# 6. Commentaries

## 6.1 Appendix A Equity analysis in transport practice

Many transport plans, strategies and policies articulate equity as a key issue to consider in transport infrastructure investments. Selected examples from different jurisdictions are presented below.

For example, in 2003, a draft transport plan from the South Australian Government (SA Government, 2003), noted ‘transport’s contribution to social inclusion through recognition that not all South Australians fare equally and some experience acute and disproportionate disadvantage’. The following groups and issues were identified:

• Age (specifically the mobility needs of older people and the young who are especially dependent on public transport and others for transport)
• Gender (specifically people who have particular travel needs regarding access to private transport and in patterns of commuting and employment)
• People with disabilities
• Other socially or economically disadvantaged groups such as indigenous people.

Western Australia’s sustainability framework (WA Government, 2003) presented a set of criteria that could be used in the process of sustainability assessment, one of which emphasises increasing ‘access, equity and human rights in the provision of material security and effective choices’.

The NSW Government (DIPNR, 2004) has used the objectives of social equity, economic development, environmental protection and financial management to guide transport planning. In relation to transport, these objectives were described as follows:

• Social equity reflects access to jobs and services, the affordability of housing and transport, and the provision of transport choice.
• Economic development includes creating sustainable jobs, supporting exports, developing regions and minimising the cost of congestion.
• Environmental protection includes minimising the environmental impacts of transport on air, water, soils, vegetation and noise.
• Financial management includes ensuring taxpayers receive value for money from public investment, and considering inter-generational equity issues such as not overburdening future generations with excessive debts or capital requirements.

All these Australian examples emphasise the consideration of equity impacts during strategic planning and decision-making levels. There is scope to further develop and refine methodologies, tools and techniques for equity assessment of transport initiatives. Some progress has been made overseas. The Applied Research Centre in California, for example, has developed guidance for policymakers on preparing an Equity Impact Statement (ARC, 2004). The approach includes identification of the following elements:

• Communities of concern (including, for example, gender, income, disability characteristics)
• adverse effects (including social, cultural, economic, environmental, individual and cumulative effects)
• Key questions that are integrated into the policy-making process that address specific issues such as compliance with legislation, access to livelihood, quality of life and the distribution of the costs and benefits.

In Europe, the German transport investment appraisal method is both detailed and explicit in its treatment of distributional effects on different regions within Germany (Bristow and Nellthorp, 2000). A unique feature of the approach is the flexibility to assign extra weight to employment impacts to reflect specific socio-economic conditions within specific regions (Bristow and Nellthorp, 2000). These same authors also report that in Finland, distributional effects are assessed and presented as part of a Supplementary Study that is made available to the decision-maker alongside the cost-benefit results and other findings. Bristow and Nellthorp (2000) conclude that in many other EU countries there is little evidence that equity and distributional impacts are given a significant role in the assessment of proposed initiatives and the reporting of results.

## 6.2 Appendix B Techniques to estimate distributional (equity) impacts on the community

There are various quantitative and qualitative techniques for considering equity impacts on the community. Some common techniques described in this section include equity indexes and weights, social impact assessment, stated preference surveys and spatial analysis techniques.

### 6.2.1 Equity weights

A number of indexes have been developed to measure equity or inequity between groups or populations. This type of analysis is mostly used to estimate income inequity in the population; however, it is also being applied to concepts such as accessibility. The choice of which index to use will depend on the decision-makers’ needs, data availability and the level of development within the community of interest.

#### Welfare index

Loeis and Richardson (1997) identified a welfare index for use in transport analysis and evaluation. Travel demand estimation and the evaluation of travel proposals often rely on personal or household income as one of the explanatory variables. However, the financial significance of a unit of income varies from person to person and household to household. For example, at the same income level, a small household can buy more for its individual members than a larger household. Consequently, it is by itself not an adequate economic explanatory variable for travel behaviour or evaluation.

Loeis and Richardson (1997) developed their ‘Welfare Index’ through practical application of the welfare economics concept of equivalence scales, used in classifying households based on the relative cumulative needs or living costs of their members. Applied in combination with after-tax income, it rates households on relatively uniform standards of financial or welfare capacity. The result is therefore a better explanatory variable for travel behaviour than personal or household income.

##### Equity weights

The text in this section is provided as information for the reader only. In line with the advice of the Department of Finance and Administration (2006), the use of equity weights is not recommended by the ATAP (see discussion in chapter 1).

