period 2000-2005 based on real salary projections provided by the EBRD. This implies that the real income per capita in 2005 will be 23 UAH/month higher than the present level.

The incremental expenditure per capita in connection with the heating and water projects is estimated at 11 UAH/month in 2005. Implementing the two projects concurrently implies that households will have to reserve almost 50% of the expected increase in real income in 2005 for increased expenditure on heating and water services.

Though this indicates that it is feasible in affordability terms to implement the two projects concurrently, close co-ordination of the implementation of the projects is required.

5.3.5 Collection rates

Sometimes the change in collection rates following former price increases can be used as a rough measure of ability to pay. If there is a significant increase in the collection rate, this could indicate that there are no affordability problems, whereas no increase in the collection rate could indicate the opposite.

The collection rate is a very indirect measure of affordability compared to income and expenditure. The collection rate depends on the collection procedure and whether non-paying customers are cut off. If non-payment does not have consequences for the non-payers, the collection rate is likely to be low.

The change in collection rate can only be used as a proxy for affordability if there is no simultaneous change in the institutional framework for billing and collection.

Text Box 25:
Collection rates in Kaliningrad

Approximately 65% of the Kaliningrad population has its water tariff collected by the municipal billing company Simplex.

After the tariff was increased by 150 per cent in July 1999, the collection rate declined from almost 80% to less than 50%. This indicates that a large number of the households got affordability problems after the tariff increase.

5.4 Affordability analysis using household data

If data on household expenditure is available or an expenditure survey is conducted for the purpose, it is possible to obtain better insight in to the households' financial situation.

If such micro-economic data is available, useful measures are:

Distribution of water expenditure as a percentage of income.

Share of income used for food.

Distribution of water cost as a percentage of income after expenditure on food.

With these measures it is possible to assess the proportion of households that will experience problems with regard to their ability to pay as the services improve and the tariffs increase.

This section describes the above measures as well as a method to conduct an expenditure survey, in the case where a detailed analysis of household expenditure is required and no data is readily available.

5.4.1 Income distribution and expenditure patterns

Water expenditure as a share of income

With detailed data on household income and household expenditure on water and wastewater services, it is possible to analyse the distribution of water expenditure across households.

This can be done both in the situation before and after new investments have been made. By choosing an appropriate assumption on income growth and tariffs, illustrative scenarios of the future situation can be made.

Text Box 26 illustrates the expenditure for water and wastewater services as a percentage of household income in 1999 and the possible distribution in 2005 with likely increases in the water tariff and income.

Text Box 26:
Cost of water and wastewater services as a percentage of income, Kaliningrad

At present, most households spend less than 2% of their income on water and wastewater services. Only 5% of the households pay more than 4% which indicates that the affordability problem is very limited, at present.

For the distribution in 2005, it is assumed that all households have a yearly increase in real income of 3%, and that the water tariff increases by approximately 200%. These increases are equivalent to the assumptions in Text Box 21.

With the assumed changes for 2005, a large proportion of the households will spend more than 5% of the income on water and wastewater services which indicates that there might be a problem of affordability.

Sources: Kaliningrad Statistical Committee, Krüger and the Ministry of Finance of the Russian Federation.

Food expenditure

The composition of expenditure provides important information on wealth, cf. Section 5.3.2. With data at household level, it is possible to analyse the proportion of households in which food expenditure dominates the budget.

If a large proportion of the households have food expenditure that exceeds 70% of the disposal income, an affordability problem is likely to occur. Text Box 27 illustrates the situation in Kaliningrad.

Text Box 27:
Food expenditures as a share of income, Kaliningrad

In Kaliningrad, 20% of the households have expenditure on food that exceed 70% of their disposable income. This indicates that 20% of the households are absolutely poor, and that these households might have difficulties in paying the water and wastewater bills when the tariffs increase.

Food expenditure as a percentage of household income

Sources: Kaliningrad Statistical Committee.

Water cost as a share of the disposable income after expenditure on food

If a large proportion of the income is used for food, it can be difficult for the households to save money on the budget in order to pay higher water tariffs. Therefore, a relevant measure is the cost of water as a share of the household income not used for food. To ease the terminology, this is called the residual income.

This gives an indication of whether households are able to save money to pay higher tariffs for water without cutting in their budget for food. A share of more than 10% indicates that the household is likely to have affordability problems.

Text Box 28:
Cost of water and available income after food expenditure, Kaliningrad

The amount spent on water and wastewater services can be compared to the income after deduction of food expenditure, hereafter called the residual income.

For most households in 1999, the share of water cost, as a percentage of the income not used for food, was less than 10%. For the vast majority, this share was lower than 4%.

In 2005, when the increases in tariffs will have been introduced, almost one third of the households will spend more than 10% of their residual income on water and wastewater services.

Water and wastewater cost as a percentage of residual household income (income available after food expenditure)

Note: Residual income is defined as the income available after expenditure on food
Sources: Kaliningrad Statistical Committee, Krüger, Ministry of Finance, Russian Federation, COWI

5.4.2 Survey on income and expenditure when data is not available

In cases where up-to-date and reliable income and household expenditure data is not available, it is suggested that a household expenditure survey be carried out. An example of a questionnaire to be used for a brief expenditure survey is given in Appendix 8.

The informal economy often plays a substantial role and makes it difficult, if not impossible, to get trustworthy data on income through a survey. To circumvent this problem, one can ask for household expenditure levels for a broad range of items.