Equity weights provide a method of explicitly incorporating concepts of fairness into an economic analysis. Weights express the extent society is prepared to sacrifice efficiency in pursuit of fairness. The greater the equity weight, the more efficiency gain a society is willing to trade-off to achieve improved fairness (Sassi et al, 2001).

The underlying assumption to the development and application of such weights is that the concepts of equity and efficiency can be traded off against each other. The application of weights is thus used to effect a balancing of conflicting, but commensurable objectives when making complex resource allocation decisions (Sassi et al, 2001). An example of how to apply equity weights is provided below.

Equity weights can be derived from two major sources: the view of a (representative) sample of the population and/or the views of decision-makers (Sassi et al, 2001).

It is important to note that the application of equity weighting can be controversial. Equity weights are subjective and a detailed description of the equity effects should be provided to the decision-maker who can assess the distributional effects of an initiative. For this reason, equity weights are not often used in practice.

Box 1 How to apply equity weights

Equity weighting is a simple concept. Say that a particular initiative provides benefits for two different population groups A and B. The net benefit is given by:

NetB = Wa ΔA + Wb ΔB, where Wa and Wb are the distributional (or equity) weights.

In situations where population groups are equal, the weights are set to one. If there are equity differences between population groups (involving, for example, different income distributions) then the ‘marginal utility of income’ will be different for these two groups. The net benefit in such cases may be defined as:

NetB = WH ΔH + WL ΔL, where WH is the marginal utility of income for the high income group and WL is the marginal utility of income for the low income group.

If we say that the marginal utility of income is 0.40 for the high income group (i.e. a $1 change in income for this group results in a 0.4 change in economic welfare) and 1.25 for the low income group (i.e. a$1 rise in income causes a 1.25 change in economic welfare), then the net benefit equation becomes:

NetB = 0.4 ΔH + 1.25 ΔL ·(example taken from Sassi et al, 2001).

### 6.2.2. Social Impact Assessment of transport initiatives

Social impacts are the likely consequences for individuals or a community of implementing a particular course of action. It is common practice (and often required by legislation) to undertake a Social Impact Assessment in conjunction with an Environmental Impact Assessment in the process of evaluating major transport initiatives.

Social Impact Assessment relates to the identification and assessment of potential impacts for an area and the community of an initiative. Sinclair Knight Merz (1998) state that a Social Impact Assessment requires:

• A description of the existing and likely future social characteristics of an area
• A description of proposed changes
• An analysis of how these changes will impact on the community at both a broad (regional level) and a local level
• An examination of measures available to ameliorate adverse impacts.

Assessment of social impacts relies on community input to gain an understanding of community concerns, values and aspirations. As such, Social Impact Assessment processes and community consultation are inextricably linked (Sinclair Knight Merz, 1998).

The range of social impacts that can result from a transport initiative can be very large. The table below provides some common social impacts of a freeway construction (or extension) initiative.

Table 3: Selection of social impacts to consider when undertaking a road development initiative
Social impactIssues to consider
Displacement or isolation of residents Adequacy of the compensation and the relocation process, reduced land value, emotional issues including grief
Displacement or isolation of commercial and community facilities Adequacy of the compensation and the relocation process, economic hardships for existing or new businesses, reduced land value, clientele cut off, inaccessibility of services or inconvenience for customers
Barrier effects: effects on social interaction Effects on community cohesion, disruption of friendships or family contact, changes in convenience and travel time
Barrier effects: effects on business, recreation or services Inconvenience, changes to accessibility and travel time
Noise effects Physiological, psychological and social changes due to increased noise levels
Safety Effects on personal, family or child safety on a localised scale i.e. dependent on proximity to freeway or changes to traffic conditions in surrounding area
Health effects Physiological changes resulting from air and water quality
Environmental quality effects Changes in air or water quality as they affect people’s lifestyle and enjoyment of their environment, recreation, indoor and outdoor living
Land use changes Changes in zoning from residential to commercial areas or development in a previously undeveloped area, loss of recreational or public space
Aesthetics Changes to visual landscape, physical intrusion, scale, loss of open space, changes in flora or fauna
Cultural heritage Disturbance or destruction of heritage sites

The data required to facilitate a Social Impact Assessment process are firmly based on community consultation campaigns. Community participation is a major component of Social Impact Assessment. It is useful to begin the participation process early in the planning process and carry on throughout the life of individual initiatives. In many transport agencies, community participation/consultation is also a legislative requirement, meaning that an initiative cannot proceed beyond the planning stage without adequate consultation with the community. The support of the community is also often needed to ensure successful implementation of a transport initiative.