Although this method does not provide for the entire household income/expenditure level, it should give a reasonably trustworthy lower limit on the level of income. Furthermore, this approach also reveals the share of each expenditure item out of the total expenditure, which makes it feasible to give an estimate of the affordability related to an increase in the water bill.

Text Box 29:
Expenditure Survey Lviv, the Ukraine 1996

There are several reasons why the collection of income data in the Ukraine was difficult. First, a new taxation scheme was implemented and, therefore, the citizens were unwilling to reveal their income. Second, there are no reliable income statistics that include the informal part of the economy.

To circumvent this problem, an expenditure survey was conducted among 500 respondents. The survey was conducted as part of a telephone survey on water and wastewater services. The expenses included were the following:

rent, heating, electricity, water and wastewater, food, beverages and tobacco, clothes, transport, telephone, newspapers, magazines, entertainment, and other.

Furthermore, the respondents were asked if they had been buying durable goods, or if they had been travelling during the past year. This included:

durable consumer goods, furniture, cars;

travelling in the Ukraine or, if outside, the country of destination

In general, the respondents were less willing to answer the questions in the expenditure survey than those in the rest of the questionnaire. 52% at least specified the monthly expenses on food, rent and heating, and those respondents were used as a basis for making a rough estimate of the average income and income distribution.

Comparing the result of the expenditure survey with official statistics on income indicates that the estimated income of the respondents in the survey was 20-30% higher.

Based on experience, a survey size of approximately 500 persons is likely to be required. The expenditure survey sample must be larger than the sample for the stated preference questionnaire due to the greater reluctance of people to respond to the expenditure survey.

Expenditure data collected in this way should be complemented by official statistics on average wage levels and official income data. These provide a consistency check between the household expenditure survey and official data.

5.5 Tariff design and financial transfers

Two issues related to tariff design and subsidies are discussed in this section: the tariff base and its structure.

First, it should be analysed whether the tariffs should be based on the actual use of water, or whether they should be based on norms. Second, it should be investigated whether there should be an equal tariff for all inhabitants, or if certain groups should be eligible for subsidies or privileges. The latter issue is relevant in most countries where certain groups benefit from subsidies. Subsidies should be designed to help minimise the affordability problem. If an increase in tariffs influences the need for financing from other parties than the consumers (e.g. tax payers), the consequences should be considered.

This section lists the issues that are recommended in order to analyse the existing design, and a brief discussion of which issues to consider if changes in the tariff structure are anticipated.

This section is closely related to the tariff base and structure in CIS countries, and mainly relevant for these countries.

5.5.1 Tariff base

The tariff can either be related to the actual use of water or not. Basing the payment on the actual water consumption gives an incentive to economise on the amount of water used. However, it requires that meters are installed in all households, which might be too expensive compared to the benefits received.

The household tariff can also be based on the number of household members or norms such as kitchen hardware or number of square meters of the residence. If the measure is not related to water consumption, even indirectly, there is no incentive to reduce water consumption.

Text Box 30:
Tariff base

	Kaliningrad, Russia: Households are billed according to the number of household members.
	Brno, the Czech Republic: Households are billed according to metered water consumption.
	Poznan, Poland: Some households are billed according to metered consumption, other households according to a norm.

5.5.2 Eligibility for subsidies

In some of the CIS countries, there are two types of reduction in tariffs: Privileges and subsidies. Privileges are often given in lieu of wages or pension payments, e.g. to veterans or citizens holding an official position, while subsidies are given to households who have difficulties in paying their bills.

Text Box 31:
Eligibility for subsidies in Lviv, the Ukraine

In Lviv, there are two types of reductions obtainable: privileges and subsidies.

In the expenditure survey described above, 25% of the respondents received privileges. Most of those respondents received their privilege because of their social status and not because they were unable to pay themselves.

The criterion of eligibility is that the payments for water, heating and electricity must exceed 15% of the household income. The household survey indicated that almost half of the respondents interviewed were eligible for a subsidy, whereas only 8% of the respondents indicated that they receive a subsidy. The reason for the discrepancy could be that the application procedure is cumbersome.

The household survey also indicated that the income distribution among respondents was similar in the group who received privileges and the group who did not. This illustrates that privileges are not closely linked to social need.

Special privileges imply that groups of households do not pay a normal price for water services. If a large proportion of the households receive privileges, or subsidies are not motivated by affordability problems, the tariff structure is likely to be inefficient. Such structure should therefore be revised critically. A suspension of privileges would ease the financing of subsidies directed at households with affordability problems. However, this is often a difficult, national political issue.

5.5.3 Budget effects

The effect on the public budget of investments in improved water and wastewater services should be considered. There are both direct effects of the contribution to the investment and possibly operation expenditure and indirect effects of the impact on privileges and subsidies that result from increased tariffs.

It is recommended to consider carefully the impact on tariff increases including increases in the price of other utilities, and on subsidies due in the future.

Currently it is often only a small share of those eligible for subsidies who apply for them. However, if tariffs increase, the propensity to apply for subsidies paid out may therefore be considerable. It is recommended to calculate alternative scenarios for the propensity to apply for subsidies an to calculate the effect on the municipal budget in these scenarios.

¹⁹	For example, the World Bank has some useful data available: http://www.worldbank.org/data/countrydata/aag.htm
²⁰	Cf. e.g. Deaton, A. (1997): The analysis of household surveys: A micro econometric approach to development policy. The Hopkins University Press. Software for calculation of inequality measures is found on http://www.worldbank.org/lsms/tools/povcal/index.htm.