An added complication to impact assessment is that social impacts are classified differently by different practitioners. For example, air pollution is classified as an environmental issue in an Environmental Impact Assessment. A Social Impact Assessment should also include air pollution as a social issue because of its consequences on the health of the community. Air pollution mitigation would also be included in a CBA due to the economic costs of pollution mitigation measures.

The practitioner is often faced with a series of complexities inherent in impact assessment statements, which can lead to serious double counting issues in the economic appraisal of initiatives. It is very important to remember that a thorough appraisal should take into account a broad range of social impacts, not just those that are easily quantifiable and monetised such as relocation, pollution mitigation measures and safety, but also those that are more difficult to monetise such as community severance or loss of character or open space.

### 6.2.3 Equity Impact Assessment

Social (equity) Impact Assessment statements consider the winners and losers of the particular initiative investment. As stated by Levinson (2002), a set of specified (winner and loser) population subgroups would be normally identified. Then the outcomes of the initiative (e.g. travel time and delay, accessibility, consumer surplus, air pollution, noise pollution, accidents) would be assessed for each of these population subgroups.

Levinson (2002) provides an Equity Impact Statement checklist as shown below. The checklist includes a range of stratification variables (for example population, gender or spatial extent), specific process requirements (such as the opportunity to participate in decision-making) as well as desired outcome areas (such as mobility, economic, environmental and health outcomes) for transport initiatives.

 Process Outcomes Stratification Opportunity to engage in decision-making process Mobility Economic Environmental Health Other Population Spatial (or jurisdictional) Temporal Modal Generational Gender Racial Ability Cultural Income

Source: Levinson, 2002.

### 6.2.4 Assessing cumulative impacts

The distribution of effects can change over time and through the cumulative effects of successive initiative activities. Transport practitioners involved in equity analysis should therefore be aware of procedures for conducting Cumulative Effects Assessment (CEA) or Cumulative Impact Assessment (CIA).

A cumulative impact on a resource is one that results from the incremental impact of an action when added to other past, present and reasonably foreseeable future actions (see below). Cumulative impacts can result from individually minor but collectively significant actions taking place over a period of time. Cumulative impacts may also include the effects of natural processes and events, depending on the specific resource in question (FHWA, undated).

Cumulative impact analysis is resource-specific and generally performed for the environmental resources directly impacted by a government action under study, such as a transportation initiative. However, not all of the resources directly impacted by an initiative will require a cumulative impact analysis. The resources subject to a Cumulative Impact Assessment should be determined on a case-by-case basis early in the process, generally as part of early coordination or scoping (FHWA, undated).

It is generally recognised among practitioners that specific methodologies for the assessment of indirect and cumulative impacts, particularly for predicting reasonable foreseeable impacts, are not as well established or universally accepted as those associated with direct impacts, such as traffic noise analysis or wetland delineation. Determining the most appropriate technique for assessing indirect and cumulative impacts of a specific initiative should include communication with the cooperating agencies and community stakeholders (FHWA, undated).

Figure 2: Cumulative impactsSource: (FHWA, undated)

### 6.2.5 Stated preference surveys

Stated preference surveys are important community consultation tools that are used to inform equity evaluations (e.g. cost-utility analysis). They are particularly useful in situations where empirical information does not exist. For example, stated preference surveys might be used because no data has yet been generated on a new type of travel mode or a special type of pricing instrument with unique characteristics (US EPA, 1998).

In a stated preference approach, it is possible to derive statistical estimates of ‘trade-off’ rates between various alternatives or their attributes by making respondents choose from among them in measured ways that indicate the relative importance of key attributes. These rates can then be assessed in relation to each traveller and their circumstances (US EPA, 1998).

The validity of the derived statistical relationships relies on how well the alternatives are portrayed to (and understood by) the respondent, and their comparison with known ‘standards’. While stated preference surveys rely on hypothetical situations, comparison of ‘elasticity’ relationships derived from stated preference with more conventional revealed preference surveys or models have shown corroboration. The results from these surveys should be used with caution, but they offer an important interim tool for agencies to estimate relationships between pricing instruments and travel behaviour response, not just in mode choice but also in relation to destination, time of day, route choice, etc (US EPA, 1998).

Stated preference methods were developed by the private market research industry and have been used successfully for many years to aid companies in identifying the critical attributes of their product, and maximising those attributes to gain market share over competitors. Use of the techniques in transport is a fairly recent development; however, there are examples where they have been used to explore time of day choice or assist in the development of a route choice model (US EPA, 1998).

### 6.2.6 Spatial analysis techniques

This section discusses the potential of spatially based analysis and micro-simulation modelling to explore distributional or equity issues.

#### Spatially based analysis

Since transport infrastructure occurs on a spatial scale, it is usually the case that physical or social impacts resulting from transport impacts can also be quantified over a spatial scale. This is most commonly undertaken with Geographic Information Systems (GIS) technology which is now readily available and widely used to quantify various effects; for example, emission of environmental contaminants or noise modelling. Most transport impacts have a geographical component; for example, property prices can be easily represented in geographic form.

Once the distributional impact is defined over a geographical scale, relevant socio-economic characteristics need to be transposed onto the geographical representation of the impact. Some of these characteristics will be derived from a community social profile.

Due to the aggregate nature of common data sources on population characteristics (such as the census), Statistical Local Area or Local Government Area population characteristics are generally used as a proxy for specific groups being examined. For example, if concern is expressed over impacts on low income or minority populations, the impacts are measured for neighbourhoods that exceed a certain percentage of those population groups, rather than for specific minority persons or households. This provides the decision-maker with a representation of the distributional effects of initiatives on the communities of interest, i.e. the ‘winners’ and ‘losers’.

The biggest problem with spatial techniques is that some factors that affect impact distribution are difficult to determine. It is often difficult to identify the geographic location of a population class according to social characteristics. An additional complicating factor is that people’s decisions about where they live may be affected by transportation investments. For example, positive externalities such as good public transport or highway access can lead to higher property values and a migration of higher-income people to the area served (FHWA, 2003).

##### Micro-simulation

Micro-simulation modelling techniques forecast travel by modelling a set of actual or synthetic individuals or households that represent the population as the basic unit of analysis rather than dealing with population averages by postcode or statistical region. A ‘synthetic’ sample is composed of a hypothetical set of people or households with characteristics that as a whole match the overall population. Results are aggregated only after the individual or household analyses are completed, allowing the user great flexibility in specifying output categories. This is more commonly referred to as sample enumeration or discrete choice analysis. Sample enumeration relies on the modelling of behaviour for a representative sample of the population generally derived from a regional home interview survey or stated preference survey (FHWA, 2003).

The benefit of this modelling approach for analysing distribution of impacts is that travel patterns, and therefore the travel benefits of transportation improvements, can be tracked across any population characteristic that is included in the sample of persons modelled. Historically, this has been done by income level, since income is commonly used to predict travel behaviour. However the characteristics of the sample can be broadened to include other attributes (FHWA, 2003). An example of a micro-simulation program from the United States (STEP) program is presented below.

##### STEP: a micro-simulation program

STEP is a travel demand analysis package composed of an integrated set of travel demand and activity analysis models, supplemented by a variety of impact analysis capabilities and a simple model of transportation supply. STEP has been used by the US Department of Transport and the US Environmental Protection Authority to analyse travel impacts of pricing scenarios (with the intention to reduce transport emissions) by income group.

STEP program models are applied using actual or forecast data on household socioeconomic characteristics, the spatial distribution of population and employment (land use), and transportation system characteristics for the selected analysis year(s). STEP reads through the household sample, attaching level-of-service and land use data to each household record as necessary. For each household, STEP uses its models to predict a daily travel and activity pattern for each individual in the household. Finally, household travel is summed up and household totals are expanded to represent the population as a whole.

Testing the effect of a change in conditions or policies is a simple matter of re-analysing the household sample using the new data values, and comparing the results with previous outputs. For example, a new highway or new transit service can be represented by changed travel times and costs for the areas served; a parking price increase can be represented by an increase in out-of pocket costs; an increase in income in a particular area or for a particular population subgroup can be represented by editing the household file to incorporate the revised incomes.

The sampling framework preserves the richness of the underlying distribution of population characteristics and permits tabulation by any subgroup with sufficient observations to be statistically significant. For example, the results can be disaggregated by income level and age, which would allow an assessment of effects for, say, various income classes among the retired population. This is a significant advantage over an aggregate model, which uses zonal averages for most socioeconomic and economic data. A possible STEP model structure is illustrated below.

Figure 3: STEP model structureSource: US EPA, 1998

## 6.3 Appendix C Community participation processes

There are varying degrees of public participation; from information provision and consultation to substantial support for community initiatives (see figure below). Higher degrees of participation are not necessarily 'better' - different levels are appropriate for different situations and interests (Wilcox, 1994). The most commonly applied form of participation is community consultation.

Figure 4: Levels of community participationSource: Adapted from Wilcox, 1994

The desired level of participation needed for a initiative will inform the selection of participatory methods and techniques. Choice of method should directly reflect the type of information needed and the purpose for which it will be used. The following table provides common purposes for which community input is sought and the methods generally effective in achieving the task.

Table 5: Matching participatory instruments to purpose
Participatory approachCharacteristicsParticipants
Purpose: To gain ideas and input from the public
Public hearing/ community meeting A public hearing is often formal, with statements going into an official record of the meeting. A community meeting will often be an informal gathering where people come to share ideas with local officials. An open gathering of people from the community who wish to be heard about a topic or issue
Focus groups A small gathering of stakeholders who meet in a confidential setting to discuss an issue or react to a proposal. The assumption is that through discussion, new information will emerge that would not otherwise come to light from individual questioning. These meetings are often facilitated by a trained individual. Local officials may or may not actively participate in the discussion. Selected stakeholders
Purpose: To complete a specific task with citizen input
Workshop A meeting focused on a predetermined task to be accomplished. Rather than soliciting general opinion, workshops are intended to focus on specific concerns and produce a predetermined product. The benefit of such meetings is that those most directly affected by an issue are directly involved in addressing it. Primary stakeholders are often involved because of a high level of interest in the issue. To be most effective in addressing a public issue, the full range of interests should be represented in the workshop
Task force Purpose is to complete a clearly defined task in the planning process. A task force is often appointed to study a particular issue and offer a report of findings and recommendations to the policy-making body. A small (usually 8 to 20 people) ad hoc citizen committee
Purpose: To have a discussion about citizen priorities associated with community initiatives
Priority-setting committee Citizen group appointed to advise local officials regarding citizen ideas and concerns in planning community initiatives. Participants who are trusted to represent the concerns of citizens and sometimes function as a ‘go-between’ with residents and local government
Purpose: To discuss citizen priorities associated with community initiatives
Delphi procedure The objective is to work toward a consensus of opinion that can be used by policymakers for decision making. Successive rounds of presented arguments and counterpoints move the group toward consensus, or at least to clearly established positions and supporting arguments. A panel of citizens chosen for their knowledge about an issue
Purpose: To quickly and quietly ascertain public sentiment about an issue
Interviews, polls, and surveys Detailed information can be gathered. While confidential, the information can be informative both in content and overall emotional/political reaction to an issue. Interested citizens are given a chance to speak directly with someone about their views
Purpose: To gain input about the alternatives and consequences of an issue.
Media-based issue balloting Coupled with a media-based effort to discuss alternatives and consequences of potential solutions, letters to the editor or radio call-in shows are monitored to gain a sense of public reaction. Unscientific and not a reliable indicator of overall community sentiment, it can be a good way to gain a quick reaction to proposals by those most likely to be active on an issue. Citizens are asked to respond through the local media
Purpose: To give citizens broad decision-making powers
Citizen advisory boards or councils An advisory board studies an issue and makes recommendations to policy makers. The range of decision-making authority can vary and, in some cases, may be binding. Appointed representatives of one or more community interests
Referenda Direct and binding decision-making authority by the electorate. Protracted campaigning leading to a referendum can become a divisive force. All eligible voters
Purpose: To stay informed about the needs of certain neighbourhoods or interest groups
Group or neighbourhood planning council This council serves as advisory to policy makers. Such councils keep decision makers informed about neighbourhood or group concerns, formulate goals and priorities on behalf of the neighbourhood or group, and evaluate plans and programs affecting the neighbourhood or group. Organised by, and composed entirely of citizens

Source: Adapted from Leatherman and Howell, 2000

## 6.4 Appendix D A distributional rules approach

Khisty (1996) attempts to draw analytical conclusions about equity effects by using distributional rules or theories of justice that can be applied depending on the outcomes sought. In this approach, the analyst needs to determine which analytical framework is the most appropriate for the situation under investigation.

This involves the application of different equity principles or theories to determine the types of outcomes that are possible or desirable. Theories of justice are used as input in the development of decision-making procedures. There is no one single theory of justice that will satisfy everyone. For example, Khisty (1996) provides the following six theories of justice chosen because they represent ideas that are either commonly used, understood by society or are documented in the literature.

To illustrate how theories of justice can be applied, Khisty (1996) developed an example of a hypothetical city showing six alternative bus configurations (1-6) as illustrated below. The income distribution (expressed from ‘low’ to ‘high’) on the route alternatives is then overlaid on the area map. Each alternative satisfies the goals and objectives set forth by the citizens of the city, and in each case the aggregate benefits exceed the aggregate costs.

Figure 5: A hypothetical city showing six bus transit configurationsSource: Khisty (1996)

There are five major socio-economic groups in the city and their population percentages are indicated in the table below. It is assumed that each group contributes taxes to the city in proportion to their income. The amount indicated under each alternative (1-6) represents units of benefit that each individual would receive.

 Alternatives 1 2 3 4 5 6 Total Net Benefits 920 1045 1825 1885 2200 2450 Income class % population High 5% 6 9 25 28 35 50 Medium high 10% 7 11 22 25 30 40 Medium 50% 9 11 19 20 25 30 Medium low 25% 10 10 16 15 15 10 Low 10% 12 9 13 12 10 6 Average net benefit 9.20 10.45 18.25 18.85 22.00 24.50 Floor 6 9 13 12 10 6 Range 6 2 12 16 25 44

Source: Khisty (1996)

Given the details of the initiative, the question is: which of the six alternatives is the most equitable? The answer to this question depends on which distributional rules or equity principles the decision-maker adopts. Khisty (1996) provides the implications for route selection based on each of the six equity principles:

• Equal share distribution (distribution based on an equal share - or as equal as possible - of the benefits among the socioeconomic groups). Alternative 2 is most consistent with this principle with the minimum range between the highest and the lowest benefit received being 2 units and an average net benefit received of 10.45 units.
• Utilitarian distribution (distribution based on maximising the benefits to the community as a whole). Alternative 6 is most consistent with this principle. While the disparity between high-income and low-income groups is glaring, this alternative has the highest net benefit among all the alternatives.
• Distribution based on maximising the average net benefit with a minimum floor benefit of 10 units (this principle ensures that an attempt to maximise the average benefit is constrained by a certain amount to ensure that certain individuals or groups, particularly those ‘at the bottom’, receive a certain minimum amount of benefit). Alternative 5 is consistent with this principle. The choice of a minimum floor is a decision that must be made in advance by the decision-maker. This principle also illustrates the nature of an efficiency-equity trade-off; the principle is achieved with a reduction in total net benefits of 250 units compared with the maximum efficiency alternative.
• Distribution based on maximising the average net benefit with a benefit range constraint not exceeding 16 units (this principle ensures that an attempt to maximise the average benefit does not allow differences in benefit between the rich and the poor segments of the society to exceed a certain amount). Alternative 4 is consistent with this principle. As above, an efficiency-equity trade-off is apparent. In this case, 565 units of net benefit need to be traded-off.
• Distribution based on the egalitarian principle (this principle of ethical conduct attempts to reduce any existing social or economic inequalities among individuals and groups in the community). Alternative 1 distributes higher benefits to the lower end of the income distribution and is therefore consistent with the egalitarian principle. Although this alternative has the lowest total benefit of all alternatives, it probably benefits income groups that are truly in need of public transportation.
• Rawls’ theory of justice (distribution based on maximising benefits to the lowest income group). Alternative 3 is consistent with this principle. It also has the highest floor among the alternatives, but indicates a need of 625 units of net benefit to be traded for the desired equity outcome.

Which distribution theory to use will depend on the policy-maker and the characteristics of the community that is represented. Invariably, when people are affected by the choice of distribution rules, or when they are offered several rules from which to choose, they tend to prefer the rule that favours them. Preferences are a function of culture, political affiliations, gender, economic standing and so forth (Khisty, 1996).

Khisty (1996) suggests that citizens are generally not bothered by ethical theories as much as they are concerned with their own welfare in terms of ‘quality of life’. Therefore, Khisty defines ‘quality of life’ as the essence of the collective economic, social and physical conditions of people in a community.

It is important to recognise that these are highly subjective choices. They involve trade-offs between, on the one hand, the efficiency focus of increasing the net benefits to society as a whole and, on the other hand, striving for more equitable outcomes. For transport and infrastructure planners and analysts, it is also essential to note that in Australia the taxation and welfare system is the prime policy tool for addressing issues of inequality.

## 6.5 Appendix E Equity Considerations in road pricing

This section provides a discussion of equity issues associated with road pricing. The purpose is to illustrate how equity considerations are a key component of transport policy decision making. An example of providing for equity in road pricing is provided from the European Communities’ AFFORD Project.

Over recent years, the concept of road pricing has been gaining momentum due to concerns about road capacity and congestion management. However, there is still a great deal of controversy surrounding the wider introduction and application of road pricing. As stated by Stough et al (2004), there are misunderstandings over what road pricing seeks to do, concerns over how the revenues will be spent and issues relating to welfare distribution (equity) consequences.

Road pricing is intended to improve transport efficiency by rationing road capacity. In terms of reducing travel demand and making traffic flow more efficient, it does not matter how road pricing revenue is allocated. From an overall economic perspective, the revenue must be used to benefit society and the greater the benefit the more economically efficient the program. There is no requirement, however that the money be allocated in any particular way (Litman, 1999).

The major equity consideration of road pricing concerns the distribution of road pricing revenue. Two components of equity that need to be considered regarding road pricing are horizontal equity and vertical equity. Many people instinctively feel that horizontal equity implies that revenues should be dedicated to road improvements or to provide other benefits to people who pay the fee. However, horizontal equity is complicated by the existence of external costs – those that are borne by non-vehicle users (see table below). So horizontal equity is only fulfilled when revenue is returned to vehicle users as a class, but only after external costs are compensated. Since most estimates of motor vehicle external costs are larger than the expected revenue of road pricing proposals, the horizontal equity justification of returning revenues to drivers is reduced or eliminated (Litman, 1999).

The vertical equity component is more complex. Vertical equity requires that disadvantaged people receive more public resources (per capita or unit of service) than those who have a relative advantage, to accommodate their greater need. So revenues must benefit low-income drivers as a class at least as much as the costs they bear, and disadvantaged residents (including non-drivers) must benefit overall. Litman (1999) explains that vertical equity can be defined with respect to the ability to drive. As a class, non-drivers tend to be economically or socially disadvantaged. Road pricing has the potential of benefiting non-drivers overall by increasing the use of alternative travel modes. Vertical equity considerations justify using road pricing revenue in a broad range of ways including the support of alternative transport programs, reduction in taxes, or funding of public services that benefit disadvantaged populations. The table below illustrates an approach developed by Litman (1999) to assess the distribution of road pricing revenues to four classes of people based on horizontal and vertical equity considerations.

Table 7: Road pricing revenue distribution equity analysis
ClassDescriptionHorizontal equityVertical equity
Non-drivers People who cannot drive, usually due to age, disability, or low income. Non-drivers use automobiles as passengers, but their overall use of congested roads is typically low. Although this group would pay little in road pricing, they deserve a share of revenue if it is considered compensation for existing external impacts of driving. Non-drivers include many people who are economically, physically and socially disadvantaged; therefore, maximum use of road pricing revenues to benefit this group is justified.
Low-income drivers People who can drive and have access to an automobile, but whose travel decisions are significantly affected by vehicle expenses. They will be frequently tolled off by road pricing. This group pays a relatively small share of road pricing fees, but incurs costs from travel charges that provide a large portion of congestion reduction benefits. They deserve a share of toll revenues in compensation. This group is, by definition, disadvantaged so use of road pricing revenues to benefit this group is justified.
Middle-income drivers People who can drive and have access to an automobile, but whose travel decisions are only moderately affected by vehicle expenses. They will sometimes be tolled off the roadway and their net benefits of travel are reduced by road pricing. These drivers pay a large portion of total road pricing and lose net benefits. They deserve to benefit from road pricing revenues on the basis of horizontal equity, but only after all external costs are compensated. Since this group is not disadvantaged there is no vertical equity justification for using road pricing revenue to benefit them.
Upper-income drivers People who can drive and have access to an automobile, but whose travel decisions are not affected by vehicle expenses. They benefit overall from road pricing due to reduced congestion. These people enjoy net benefits from reduced congestion. They deserve a share of the revenue only after external costs are compensated. Since this group is not disadvantaged, there is no vertical equity justification for using road pricing revenue to benefit them.

Source: Litman, 1999

### 6.5.1 Road use charges: an example from the AFFORD project

The European Commission undertook a study of marginal cost transport pricing in three European cities – Helsinki, Oslo and Edinburgh – as part of the ‘Acceptability of Fiscal and Financial Measures and Organisational Requirements for Demand Management’ (AFFORD) study (Fridstrom et al, 2000).

The study distinguished between ‘first-best’ and ‘second-best’ road pricing policy packages. The first-best solution involves charging the user the true cost, i.e. the marginal cost of road use determined by the level of congestion, environmental and accident costs. The second-best pricing package was based on the use of a package of policy instruments that are available for use by transport authorities (e.g. time differentiated cordon toll rates or time differentiated parking charges) (Fridstrom et al, 2000).

The study concluded that inequity within a population increased when road pricing is implemented (based on a Gini coefficient defined in terms of household income per consumption unit before and after revenue redistribution). However, in most cases the changes to income distribution appeared to be relatively moderate.

Fridstrom et al (2000) noted that if revenue is redistributed proportionately by personal income, which is given as a percentage point relief in the income tax rate, it does nothing to correct the initial, adverse equity effects between people in the different income brackets. It does, however, reverse the potentially unpopular transfer of funds from private consumers to the public treasury.

However, if the same, absolute amount of money is redistributed to each adult individual (a ‘poll transfer’ or ‘flat distribution’) income inequity in the population improves considerably. According to model simulations, this is because the out-of-pocket expenditure on road charges represents a higher share of the household income in low income groups than among the more affluent. Both of these scenarios represent clear trade-offs between equity and efficiency: equity can be improved by redistribution but only at the expense of the efficiency gains from the road pricing strategy.

Fridstrom et al (2000) suggest that in principle it is possible to conceive of a road-pricing scheme with revenue redistribution that enhances economic efficiency as well as equity. It will usually be sufficient to redistribute a certain component of the revenue generated in a progressive manner, in order to keep the less affluent households at least equally well off. The main reason why road pricing schemes do not lead to any deterioration in income distribution is that the more affluent people, exhibiting higher rates of car ownership and use, tend – in general – to incur higher road pricing expenditure.

### 6.5.2 Non-pricing mechanisms for providing equity in road use

While road pricing is one method of rationing road capacity, there are other transport demand management mechanisms that do not involve pricing. These include priority measures such as high occupancy vehicle lanes and alternative rationing schemes. Travel behaviour change initiatives are another non-pricing mechanism to improve equity by encouraging more efficient modes of transport and better access for people without a vehicle. These measures are aimed at reducing total vehicle traffic and encouraging the use of efficient modes. Many of these strategies support equity objectives by improving travel choices/alternatives or affordability, especially for low income or mobility-disadvantaged groups (Litman, 2000).

Australia currently has a number of high occupancy vehicle lanes, commonly referred to as ‘transit lanes’. Transit lanes provide travel priority by allowing specified users (usually two or more people per private vehicle and public transport vehicles) exclusive use of part of the roadway to travel through congested sections of road. Transit lanes provide a high degree of horizontal equity (because they do not discriminate in regard to who can participate). This option benefits all existing users, especially public transport users by reducing travel times.

Road rationing schemes designate a certain percentage of the travelling population to use a road link on certain days or times of day. Those who have not been designated to use the road link at a particular time may still do so upon payment of a toll. Rationing schemes have been applied in many countries, for example in Athens and several Brazilian cities, with varied results. In these cities, access prohibitions have led to increased multiple car ownership and average fleet age, and after some years they lose their effectiveness (Viegas, 2001). Because of these results Viegas (2001) suggests that the ‘ration’ should be attributed to individuals, not to vehicles, so it is useable for driving and for riding on public transport (this also serves as an incentive to shift to public transport). Attributing the ration to individuals instead of vehicles prevents misuse of the system by those who own more than one car (Viegas, 2001). Nevertheless, rationing schemes are associated with high administration costs and are open to abuse by both users and administrators.

These are ‘second-best’ options because of administrative, spatial or other deficiencies. However, under certain scenarios, they provide a valid response to tackling complex equity issues.